JP2006017633A - Stereoscopic photograph analyzer - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To easily prepare a survey drawing in a stereoscopic photograph analyzer to enhance survey precision. <P>SOLUTION: An index is photographed by a stereo camera having a parallax. The analyzer reads a plurality of paired photographed photographs in. Two sets of paired photographs are displayed on connection photograph display areas CPD1, CPD 2 of a monitor. Image coordinates of the index and a mark are detected when an automatic button ATB is pushed. Coordinates of the same indexes in two images are correlated based on the same mark, when pushing a correlation button. A conversion parameter is calculated based on the coordinates of the indexes in two images, when pushing a calculation starting button PSB. The coordinates of a survey point in the two sets of paired photographs conversion-parameter-calculated thereon are converted by the conversion parameter. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to image processing in photogrammetry.

  Photogrammetry is widely used to create maps and is used as an extremely effective means for recording local situations such as on-site verification of traffic accidents. Conventional photogrammetry uses a stereo camera in which two cameras are fixed apart from each other, and calculates three-dimensional coordinates of each survey point from a pair of pair images taken by both cameras. In such a photogrammetry method, a paired image is read by a stereo photographic analyzer, and the three-dimensional coordinates of survey points that are commonly captured in the paired image are obtained to create a survey map. When surveying covers a wide range, a large number of pair images are used, and survey maps obtained for each pair image are connected.

  In order to easily connect survey maps, the marks on two or more point indicators are copied to both of the two pairs of images, and 2 in the stereo photo analyzer by manual operation using an input means such as a mouse. A survey map is connected by designating a mark imprinted on a pair image (see Patent Document 1).

However, when the number of pair images to be connected increases, the manual operation for specifying the mark is complicated, and it takes time to connect. Further, since the connection accuracy affects the overall survey accuracy, the overall survey accuracy is also affected by the skill level of the operator who performs the connection.
Japanese Patent Laid-Open No. 2001-21351

  Accordingly, an object of the present invention is to provide a stereo photograph analyzer that can easily and quickly connect a large number of survey maps obtained by pair images taken by a stereo camera.

  The stereo photo analyzer of the present invention is an acquisition means for acquiring, as digital data, a group of photographs taken by a stereo camera having two or more cameras and including a plurality of indices, and each photograph of the indices in each photograph of the photograph group. A recognition means for recognizing coordinates based on the image, and an index that is a coordinate of the index based on the shooting position at which the photograph group was photographed based on the coordinates based on the respective photographs for the index recognized by the recognition means In a coordinate system based on a second imaging position that is different from the first index coordinates and the first index coordinates that are index coordinates in the coordinate system based on the first imaging position for the same index Corresponding means for associating the second index coordinates, which are the index coordinates, for each index, and the first index coordinates and the second finger for the same index associated by the associating means In a single first and second reference coordinate system based on coordinates in the coordinate system based on the first imaging position and coordinates in the coordinate system based on the second imaging position based on a plurality of combinations with coordinates Conversion value calculation means for calculating first and second conversion parameters for conversion into coordinates.

  Identification information that includes different identification information for each index, the recognition means recognizes the identification information for each index in association with the index coordinates corresponding to the index, and the association means associates with the first index coordinates recognized by the recognition means It is preferable that the association is performed based on a combination of the same identification of the first identification information and the second identification information that is the identification information associated with the second index coordinates. And a monitor for displaying a photograph image included in the photograph group, and an input means for designating an arbitrary point or region on the photograph image included in the photograph group displayed on the monitor. It is preferable that the first identification information and the first index coordinates or the second identification information and the second index coordinates are recognized in the area designated by or in the vicinity of the designated area.

  Alternatively, the image processing apparatus includes a monitor that displays a photo image included in the photo group, and an input unit that specifies an arbitrary point or region on the photo image included in the photo group displayed on the monitor, and the association unit inputs Photographs included in the second photograph group photographed from the second photographing position with the same index as the index within the area designated on the image of the photograph included in the first photograph group photographed from the first photographing position by the means It is preferable that the first index coordinates and the second index coordinates for the same index are associated with each other by designating a region on the image. Further, it is preferable that the recognition means recognizes the first identification information and the first index coordinates or the second identification information and the second index coordinates in the area designated by the input means or in the vicinity of the designated area.

  Further, the surveying means is a survey point coordinate based on the photographing position at which the photograph group was photographed on the basis of the coordinates based on each photograph at the point designated by the input unit as the survey point to be surveyed It is preferable to provide conversion means for calculating point coordinates and converting the survey point coordinates to coordinates in the reference coordinate system based on the conversion parameters calculated by the conversion value calculation means.

  For each index, the second index coordinate for the same index is associated with the first index position and the third index coordinate which is the index coordinate in the coordinate system based on the third imaging position different from the second imaging position. Coordinates based on the second imaging position based on a plurality of combinations of the second index coordinates and the third index coordinates for the same index associated by the association means by the conversion value calculation means. Second and third conversion parameters for converting the coordinates in the system and the coordinates in the coordinate system based on the third shooting position into the coordinates in a single second and third reference coordinate system as a reference can be calculated; It is preferable to provide composite conversion means for converting the coordinates in the coordinate system based on the third imaging position to the coordinates in the first and second reference coordinate systems based on the first and second conversion parameters and the second and third conversion parameters. Arbitrariness.

  Furthermore, each index includes different identification information, the recognition unit recognizes the identification information for each index in association with the index coordinate corresponding to the index, and the association unit is associated with the second index coordinate recognized by the recognition unit. The association is preferably performed based on a combination of the same identification of the second identification information that is identification information and the third identification information that is identification information associated with the third index coordinates.

  Alternatively, the image processing apparatus includes a monitor that displays a photo image included in the photo group, and an input unit that specifies an arbitrary point or region on the photo image included in the photo group displayed on the monitor, and the association unit inputs Photographs included in the third photograph group photographed from the third photographing position with the same index as the index in the area designated on the image of the photograph included in the second photograph group photographed from the second photographing position by the means It is preferable that the second index coordinates and the third index coordinates for the same index are associated with each other by designating an area on the image.

  It is preferable to include a selection unit that selects the first photo group and the second photo group from a plurality of photo groups acquired from the acquisition unit, and the selection unit detects an input in the selection input unit that performs image selection input. It is preferable to select the first photograph group and the second photograph group. Further, the acquisition unit acquires the shooting direction of the stereo camera at the time of shooting detected by the orientation sensor provided in the stereo camera in association with the shot photo group, and the selection unit sets the first photo group selected first. It is preferable to select a group of photographs associated with a direction substantially the same as the associated direction as a candidate for selection of the second group of photographs. Alternatively, the acquisition unit acquires the shooting time detected by a timer provided in the stereo camera in association with the shot photo group, and the selection unit continuously acquires the shooting time associated with the first photo group selected first. It is preferable to list the photo group associated with the shooting time to be selected as the second photo group selection candidate.

  It is preferable to include a selection unit that selects the first photo group and the second photo group from a plurality of photo groups acquired from the acquisition unit, and the selection unit detects an input in the selection input unit that performs image selection input. It is preferable to select the first photograph group and the second photograph group. Further, the acquisition unit acquires the shooting direction of the stereo camera at the time of shooting detected by the orientation sensor provided in the stereo camera in association with the shot photo group, and the selection unit sets the first photo group selected first. It is preferable to select a group of photographs associated with a direction substantially the same as the associated direction as a candidate for selection of the second group of photographs. Alternatively, the acquisition unit acquires the shooting time detected by a timer provided in the stereo camera in association with the shot photo group, and the selection unit continuously acquires the shooting time associated with the first photo group selected first. It is preferable to list the photo group associated with the shooting time to be selected as the second photo group selection candidate.

  Preferably, the acquisition means acquires digital data of a group of photographs taken by a digital camera, or acquires digital data by reading an image of a group of photographs that is a silver-lead photograph with a scanner and performing A / D conversion. .

  Further, it is preferable that the identification information has a different color for each index, or is an identification character, figure, or symbol written in the vicinity of the index.

  The point indicator of the present invention includes a plate member that can be fixed in a substantially upright state, and an index that is provided on a side surface of the plate member at opposite positions on the front and back sides.

  It is preferable that an identification character, figure, or symbol is written in the vicinity of the index. Or it is preferable that it is a different color for every parameter | index.

  The photogrammetry method of the present invention includes a plurality of indicators having an index on an effective surveying range, which is a photographing range for satisfying a design surveying accuracy in photogrammetry using a stereo camera having a plurality of cameras at any photographing position. A step of arranging the point indicator at the survey point so as to include a plurality of point indicators, a first photographing position for including a plurality of point indicators in the effective surveying range of the stereo camera, and A first positioning step for determining a first shooting posture of the stereo camera; a first shooting step for acquiring first image data by shooting with the stereo camera at the first shooting position determined in the first positioning step; A plurality of point indicators included in the effective surveying range in the first shooting posture at the first shooting position moved from the first shooting position of the stereo camera A second positioning step for determining a second shooting position and a second shooting posture to be included in the effective measurement range, and a second image data obtained by shooting with a stereo camera at the second shooting position determined in the second positioning step. Using the stereo photograph analyzer of the present invention on the basis of the second photographing step for acquiring the first image data obtained in the first photographing step and the second image data obtained in the second photographing step. And an analysis step.

  In the arrangement step, the effective surveying ranges are arranged so as to include two or more point indicators. In the first positioning step and the second positioning step, two or more point indicators are included in the effective surveying range of the stereo camera. It is preferable that the inclination of the stereo camera with respect to the horizontal plane when photographing in the two photographing steps is substantially the same as the inclination when photographing at the first photographing position, or three or more point indicators in the effective surveying range in the arranging step. It is preferable that three or more point indicators are included in the effective surveying range of the stereo camera at the first photographing position in the positioning step.

  Further, the point indicator in the arranging step is the point indicator of the present invention, and in the second positioning step, the direction of the stereo camera in the first photographing posture is opposite to the direction of the stereo camera in the second photographing posture. preferable.

  The photogrammetry program of the present invention is an acquisition means for acquiring, as digital data, a group of photographs taken by a stereo camera having two or more cameras and including a plurality of indices, and each photograph of the indices reflected in each photograph of the photograph group. Recognizing means for recognizing the coordinates based on the reference, and index coordinates that are the coordinates of the index based on the shooting position at which the group of photographs was taken based on the coordinates based on the respective photographs for the indices recognized by the recognizing means. The calculation means for calculating and the index coordinates in the coordinate system based on the second indexing position different from the first indexing position and the first indexing coordinate in the coordinate system based on the first imaging position for the same index The association means for associating the second index coordinates for each index, and the first index coordinates and the second index for the same index associated by the association means In a single first and second reference coordinate system based on coordinates in the coordinate system based on the first imaging position and coordinates in the coordinate system based on the second imaging position based on a plurality of combinations with the standard. The computer is caused to function as conversion value calculation means for calculating first and second conversion parameters for conversion into coordinates.

  According to the present invention, the stereo photo analyzer can automatically designate the mark of the point indicator, and it is possible to shorten the survey time and maintain the survey accuracy at a high level regardless of the surveyor's skill. Become.

Hereinafter, embodiments of the present invention will be described with reference to the drawings.
FIG. 1 is a diagram showing a survey vehicle used for photographing in the photogrammetry method according to the first embodiment of the present invention.

  The surveying automobile 40 is provided with a photographing device 41 which is a stereo camera on the upper surface. The photographing device 41 is configured to include digital cameras 43 and 44 on the left and right sides of the base line rod 42 held substantially horizontally. The left and right cameras 43 and 44 are fixed so that their optical axes are substantially parallel and face the same direction. The direction in which the photographing device 41 faces can be adjusted. The left and right cameras 43 and 44 are simultaneously photographed by the release operation of the operator.

  Note that the photographing apparatus 41 may include a plurality of cameras other than the left and right cameras 43 and 44. If the position and orientation are fixed with respect to the left and right cameras 43 and 44, photogrammetry can be performed using photographs of the other cameras and the left and right cameras 43 and 44. When performing photogrammetry with a plurality of cameras, it is possible to perform surveying points with a group of photos taken from the same shooting position instead of a paired photo. Further, the cameras do not have to be arranged on the left and right, for example, above and below, and photogrammetry is possible as long as they capture the same subject and have parallax.

  FIG. 2 is a diagram showing a photographing situation in the photogrammetry method according to the first embodiment of the present invention, and is a plan view depicting the outline of a road to be surveyed and a white line.

  When photographing using the automobile 40 at the surveying point, the parallax of the pair photograph taken in the fixed photographing posture is generated in the horizontal direction. It is possible to measure the subject that appears in the central area (shooting area) where the shooting range of the left camera 43 and the shooting range of the right camera 44 overlap. On the other hand, the surveying accuracy allowed in design in the photogrammetry by the photographing apparatus 41 is determined. In order to ensure this surveying accuracy, the distance (surveying distance) from the photographing position to the subject that can be surveyed is determined. The imaging area within the survey distance range becomes the effective survey range (the shaded portion in FIG. 2), and photogrammetry of survey points appearing within the effective survey range is possible.

  It should be noted that parameters such as the focal length of the left and right cameras 43 and 44, the distance between the left and right cameras 43 and 44, (base line length), resolution, distortion and the like are accurately measured, and the allowable measurement accuracy for these parameters. The survey distance that can be secured is determined. In the present embodiment, the effective surveying range is an imaging region in which the depth from the left and right cameras 43 and 44 is within a range of approximately 30 m, for example.

  The point indicator PS is arranged on a substantially straight line on the road where surveying is performed. The automobile 40 is moved from the first indicator to the first imaging position RP1 where the fourth indicators PS1, PS2, PS3, PS4 are included in the first effective surveying range MA1 of the imaging device 41. After the first shooting posture, which is the direction and inclination of the shooting device 41, is adjusted and determined, shooting is performed. Photos taken by the left and right cameras 43 and 44 are first paired photos, and first image data corresponding to the first paired photos is acquired and stored in the memory card 15 (see FIG. 4).

  As shown in FIG. 3, the indicator PS is formed with an indicator 32 and a mark 35 that is identification information of the indicator 32 on the side surface of the quadrangular pyramid body member 31. The side surface of the main body member 31 has a single color, and the indicator 32 is formed in a round shape with a single color different from that of the main body member 31. Therefore, the color around the index 32 is different from the color of the index 32. The mark 35 is different for each index 32 and is provided in the vicinity of the index 32.

  In FIG. 2, after photographing at the first photographing position, the vehicle 40 is moved to the second photographing position RP <b> 2 by moving about 10 m along the arrangement of the indicator PS. The second imaging posture is adjusted so that the second indicator to the fifth indicator PS2, PS3, PS4, PS5 are included in the second effective surveying range MA2 of the imaging device 41. Therefore, the indicators PS2, PS3, PS4 that fall into the first effective surveying range MA1 at the first shooting position RP1 are included in the second effective surveying range MA2 at the second shooting position RP2. Photographing is performed at the second photographing position RP2, and the second image data is stored in the memory card 15. Similarly, the automobile 40 is moved, and photographing is performed repeatedly while changing the photographing position until all the surveying points are included in the effective surveying range.

  The image data acquired as described above is supplied to a stereo photograph analyzer to which the first embodiment of the present invention to be described later is applied, and surveying is performed to connect survey results for each image data, that is, survey coordinates. And survey maps are easily connected.

  The photographing device 41 is provided with a timer, an orientation sensor, and a tilt sensor (not shown), and the time, orientation, and tilt when photographing at each photographing position are detected. It is stored in the memory card 15 in association with the image data at the shooting position.

  In the present embodiment, four or more indicators PS are arranged so as to be included in the effective surveying range at each imaging position, and the imaging positions and attitudes are set so that four or more indicators PS are included in the effective surveying range. However, when correcting the shooting angle by the tilt sensor, or when shooting with a constant tilt of the shooting device 41, it is sufficient to include at least three indicators PS. In FIG. 2, three or more indicators PS are arranged so as to be included in the effective surveying range at one imaging position, and the imaging positions and imaging for including three or more indicators PS in the effective surveying range. You just have to decide your posture. Since three or more indicators PS are included in the effective surveying range, three or more indicators PS are shown in the pair photo.

  When changing the shooting position while keeping the direction of the shooting device 41 substantially constant in order to connect three or more image data, it is convenient to sequentially change the three indicators PS to be copied as the basis of the connection. Further, it is convenient to arrange the indicators PS along the direction of connection. For example, when the first pair photo to the third pair photo ahead of the traveling direction are taken while the automobile 40 is moving, the sign PS PS that appears in front of the first pair photo is not captured in the second pair photo. On the other hand, another indicator PS is newly included in the effective surveying range of the second pair photo. The same applies to the second pair photograph and the third pair photograph.

  Considering the second pair photo, the first pair photo is connected by the three front indicators PS, and the third pair photo is connected by the three back indicators PS. If the two middle signs PS are overlapped, the first pair photo and the two last signs PS that overlap with the innermost sign PS by the two signs PS that overlap with the frontmost sign PS. Can be connected to the third pair photo. Therefore, it is convenient to include at least four indicators PS in the effective surveying range.

  In the present embodiment, the first effective surveying range MA1 and the second effective surveying range MA2 include three identical indicators PS, but the inclination of the imaging device 41 with respect to the horizontal plane is substantially the same at each imaging position. Or when correcting using the inclination calculated | required with the tilt sensor, it is sufficient if the two same indicator PS is included. For example, when the inclination angle of the photographing device 41 with respect to the horizontal plane is kept constant and a planar road is surveyed, the connection can be made by two indicators PS. Therefore, a plurality of indicators PS may be included in the effective survey range for the photogrammetry in the present embodiment.

  Next, the configuration of the stereo photo analyzer to which the first embodiment of the present invention is applied will be described. FIG. 4 is a block diagram showing the configuration of the analyzer. The stereo photo analyzer 10 includes a CPU 11, a card reader 12, and a monitor 13. The CPU 11, card reader 12, and monitor 13 are connected to the bus 14 directly or indirectly. The operation of the entire apparatus is controlled by the CPU 11. Image data or the like is read from the memory card 15 by the card reader 12. Some or all of the pair photographs corresponding to the image data can be displayed on the monitor 13.

  Further, the photo analysis apparatus 10 is provided with a mouse 16 and a keyboard 17 used for image point designation and various instructions. The mouse 16 and the keyboard 17 are connected to the bus 14 via the input control unit 18, and inputs from the mouse 16 and the keyboard 17 are transferred to the CPU 11 via the bus 14, and their input modes and the like are set.

  The memory card 15 is inserted into the card reader 12, and a plurality of image data stored in the memory card 15 is read into the device. That is, it is possible to acquire image data corresponding to a pair photo obtained by photographing. The reading operation and writing operation of the card reader 12 are controlled by a recording medium control unit 19 connected to the bus 14. A hard disk 20 is connected to the recording medium control unit 19. The hard disk 20 acquires and stores image data read from the memory card 15 together with shooting related data.

  Next, image connection in the analyzer 10 will be described. The photo analyzer 10 has a three-dimensional coordinate calculation function and a three-dimensional coordinate connection function. These functions work together to achieve image connection. With the three-dimensional coordinate calculation function, three-dimensional coordinates based on the shooting position of the pair photo are calculated for an arbitrary survey point within the effective survey range of the pair photo pair. With the three-dimensional coordinate connection function, the three-dimensional coordinates based on the photographing positions of the two pairs of pair photographs are converted into three-dimensional coordinates in a single coordinate system. The three-dimensional coordinate connection function is executed by the following conversion parameter calculation operation and coordinate conversion operation.

  The three-dimensional coordinates calculated for each pair photo are based on a three-dimensional coordinate system based on a shooting position that is different for each pair photo. In order to convert a plurality of three-dimensional coordinate systems into a single coordinate system, movement conversion for matching the origins and rotation conversion for matching the coordinate axes may be performed. Note that the three-dimensional coordinate system based on the shooting position is based on a position relatively determined by the positions of the left and right cameras 43 and 44 at each shooting position, and any three directions intersecting with each other at the origin are coordinate axes. Coordinate system.

  Conversion parameters for performing movement / rotation conversion are calculated by a conversion parameter calculation operation. By the coordinate conversion operation, the three-dimensional coordinates of the survey points in the plurality of coordinate systems are converted into the three-dimensional coordinates in the single three-dimensional coordinate system based on the conversion parameters.

  Next, specific operations for executing these functions will be described. First, the three-dimensional coordinate calculation function in the analyzer 10 will be described. FIG. 5 is a diagram showing a screen for displaying a read photograph on the monitor. The read photo display screen is displayed on the display surface of the monitor 13 by the operation of the analyzer 10.

  When a read button RDB displayed on the upper left of the display surface is pressed, image data stored in the hard disk 20 or the memory card 15 is read into the CPU 11. In the read photo display area APD, one of the pair photos corresponding to the image data stored in the hard disk 20 or the memory card 15 is displayed.

  One of the pair photos is displayed side by side in two rows, and character information can be displayed on the right side. As this character information, for example, an image data number 50 which is a number indicating the order of image data read by the CPU 11, information corresponding to shooting related data, shooting time 51 and shooting direction 52, etc. are displayed. The The image data number 50 is assigned to each image data in the order read by the CPU 11. A read button RDB, a survey button MSB, a parameter calculation button PMB, and a connection button CNB are displayed above the read photo marking area APD.

  When the survey button MSB is pressed on the read photo display screen, it is possible to select a pair of pair photos for calculating the three-dimensional coordinates of the survey point. When a pair photo is selected with the mouse 16 or the like, an operation screen for surveying as shown in FIG. 6 is displayed on the display surface of the monitor 13. The selected pair photo is displayed in the selected pair photo display area SPD on the upper side of the monitor 13.

  A left image IML that is an image of a photograph taken by the left camera 43 is displayed on the left side, and a right image IMR that is an image of a photograph taken by the right camera 44 is displayed on the right side. The operator can designate any point on the left image IML and the right image IMR by operating the mouse 16. A survey point for surveying is designated by the operator on the left image IML and the right image IMR.

  Based on the designated survey point, the three-dimensional coordinates of the survey point based on the position where the pair photograph was taken are calculated as described below. By specifying the survey point, the image coordinates of the survey point in each of the left image IML and the right image IMR are recognized. Since the left camera 43 and the right camera 44 have parallax, based on the image coordinates in the left image IML and the image coordinates in the right image IMR of the survey point, the shooting position where the pair photo was taken by the principle of triangulation A reference three-dimensional coordinate is obtained.

  For example, a survey point corresponding to one point P1 on the same road is designated as a point P1L in the left image IML, and a point P1R is designated in the right image IMR, whereby the three-dimensional coordinates of the survey point P1 are calculated. By designating the same survey point (P2, P3, P4,...) On both images IML and IMR, the three-dimensional coordinates of each survey point are calculated.

  In the survey map display area GVD below the monitor 13, each survey point is displayed on the two-dimensional coordinate based on the calculated three-dimensional coordinate of each survey point. The two-dimensional coordinates are the coordinates of a plane when the survey point is viewed from above. A plurality of survey points displayed in the survey map display area GVD can be connected with a straight line or a curve by operating the mouse 16. By specifying the outer edge of the road as a survey point and connecting the survey points, a survey map of the road in the effective survey range of the pair photo is drawn. Although FIG. 6 shows a method of inputting the left and right images independently, the image may be input using a conventionally known stereoscopic input device using a deflection filter and deflection glasses.

  When surveying and plotting for a pair of paired photos are completed, the read photo display screen is displayed again. As shown in FIG. 5, a symbol 53 indicating that surveying has been completed is added to the character information on the right side for displaying one of the paired photos of the paired photo for which the surveying operation has been completed.

  As described above, the three-dimensional coordinate calculation function is executed. Next, the conversion parameter calculation operation will be described.

  In FIG. 5, when the operator operates the mouse 16 or the like and presses the parameter calculation button PMB, it is possible to select two sets of pair photographs for obtaining conversion parameters. Two pairs of photographs are designated by operating the mouse 16 or the like. When the first pair photo is selected, selection candidates for the second pair photo to be connected are displayed based on the shooting related data associated with the image data of the first selected pair photo.

  For example, the outer frame of one of the two pairs of photographs taken before and after the time when the pair photograph selected first is taken is displayed in red in the read photograph display area APD. These two pairs of photos are listed as candidates for selecting a pair photo to be connected to the first selected pair photo. Normally, photographs that are connected to each other are continuously photographed, so photographs that have consecutive photographing times are considered as connection candidates.

  Alternatively, the outer frame of one of the pair photos taken in the substantially same orientation when the first selected pair photo was taken is displayed in blue in the read photo display area APD. Therefore, these pair photos are listed as selection candidates for pair photos to be connected to the first selected pair photo.

  As shown in FIG. 2, there are two rows in which the indicator PS is arranged, and the pair of photos are moved without moving the shooting direction at each shooting position along one row. There is. In such a case, since the photographs connected to each other are taken in substantially the same direction, photographs taken in the same direction are considered as connection candidates. In the case where the selection candidate is selected depending on the shooting time and direction, the outer frame of one of the paired photos is displayed in green in the read photo display area APD.

  Two pairs of photographs are selected: a first pair photograph photographed at the first photographing position RP1 and a second pair photograph photographed at a second photographing position RP2 different from the first photographing position RP1. Selection completion or reselection can be input by operating the mouse 16 or the like, and when the selection is completed, a screen for performing a conversion parameter calculation operation is displayed. If reselection is input, selection can be made again. It should be noted that the second pair photograph to be selected must include a part or all of the same index 32 as the index 32 shown in the first pair photo.

  FIG. 7 is a diagram showing a sign detection operation screen for conversion parameter calculation operation. The image of one photo of the selected first pair photo is displayed in the connection photo display area CPD1 on the left side of the display surface of the monitor 13, and the image of one photo of the second pair photo is displayed in the connection photo display region CPD2 on the right side. The An automatic button ATB, a semi-automatic button HAB, an association start button CSB, and a calculation start button PSB are displayed above the connection photograph display areas CPD1 and CPD2.

  When the automatic button ATB is pressed, the automatic mode, that is, the position and range of the indicator PS displayed in the connected photograph display areas CPD1 and CPD2 are automatically detected. Detection of the sign PS is performed by image recognition such as known feature extraction. When the detection of the sign PS is finished, the detected sign PS is displayed blinking on each image. When a subject other than the sign PS is detected as the sign PS, an input for excluding the subject other than the sign PS can be performed by operating the mouse 16 or the like. Note that the position and range of the indicator PS on the photo that is the same pair as one photo displayed in the regions CPD1 and CPD2 and is not displayed is also detected. In addition, when using a stereoscopic analysis device, it is displayed on a screen that is viewed stereoscopically.

  Further, when the semi-automatic button HAB is pressed, it can be detected in the semi-automatic mode. In the semi-automatic mode, first, the position of the indicator PS is roughly designated by the operator in the connection photograph display areas CPD1 and CPD2. The designation may be one point or one area in the vicinity of one indicator PS in the areas CPD1 and CPD2. When the position of the sign PS is roughly designated, the vicinity of the designated point or area, or the position and range of the sign PS within the area are detected. The semi-automatic mode is ended by inputting detection end by operating the mouse 16 or the like.

  According to the automatic mode, when the appearance of the subject in the effective surveying range is similar to the sign PS, the subject may be detected as the sign PS. According to the semi-automatic mode, it is possible to accurately detect only the sign PS.

  When the sign PS is detected in the automatic mode or the semi-automatic mode, the mark 35 described for each sign PS is recognized in each of the connected photograph display areas CPD1 and CPD2. Further, the image coordinates of the index 32 provided for each indicator PS, that is, the coordinates based on each photograph constituting the pair photograph are recognized. The image coordinates of the index 32 are recognized in association with the mark 35 described for each indicator PS, that is, the mark 35 corresponding to the index 32. The detection of the mark 35 is performed by image recognition such as publicly known feature extraction, and the detection of the image coordinates is performed by image recognition such as publicly known luminance center of gravity detection.

  Further, based on the detected image coordinates of the index 32, first index coordinates and second index coordinates, which are three-dimensional coordinates based on the first imaging position RP1 and the second imaging position RP2, are calculated. The calculation of the three-dimensional coordinates is automatically performed by executing the above-described three-dimensional coordinate calculation function. Further, the first mark associated with the image coordinates used for the calculation of the first index coordinates is also directly associated with the first index coordinates. Similarly, the second index coordinates are similarly associated with the second mark.

  When the detection of the sign PS, the recognition of the mark 35, and the calculation of the three-dimensional coordinates of the index are completed, the association start button CSB can be pressed. When the association start button CSB is pressed, the first index coordinates and the second index coordinates are associated with the same index 32. The association is performed by associating a combination in which the first mark 35 associated with the first index coordinates and the second mark 35 associated with the second index coordinates are the same. An association number 54 is assigned to each index 32 that has been associated.

  When the association process is completed, the association number 54 of the index 32 associated with the association display area CRD below the connection photograph display area CPD1 is displayed as shown in FIG. When each association number 54 is designated by the operation of the mouse 16 or the like, the indicator 32 corresponding to the association number 54 blinks in the areas CPD1 and CPD2, and the association can be confirmed. When the association is incorrect, the association can be corrected by operating the mouse 16 or the like, and the association can be excluded from the conversion parameter calculation described later.

  When the confirmation of the association is finished and the parameter calculation button PSB is pressed, the second coordinate system is a three-dimensional coordinate system based on the first photographing position RP1 from the second coordinate system that is a three-dimensional coordinate system based on the second photographing position RP2. A 1-2 conversion parameter for conversion to the first coordinate system is calculated. The first-second conversion parameter is calculated by a successive approximation method based on a plurality of combinations of the first index coordinates and the second index coordinates associated with the same index 32.

  When calculation of the conversion parameter is completed, the read photo display screen is displayed again. When the conversion parameter is calculated as shown in FIG. 5, the image data number 50 of the other paired photo linked by the conversion parameter is added as character information displayed on the right side of one of the paired photos. Since the conversion parameters of the first pair photo and the second pair photo were calculated, (2) is one of the pair photos of the image data number 2 in the column of the photo display of one of the pair photos of the image data number 1. (1) is displayed in the display column. Similarly, two pairs of pair photographs to be connected are selected, and a conversion parameter is calculated for each pair photograph.

  As described above, after the conversion parameter calculation work and the three-dimensional coordinate calculation function are executed, the coordinate conversion operation can be performed. When the connection button CNB is pressed on the read photo display screen (see FIG. 5), a three-dimensional coordinate connection screen (see FIG. 9) is displayed. A combined display area CGD is provided above the display surface of the monitor 13. In the combination display area CGD, an image data number 50 of a pair photo that has been surveyed and for which a conversion parameter with another pair photo has been calculated is displayed. The image data number 50 is displayed in association with the image data number 50 of another pair photo that can be converted.

  For example, the first pair photo and the second pair photo, the second pair photo and the third pair photo, the third pair photo and the fourth pair photo, and the fifth pair photo and the sixth pair photo can be converted. If calculated, the combination display area CGD displays the combination of the image data numbers “1-2-3-4” and the combination of the numbers “5-6”. It can be seen that a pair photo or a fifth pair photo and a sixth pair photo can be connected. The shooting positions at which the first pair photo, the second pair photo, the third pair photo, the fourth pair photo, the fifth pair photo, and the sixth pair photo are taken are different from each other.

  When a combination to be connected is selected and conversion is started by operating the mouse 16 or the like, the three-dimensional coordinates of the survey point based on the shooting position where each pair photo included in the combination is taken is a single three-dimensional Converted to the coordinates of the reference coordinate system, which is the coordinate system. Note that the coordinate conversion is calculated based on the conversion parameters described above. The survey map created for each pair photo is connected and displayed in the survey map display region GVD below the combination display region CGD. Each survey point is displayed in the survey map. It is also possible to create a survey map of the road by connecting survey points in the region GVD.

  Note that connection of three or more pair photographs, for example, connection of the first pair photograph, the second pair photograph, and the third pair photograph can be performed as described below. The 1-2 conversion parameters of the first pair photograph and the second pair photograph are calculated as described above. Similar to the 1-2 conversion parameter, the 2-3 conversion parameter for converting the coordinates of the survey point in the third coordinate system with the third imaging position as a reference into the coordinates of the second coordinate system is calculated. When the first coordinate system is a reference coordinate system that is a single three-dimensional coordinate system, the coordinates of the survey point in the second coordinate system are the first to second conversion parameters, and the coordinate of the survey point in the third coordinate system is the first coordinate system. It is possible to perform conversion using the 1-2 conversion parameter and the 2-3 conversion parameter.

  Next, referring to the flowcharts of FIGS. 10 to 14, the pair photo connection processing by the CPU 11 will be described.

  FIG. 10 is a flowchart showing a main routine showing a processing procedure of the whole connection work. In step S100, image data of a pair photograph is read from the card reader 12 or the left and right cameras 43 and 44. In step S101, it is confirmed whether there are two or more read pair photographs. If only one pair photo is read, the connection is not possible and the process ends.

  If there are two or more pairs of photographs, the process proceeds to step S102, a read photograph display screen is displayed on the monitor 13, and the process proceeds to step S103. In step S103, the presence / absence of input from any one of the survey button MSB, the parameter calculation button PMB, and the connection button CNB is detected. When the survey button MSB is pressed, the survey subroutine of step S200 is executed to calculate the three-dimensional coordinates of the survey point with reference to the shooting position for each paired photo, and the process returns to step S102.

  When the parameter calculation button PMB is pressed in step S103, the index association subroutine in step S300 is executed, and then the parameter calculation subroutine in step S400 is executed. A conversion parameter for conversion to the coordinate system is calculated, and the process returns to step S102.

  When the connection button CNB is pressed in step S103, the three-dimensional coordinate conversion subroutine of step S500 is executed to convert the three-dimensional coordinates of the survey points in the coordinate system for each paired photograph into a single three-dimensional coordinate. When the three-dimensional coordinate conversion subroutine ends, the pair photo connection processing ends.

  FIG. 11 is a flowchart showing the surveying subroutine. In step S201, an input for selecting a pair of pair photographs by the operator is detected. In step S202, a surveying operation screen is displayed on the display surface of the monitor 13, and the image of the left camera 43 and the image of the right camera 44 of the pair photo detected in step S201 are displayed in the selected photo display area SPD. In the next step S203, the internal localization elements and external localization elements of the left and right cameras 43, 44 are read. In addition, since the imaging device 41 is a baseline length fixed system, an internal localization element may be preset.

  Proceeding to step S204, an input for designating survey points in the left image IML and the right image IMR by the operator is detected. In the next step S205, image coordinates in the designated left image IML or right image IMR are detected based on the survey point designation detected in step S204. Next, in step S206, based on the image coordinates detected in step S205 and the internal localization elements and external localization elements read in step S203, three-dimensional coordinates based on the position where the pair photograph was taken are calculated.

  Proceeding to step S207, two-dimensional coordinates obtained by viewing the three-dimensional coordinates calculated in step S206 from above the survey point are calculated, and the process proceeds to step S208. In step S208, the two-dimensional coordinates of the survey point calculated in step S207 are displayed in the survey map display area GVD. In step S209, it is checked whether there is an input for ending surveying or an input for designating another survey point. If there is an input for designating another survey point, the process returns to step S205. The processing from step S205 to step S209 can be repeated until the three-dimensional coordinates and two-dimensional coordinates of the surveying points to be surveyed are obtained.

  If it is determined in step S209 that surveying has been completed, the process proceeds to step S210. In step S210, it is confirmed whether or not there is an input connecting the survey points to the survey points in the survey map display area GVD. If there is no input connecting the survey point to the survey point in step S210, the survey subroutine ends. If there is an input to be connected on the other side, the process proceeds to step S211, and a straight line or a curve connecting the survey points and the survey points in the survey map display area GVD is displayed.

  In step S212, it is confirmed whether there is an input for ending the drawing or another input for connecting the survey points. If there is another input, the process returns to step S211 and the processes of steps S211 to S212 are repeated. When the end of drawing is input in step S212, the surveying subroutine ends, and the process returns to step S102 of the main routine.

  FIG. 12 is a flowchart showing the index association subroutine. In step S301, an input for selecting the first pair of photographs by the operator is detected. Next, the process proceeds to step S302, and the second pair photo selection candidate is displayed on the display surface of the monitor 13 on the basis of the shooting related data of the pair photo detected in step S301. In the next step S303, an input for selecting the second pair of pair photographs is detected, and the process proceeds to step S304.

  In step S304, it is confirmed whether or not to confirm the selection of two pairs of pair photos. If reselection is input, the selection is not confirmed and the process returns to step S301, and steps S301 to S304 are repeated. When the selection confirmation input is performed, the process proceeds to step S305.

  In step S305, a sign detector detection operation screen is displayed, and the process proceeds to step S306. In step S306, it is determined which of the automatic button ATB and the semi-automatic button HAB is pressed. If the automatic button ATB is pressed, the process proceeds to step S307, where the image coordinates and range of the indicator PS in the images displayed in the connected photograph display areas CPD1 and CPD2 are detected, and the process proceeds to step S311.

  When the semi-automatic button HAB is pressed in step S306, the process proceeds to step S308, and a position roughly designated on the connection photograph display areas CPD1 and CPD2 by the operator is detected. Next, in step S309, the image coordinates and range of the indicator PS within the range of the region roughly designated in step S308 or in the vicinity of the designated position are detected, and the process proceeds to step S310.

  In step S310, whether or not another sign PS is detected is determined based on an input from the operator. That is, when the operator inputs again the rough designation of the position of the sign PS, the process returns to step S308. Thereafter, the semi-automatic detection process of the sign PS in steps S308 to S310 is repeated. If the sign detection end is input, the process proceeds to step S311.

  In step S311, the mark 35 marked on the indicator PS detected automatically or semi-automatically is identified, and the image coordinates of the indicator 32 similarly provided on the indicator PS are detected, and the process proceeds to step S312. In step S312, based on the image coordinates of the index 32 detected in step S311, three-dimensional coordinates for each index 32 with respect to each shooting position are calculated. Thereafter, the process proceeds to step S313, and the association operation screen is displayed.

  In the next step S314, based on the mark 35, the three-dimensional coordinates based on the respective shooting positions of the first pair photo and the second pair photo are associated with each other for the same index 32. An association number 54 is assigned for each index. Next, in step S315, the association number 54 is displayed for each associated index 32. Thereafter, the index association subroutine ends, and the processing of the parameter calculation subroutine in step S400 is started.

  Next, a process for calculating the conversion parameter after describing the conversion parameter will be described.

  The first pair photo and the second pair photo, which are two pairs of photos to be connected, are taken at the first shooting position RP1 and the second shooting position RP2, respectively. When performing photogrammetry without surveying the absolute three-dimensional position of the stereo camera, the survey points that appear in the first pair of photographs are surveyed in the first coordinate system based on the first imaging position. Similarly, the survey points appearing in the second pair photo are surveyed in the second coordinate system with the second photographing position as a reference.

As described above, the first coordinate system and the second coordinate system are mutually converted by movement and rotation. The coordinates of the origin of the second coordinate system in the first coordinate system are (X ₂₀ , Y ₂₀ , Z ₂₀ ), and the inclination of the coordinate axis of the second coordinate system with respect to the coordinate axis of the first coordinate system is (ω, φ, κ) In this case, all the coordinates of the second coordinate system are translated to the coordinates of the first coordinate system by moving in parallel at (X ₂₀ , Y ₂₀ , Z ₂₀ ) and rotating at (ω, φ, κ). Is done. Therefore, (X ₂₀ , Y ₂₀ , Z ₂₀ , ω, φ, κ) is a conversion parameter to be obtained.

  FIG. 13 is a flowchart showing a parameter calculation subroutine. First, in step S401, a calculation start input by pressing the calculation start button PSB is detected, and the process proceeds to step S402.

In step S402, appropriate initial values (X _G20 , Y _G20 , Z _G20 , ω _G , φ _G , κ _G ) are approximated to the conversion parameters (X ₂₀ , Y ₂₀ , Z ₂₀ , ω, φ, κ). give. Next, in step S403, the index a in the first coordinate system is determined from the coordinates (X _2ai , Y _2ai , Z _2ai ) of the index a _i (i = 1, 2, 3) in the second coordinate system using the given initial value. Approximate coordinates of _i (X _GC1ai , Y _GC1ai , Z _GC1ai ) are calculated.

ここで行列｛Ｔ_jk｝は回転行列であり、各成分Ｔ_jkは例えば次式で表される。
Ｔ₁₁＝ｃｏｓφ・ｃｏｓκ
Ｔ₁₂＝ｃｏｓω・ｓｉｎκ＋ｓｉｎω・ｓｉｎφ・ｃｏｓκ
Ｔ₁₃＝ｓｉｎω・ｓｉｎκ−ｃｏｓω・ｓｉｎφ・ｃｏｓκ
Ｔ₂₁＝−ｃｏｓφ・ｓｉｎκ
Ｔ₂₂＝ｃｏｓω・ｃｏｓκ−ｓｉｎω・ｓｉｎφ・ｓｉｎκ
Ｔ₂₃＝ｓｉｎω・ｃｏｓκ＋ｃｏｓω・ｓｉｎφ・ｓｉｎκ
Ｔ₃₁＝ｓｉｎφ
Ｔ₃₂＝−ｓｉｎω・ｃｏｓφ
Ｔ₃₃＝ｃｏｓω・ｃｏｓφ That is, the coordinates of the index a _i in the first coordinate system _{(X C1ai, Y C1ai, Z} C1ai) the coordinates of the index a _i in the second coordinate system _{(X 2ai, Y 2ai, Z} 2ai) by equation (1) Therefore, the approximate values (X _G20 , Y _G20 , Z _G20 , ω _G , φ _G , κ _G ) of the conversion parameters and the coordinates (X _2ai , Y _2ai , Z _2ai ) of the index a _i in the second coordinate system are _obtained. ) _Is substituted into equation (1), approximate coordinates (X _GC1ai , Y _GC1ai , Z _GC1ai ) of the index a _i in the first coordinate system can be obtained.

Here, the matrix {T _jk } is a rotation matrix, and each component T _jk is expressed by the following equation, for example.
T ₁₁ = cosφ · cosκ
T ₁₂ = cosω · sinκ + sinω · sinφ · cosκ
T ₁₃ = sinω · sinκ−cosω · sinφ · cosκ
T ₂₁ = −cosφ · sinκ
T ₂₂ = cosω · cosκ−sinω · sinφ · sinκ
T ₂₃ = sinω · cosκ + cosω · sinφ · sinκ
T ₃₁ = sinφ
T ₃₂ = −sinω · cosφ
T ₃₃ = cosω ・ cosφ

すなわち本実施形態においてメリット関数Φは、同一の指標について関連付けられた、実際に認識された第１座標系における指標の座標（Ｘ_1ai、Ｙ_1ai、Ｚ_1ai）と第２座標系における指標の座標（Ｘ_2ai、Ｙ_2ai、Ｚ_2ai）から近時的に求めた（Ｘ_GC1ai、Ｙ_GC1ai、Ｚ_GC1ai）との距離の２乗の和である。 In the next step S404, whether or not the approximate translation amount of the second coordinate system relative to the first coordinate system and the inclination of the coordinate axes (X _G20 , Y _G20 , Z _G20 , ω _G , φ _G , κ _G ) are appropriate. A merit function Φ for determining is calculated. The merit function Φ is defined by, for example, equation (2).

In other words, in this embodiment, the merit function Φ is related to the same index, and the actually recognized index coordinates (X _1ai , Y _1ai , Z _1ai ) in the first coordinate system and the index coordinates in the second coordinate system. The sum of the squares of the distances to (X _GC1ai , Y _GC1ai , Z _GC1ai ) recently obtained from (X _2ai , Y _2ai , Z _2ai ).

Next, in step S405, it is determined whether or not the merit function Φ is smaller than a predetermined value. In other words, the index a _i calculated based on the approximate translation amount of the second coordinate system and the inclination of the coordinate axes (X _G20 , Y _G20 , Z _G20 , ω _G , φ _G , κ _G ). The three-dimensional coordinates (X _GC1ai , Y _GC1ai , Z _GC1ai ) in the approximate first coordinate system of the three-dimensional coordinates (X _1ai in the first coordinate system obtained from the first pair photograph of the index a _i , Y _1ai , Z _1ai ).

If it is determined that Φ ≧ predetermined value, the process proceeds to step S406, where the approximate translation amount of the second coordinate system and the inclination of the coordinate axes (X _G20 , Y _G20 , Z _G20 , ω _G , Φ _G , κ _G ), correction amounts (δX ₂ , δY ₂ , δZ ₂ , δω, δφ, δκ) are obtained by, for example, the least square method.

In the next step S407, based on the correction amounts (δX ₂ , δY ₂ , δZ ₂ , δω, δφ, δκ) calculated in step S406, the translation amount of the second coordinate system and the inclination of the coordinate axis (X _G20 , Y _G20 , Z _G20 , ω _G , φ _G , κ _G ). That is, (X _G20 , Y _G20 , Z _G20 , ω _G , φ _G , κ _G ) are respectively converted into (X _G20 + δX ₂ , Y _G20 + δY ₂ , Z _G20 + δZ ₂ , ω _G + δω, φ _G + δφ, κ _G + Δκ). Thereafter, the process returns to step S403, and steps S403 to S407 are repeatedly executed until it is determined in step S404 that Φ <predetermined value.

On the other hand, if Φ <predetermined value in step S405, the process proceeds to step S408, and the currently given translation amount of the second coordinate system relative to the first coordinate system and the inclination of the coordinate axes (X _G20 , Y _G20 , Z _G20 , ω _G , φ _G , κ _G ) are determined as conversion parameters that are the amount of translation of the second coordinate system relative to the first coordinate system and the inclination of the coordinate axes. Thereafter, the parameter calculation subroutine ends, and the process returns to step S102 of the main routine.

  FIG. 14 is a flowchart showing a three-dimensional coordinate conversion subroutine. First, in step S501, an input of connection start by pressing the connection button CNB is detected, and the process proceeds to step S502 to open the three-dimensional coordinate connection screen. Further, the process proceeds to step S503, and a pair photo combination that can be connected is displayed in the connectable combination display area CGD. Thereafter, in step S504, an input for selecting a pair photo pair to be connected is detected. If an input is detected, it will progress to step S505.

  In step S505, the three-dimensional coordinates of the survey points obtained in each pair photo included in the selected combination are converted into coordinates in a single three-dimensional coordinate system. In step S506, a road survey map is displayed in the survey map display area GVD based on the coordinates of the survey points in the single three-dimensional coordinate system obtained by the conversion. The above processing is executed, the three-dimensional coordinate conversion subroutine is finished, and the main routine is also finished.

  The conversion parameters are six unknowns, and a solution can be obtained with six equations. However, since the three unknowns are rotation angles, the six equations can have two solutions for each rotation angle. Since a set of conversion parameters that can convert all survey points in the second coordinate system to coordinates in the first coordinate system, three equations are necessary to obtain a single rotation angle solution. Therefore, at least nine equations are required.

  When one index 32 is connected, three equations are obtained. Therefore, in order to calculate a single conversion parameter, it is necessary that three or more identical indicators 32 are included in the effective surveying range of the two pairs of paired photos to be connected. Note that when the inclination of the photographing device 41 with respect to the horizontal plane is constant at each photographing position, the two rotation angles are constant. Therefore, the conversion parameters are four unknowns including one unknown rotation angle. In this case, at least five equations are required. Therefore, when the inclination of the photographing device 41 with respect to the horizontal plane is constant at each photographing position, it is sufficient that two or more identical indexes are included.

  With the configuration as described above, according to the analyzer to which the present embodiment is applied, a plurality of pairs are converted by converting the coordinates in the coordinate system for each pair photo obtained by the pair photo survey into coordinates in a single coordinate system. It is possible to connect surveying maps obtained from photographs and surveying points accurately and easily.

  In the analyzer to which the present embodiment is applied, the second coordinate system is converted to the first coordinate system, but a 2-1 conversion parameter for converting the first coordinate system to the second coordinate system is calculated, The coordinates of the survey point in the first coordinate system may be converted into the coordinates of the second coordinate system. Further, the coordinates of the survey point in the first coordinate system and the coordinates of the survey point in the second coordinate system may be converted into coordinates of a single three-dimensional coordinate system different from the first coordinate system and the second coordinate system. That is, the first and second conversion parameters to be converted into the first and second reference coordinate systems, which are the first coordinate system, the second coordinate system, or the single three-dimensional coordinate system, are calculated and converted into the single three-dimensional coordinate system. Any configuration can be used. This is because the purpose of this embodiment is to perform surveying in a single coordinate system. The connection is made in the same way when there are four or more coordinate systems to be connected.

  In addition, when converting to a coordinate system different from the three-dimensional coordinate system based on the shooting position of the pair photo to be connected by the calculated conversion parameter, connection of three or more pair photos is performed as follows. . First, first and second conversion parameters for converting the first coordinate system and the second coordinate system to the first and second reference coordinate systems are calculated. Similarly, second and third conversion parameters for converting the second coordinate system and the third coordinate system into the second and third reference coordinate systems are calculated. The coordinates of the survey point in the third coordinate system are converted to the second and third reference coordinate systems, then converted to the second coordinate system, and further converted to the coordinates in the first and second reference coordinate systems.

  Further, in the analyzer 10 to which this embodiment is applied, the mark 35 is used as the identification information of the index 32, but may be a character, a figure, or a symbol that can be distinguished from each other. Alternatively, a configuration may be adopted in which association is performed by using different colors for each index, and colors recognized in the indicator detection operation.

  Moreover, in the analyzer 10 to which the present embodiment is applied, image data that is digital data is directly acquired from a stereo camera that is a digital camera, but a silver lead photograph taken by a silver lead camera is scanned by a scanner (not shown). ) And the image signal read by the scanner is converted into digital data by an A / D converter (not shown), and the image data may be acquired.

  Moreover, in the analyzer 10 to which this embodiment is applied, a plane survey map can be created based on the three-dimensional coordinates of the surveyed survey points, but a stereo survey map may be created. .

  It should be noted that a configuration may be adopted in which images of both the paired photographs are displayed in the read photograph display area APD or the connected photograph display areas CPD1 and CPD2.

  In addition, the analyzer to which this embodiment is applied can be applied to a general-purpose personal computer, and the photo analysis program to which the embodiment of the present invention is applied is stored in a recording medium and is read by the personal computer. It is possible to function as a machine.

  Next, according to the analyzer to which the second embodiment of the present invention is applied, it is possible to connect a pair photo even when a sign PS not marked with a mark 35 is photographed. In this case, after the position and range of the sign PS are detected, the association designation screen is displayed as shown in FIG. The indicator PS detected on the image displaying the first pair photograph and the second pair photograph is indicated by blinking display or the like.

  One of the indicators PS detected on the image displaying the first pair photograph is selected by specifying an area by operating the mouse 16 or the like of the operator. Furthermore, the same indicator PS as the selected indicator PS is selected from the indicator PS detected on the image displaying the second pair photograph by area designation. Based on the selected indicator PS, the first index coordinates and the second index coordinates for each same index 32 can be associated. Other configurations are the same as those of the analyzer 10 to which the first embodiment is applied.

  FIG. 16 is a diagram showing an index association subroutine (1300) of the analyzer 10 to which the second embodiment is applied. This subroutine corresponds to the index association subroutine in the first embodiment, and the processing from step S1301 to step S1310 corresponds to the processing from step S301 to step S310. In step S1311, it is determined whether or not the mark 35 of the indicator PS can be detected.

  If the mark 35 can be detected, the process proceeds to step S1312. The subsequent processing from step S1312 to step S1314 corresponds to the processing from step S311 to step S313, and the processing from step S1319 to step S1320 corresponds to the processing from step S314 to step S315. In other words, the processes in steps S1312-S1314, S1319, and S1320 are performed in the same manner as in the first embodiment.

  If the mark 35 cannot be detected in step S1311, the process proceeds to step S1315. In step S1315, the image coordinates of the index 32 provided on the indicator PS are detected, and the process proceeds to step S1316. In step S1316, based on the image coordinates of the index 32 detected in step S1315, three-dimensional coordinates for each index 32 with respect to each shooting position are calculated. Thereafter, the process proceeds to step S1317, and the association operation screen is displayed.

  Next, in step S1318, an input by the operator's operation of the mouse 16 or the like that designates an area including the same indicator PS on the image of the first pair photograph and the image of the second pair photograph is detected. Proceeding to the next Step S1319, the subsequent processing is the same as in the first embodiment.

  Next, a photogrammetry method according to a second embodiment of the present invention and a point indicator as an embodiment of the present invention will be described. 17-19 is a figure which shows the point indicator which is one Embodiment of this invention. 17 is a front view, FIG. 18 is a side view, and FIG. 19 is a rear view.

  The indicator PS is composed of a base 33 and a plate member 34. The plate member 34 is supported by the base 33 and is fixed in a substantially upright state. An index 32F is provided at the center of one surface (reference F) of the plate member 34. An index 32B is also provided at the center of the other surface (reference B). The index 32F and the index 32B are opposed to each other on the front and back of the plate member 34. In the vicinity of the indicators 32F and 32B, a mark 35 which is identification information of the indicators 32F and 32B is written. In addition, the structure which makes the color of the parameter | index 32 different for every indicator PS and uses the color of a different parameter | index as identification information may be sufficient.

  FIG. 20 is a diagram showing a photographing situation in the photogrammetry method according to the second embodiment of the present invention, and is a plan view depicting the outline of a road to be surveyed and a white line. In the present embodiment, a photograph is taken from two directions facing the surveying point. Photographing is performed at a photographing position and a photographing posture (symbol a) that include more than half of the road RD in the effective surveying range while proceeding along the road RD to be surveyed in one direction (symbol A). Next, while traveling on the road RD in the opposite direction (symbol B), taking the direction (symbol b) opposite to the shooting direction when traveling in the direction A as the shooting direction, the remaining half or more of the road RD is set as the effective surveying range. Shoot at the included shooting position.

  The indicator PS is arranged so that the indicator 32 can be seen from both the photographing directions for photographing (symbol a and symbol b). For example, an arrangement in which the orientations of the optical axes of the left and right cameras 43 and 44 and the surface of the plate member 34 with the index 32 are substantially orthogonal is preferable. However, since the image coordinates of the index 32 need only be recognized by the analyzer 10, it is sufficient that the plate member 34 substantially intersects with the direction of the optical axis. Other method steps are the same as in the first embodiment.

  By supplying the image data acquired as described above to the analyzer and connecting the images, it is possible to easily connect a pair photo taken from the opposite direction. When the range to be surveyed is wide, it is possible to connect a pair of photographs by taking pictures while reciprocating multiple times.

  According to the sign PS of the present embodiment, it is possible to increase the connection accuracy of a pair photo whose shooting direction is opposite. In the case of using a triangular pyramid type indicator with an index on the side, the position of the index 32 that appears in the pair photo taken in the shooting direction of the symbol a and the position of the index 32 that appears in the pair photo shot in the shooting direction of the symbol b Different from each other. Therefore, connection accuracy is reduced when these indicators are connected.

  On the other hand, since the indicator PS of the present embodiment includes the indicators 32F and 32B at positions opposite to each other of the plate member 34, the indicator 32 shown in the pair photograph taken in the shooting direction of the symbol a and the shooting direction of the symbol b. Since the positions of the indicators 32 appearing in the taken pair photographs are substantially the same, it is possible to keep the connection accuracy high. That is, since the plate thickness of the plate member 34, particularly the thickness of the indicator portion, can be reduced, the error caused by the inclination of the indicator PS with respect to the plate thickness and the traveling direction can be reduced. Further, it is possible to reduce the deviation without correcting the plate thickness in the traveling direction.

It is a figure which shows the motor vehicle for surveying used for photogrammetry. It is a figure which shows the photography situation in photogrammetry conceptually, Comprising: It is the top view which drew the outline and white line of the road which are survey objects. It is a perspective view of a marking device. It is a block diagram which shows the electrical constitution of the stereo photograph analyzer which is the 1st Embodiment of this invention. It is a conceptual diagram which shows the screen which displays a reading photograph. It is a conceptual diagram which shows the display screen when performing surveying from a pair photograph. It is a conceptual diagram which shows the display screen when calculating a conversion parameter. It is a conceptual diagram which shows the display screen after finishing matching. It is a conceptual diagram which shows the display screen when connecting a three-dimensional coordinate. It is a flowchart which shows the main routine of the connection process performed in CPU. It is a flowchart which shows a surveying subroutine. It is a flowchart which shows a parameter | index correlation subroutine. It is a flowchart which shows a parameter calculation subroutine. It is a flowchart which shows a three-dimensional coordinate transformation subroutine. It is a conceptual diagram which shows the display screen when calculating the conversion parameter of the stereo photograph analyzer which is the 2nd Embodiment of this invention. It is a flowchart which shows the parameter | index correlation subroutine performed in CPU of the stereo photograph analyzer which is the 2nd Embodiment of this invention. It is a front view which shows the point indicator which is one Embodiment of this invention. It is a side view which shows the point indicator which is one Embodiment of this invention. It is a rear view which shows the point indicator which is one Embodiment of this invention. It is a figure which shows notionally the imaging | photography condition in the photogrammetry method which is the 2nd Embodiment of this invention, Comprising: It is the top view which drawn the external shape and white line of the road which is a survey object.

Explanation of symbols

10 Stereo photo analyzer 11 CPU
13 monitor 16 mouse 32 index 34 plate member 35 mark 41 photographing device 43 left digital camera 44 right digital camera RD road PS point indicator MA1 first effective surveying range MA2 second effective surveying range RP1 first photographing position RP2 second photographing position P1, P2, P3, P4, ... Surveying point IML Left image IMR Right image

Claims (27)

Acquisition means for acquiring, as digital data, a group of photographs taken by a stereo camera having two or more cameras and in which a plurality of indices are reflected;
Recognizing means for recognizing coordinates on the basis of the respective photographs for the indicators appearing in the respective photographs of the photograph group;
A calculation means for calculating an index coordinate which is a coordinate of the index based on a photographing position where the photograph group was photographed, based on coordinates of the index recognized by the recognition means on the basis of the respective photographs; ,
A first index coordinate which is an index coordinate in a coordinate system based on the first imaging position for the same index, and an index coordinate in a coordinate system based on a second imaging position different from the first imaging position. An association means for associating two index coordinates for each index;
Based on a plurality of combinations of the first index coordinates and the second index coordinates for the same index associated by the associating means, the coordinates in the coordinate system with the first imaging position as a reference, and the Conversion value calculation means for calculating first and second conversion parameters for converting the coordinates in the coordinate system with the second imaging position as a reference into the coordinates in a single first and second reference coordinate system as a reference. Stereo photo analyzer characterized by that. The indicators each have different identification information,
The recognition means recognizes the identification information for each index in association with the index coordinates corresponding to the index,
The identification means recognized by the recognition means is first identification information that is identification information associated with the first index coordinates, and second identification information that is identification information associated with the second index coordinates. The stereo photo analyzer according to claim 1, wherein the association is performed based on a combination of the same indices with the two. A monitor for displaying an image of a photograph included in the photograph group;
An input means for designating an arbitrary point or region on the image of the photograph included in the photograph group displayed on the monitor;
The associating means sets the same index as the index in the area designated on the image of the photograph included in the first photograph group photographed from the first photographing position by the input means to the second photographing position. The first index coordinate and the second index coordinate for the same index are associated with each other by designating a region on a photograph image included in the second photograph group taken from Stereo photo analyzer described in 1. A monitor for displaying an image of a photograph included in the photograph group;
An input means for designating an arbitrary point or region on the image of the photograph included in the photograph group displayed on the monitor;
The recognizing unit recognizes the first identification information and the first index coordinate, or the second identification information and the second index coordinate in the region designated by the input unit or in the vicinity of the designated region. The stereo photograph analyzer according to claim 2.   The recognizing means recognizes the first identification information and the first index coordinates, or the second identification information and the second index coordinates in the area designated by the input means or in the vicinity of the designated area. The stereo photograph analyzer according to claim 3. Based on the coordinates based on the respective photographs for the points designated by the input means as surveying points to be surveyed by the calculation means, Calculate the survey point coordinates that are coordinates,
The stereo according to claim 3 or 4, further comprising conversion means for converting the survey point coordinates into coordinates in the reference coordinate system based on the conversion parameter calculated by the conversion value calculation means. Photo analysis machine. A third index which is an index coordinate in a coordinate system based on the second index coordinate for the same index and the third shooting position different from the first shooting position and the second shooting position; Coordinates can be associated with each indicator,
Based on a plurality of combinations of the second index coordinates and the third index coordinates for the same index associated by the association means, the conversion value calculation means is based on the second imaging position. The second and third conversion parameters for converting the coordinates in the coordinate system and the coordinates in the coordinate system based on the third shooting position into the coordinates in the single second and third reference coordinate system as a reference can be calculated. And
Based on the first and second conversion parameters and the second and third conversion parameters, composite conversion means for converting coordinates in the coordinate system with the third imaging position as a reference into coordinates in the first and second reference coordinate systems The stereo photograph analyzer according to claim 1, comprising: The indicators each have different identification information,
The recognition means recognizes the identification information for each index in association with the index coordinates corresponding to the index,
Second identification information, which is identification information associated with the second index coordinates, recognized by the recognition means, and third identification information, which is identification information associated with the third index coordinates. The stereo photograph analyzer according to claim 7, wherein the association is performed based on a combination of the same indices with. A monitor for displaying an image of a photograph included in the photograph group;
An input means for designating an arbitrary point or region on the image of the photograph included in the photograph group displayed on the monitor;
The associating means sets the same index as the index in the area designated on the image of the photograph included in the second photograph group photographed from the second photographing position by the input means to the third photographing position. 8. The second index coordinate and the third index coordinate for the same index are associated with each other by designating a region on a photograph image included in the third photograph group taken from Stereo photo analyzer described in 1.   The stereo photograph analyzer according to claim 1, further comprising selection means for selecting a first photograph group and a second photograph group from a plurality of photograph groups obtained from the obtaining means.   The stereo photograph analyzer according to claim 10, wherein the selection means detects an input in a selection input means for inputting an image selection and selects the first photo group and the second photo group. . The acquisition means acquires the shooting direction of the stereo camera at the time of shooting detected by an orientation sensor provided in the stereo camera in association with the shot photo group,
The selection means lists the photograph group associated with the orientation substantially the same as the orientation associated with the first selected photograph group as a selection candidate for the second photograph group. Stereo photo analyzer described in 1. The acquisition means acquires the shooting time detected by a timer provided in the stereo camera in association with the photographed group,
The selection means lists the photo group associated with the photographing time that is continuous with the photographing time associated with the first selected photo group as a selection candidate for the second photo group. Item 13. The stereo photo analyzer according to Item 10.   The stereo photograph analyzer according to claim 1, wherein the obtaining unit obtains digital data of the group of photographs taken by a digital camera.   The stereo photograph analyzer according to claim 1, wherein the obtaining unit obtains digital data by reading an image of the group of photographs that is a silver-lead photograph with a scanner and performing A / D conversion.   The stereo photograph analyzer according to claim 2 or 8, wherein the identification information has a different color for each index.   The stereo photograph analyzer according to claim 2 or 8, wherein the identification information is an identification character, a figure, or a symbol written in the vicinity of the index.   The stereo photograph analyzer according to claim 1, wherein the index is a single-colored round dot, and the periphery of the index of a point indicator provided with the index on a side portion is a color different from the index.   The stereo photograph analyzer according to claim 1, wherein the index is provided at opposite positions on the front and back of a plate member that can be fixed in a substantially upright state. A plate member that can be fixed in a substantially upright state, and an indicator provided on the side surface of the plate member at opposite positions on the front and back sides,
A point indicator characterized by being used for taking a photograph for surveying by the stereo photograph analyzer according to claim 1.   The point indicator according to claim 20, wherein a character, a figure, or a symbol for identification is written in the vicinity of the indicator.   21. The point indicator according to claim 20, wherein the color is different for each index. Include multiple point indicators with indicators on the surface in the effective surveying range, which is the shooting range to meet the design survey accuracy in photogrammetry with stereo cameras with multiple cameras at any shooting position An arrangement step of arranging the point indicator at a survey point so that
A first positioning step for determining a first photographing position and a first photographing posture of the stereo camera for including a plurality of the point indicators in the effective surveying range of the stereo camera;
A first photographing step of performing photographing by the stereo camera at the first photographing position determined in the first positioning step and acquiring first image data;
A second for moving from the first shooting position and including the plurality of point indicators included in the effective surveying range in the first shooting posture at the first shooting position in the effective surveying range of the stereo camera. A second positioning step for determining a photographing position and a second photographing posture;
A second shooting step of performing shooting by the stereo camera at the second shooting position determined in the second positioning step and acquiring second image data;
Based on the first image data acquired in the first shooting step and the second image data acquired in the second shooting step, surveying is performed using the stereo photograph analyzer according to claim 1. A photogrammetry method comprising: an analysis step to perform. In the arranging step, the effective surveying range is arranged so as to include two or more of the point indicators,
In the first positioning step and the second positioning step, the effective surveying range of the stereo camera includes two or more point indicators,
The photogrammetry method according to claim 23, wherein, in the second photographing step, an inclination of the stereo camera with respect to a horizontal plane when photographing is substantially the same as an inclination when photographing at the first photographing position.   In the arranging step, the effective surveying ranges are arranged so as to include three or more point indicators, and in the positioning step, three or more points are included in the effective surveying range of the stereo camera at the first photographing position. The photogrammetric method according to claim 23, comprising a sign.   The point indicator according to claim 20, wherein the point indicator in the arranging step is the stereo camera in the second photographing posture in which the direction in which the stereo camera faces in the first photographing posture in the second positioning step. The photogrammetric method according to claim 23, wherein the photogrammetric method is opposite to the direction of. Acquisition means for acquiring, as digital data, a group of photographs taken by a stereo camera having two or more cameras and in which a plurality of indices are reflected;
Recognizing means for recognizing coordinates on the basis of the respective photographs for the indicators appearing in the respective photographs of the photograph group;
A calculation means for calculating an index coordinate which is a coordinate of the index based on a photographing position where the photograph group was photographed, based on coordinates of the index recognized by the recognition means on the basis of the respective photographs; ,
A first index coordinate which is an index coordinate in a coordinate system based on the first imaging position for the same index, and an index coordinate in a coordinate system based on a second imaging position different from the first imaging position. An association means for associating two index coordinates for each index;
Based on a plurality of combinations of the first index coordinates and the second index coordinates for the same index associated by the associating means, the coordinates in the coordinate system with the first imaging position as a reference, and the A computer is used as conversion value calculation means for calculating first and second conversion parameters for converting the coordinates in the coordinate system with the second imaging position as a reference into the coordinates in a single first and second reference coordinate system as a reference. Photogrammetry program characterized by functioning.

Images