JP5478366B2 - Video surveillance device - Google Patents

Description

  The present invention relates to a video monitoring apparatus that processes and synthesizes a plurality of images obtained by photographing a monitoring target area and displays the images as one synthesized image.

  Video (images) on the near side and depth side of the station platform are shot from different positions with overlapping areas, and the depth side image is aligned with the near side image, and the same subject is displayed in the same size. Then, the image is converted as described above, and a composite image is obtained by superimposing the overlap area with a predetermined transmittance (see, for example, Patent Document 1).

JP 2008-199505 A

  A conventional video monitoring apparatus can combine multiple images of the monitoring target area into one image, but the depth side image is in a state of being fitted into the near side image, so that the depth on the screen to be viewed There is a problem that the size of the side image is reduced, and visual observation is difficult.

  The present invention has been made to solve the above-described problem, and converts the depth-side image on the screen to be viewed into a size that is easy to view, and does not cause a missing portion of the monitoring target region. Thus, the object is to process the near-side image so as to eliminate the sense of perspective and obtain a composite image.

  The video monitoring apparatus according to the present invention includes a first monitoring camera that captures a first monitoring area that is a monitoring target area, a second monitoring camera that is the monitoring target area, and has an overlapping area with the first monitoring area. The second monitoring camera that captures the monitoring area, the first image of the first monitoring area captured by the first monitoring camera is captured, and the first image overlaps the second monitoring area. A first part that obtains a first converted image by dividing an overlapping part including a region and a non-overlapping part other than the overlapping part, enlarging the overlapping part, reducing the non-overlapping part, and performing image conversion The image conversion unit captures the second image in the second monitoring area, changes the display magnification of the second image so as to match the subject size of the first converted image, and performs image conversion. A second image conversion unit for obtaining a second converted image, An image composition processing unit that captures the first and second converted images and obtains a composite image in which the second converted image is superimposed on the overlapping portion of the first converted image, and a display that displays the composite image It has a part.

  According to the video monitoring apparatus of the present invention, in order to perform image conversion for enlargement and reduction in each of the overlapping portion and the non-overlapping portion of the first and second images, In addition, image conversion that reduces the non-overlapping part can be performed, and more information about the first image is displayed in the composite image on the display unit than when the non-overlapping part is not reduced. Is possible.

It is explanatory drawing which shows the imaging | photography condition of the monitoring area by the monitoring camera by Embodiment 1 of this invention. It is a block diagram which shows the relationship of arrangement | positioning with the video monitoring apparatus of this invention, and a station facility. It is a block diagram of the image | video monitoring apparatus of this invention. It is a figure regarding the image | video synthesis | combination required for description of this invention. It is a figure regarding the image | video synthesis | combination required for description of this invention. It is explanatory drawing which divides | segments the image | video obtained by the video monitoring apparatus of this invention. It is explanatory drawing of the image conversion process of the divided area by this invention. It is explanatory drawing of the image conversion process of another division area by this invention. It is explanatory drawing of the image conversion process of another division area by this invention. It is a figure which shows the near side conversion image by this invention. It is explanatory drawing which synthesize | combines the depth side image obtained by the video monitoring apparatus of this invention with a near side conversion image. It is a figure which shows the synthesized image obtained by the video monitoring apparatus of this invention.

Embodiment 1 FIG.
Next, Embodiment 1 of the present invention will be described with reference to FIGS.
FIG. 1 is a diagram showing a state in which the surveillance camera constituting the video surveillance apparatus of the present invention is installed in the vicinity of the platform of the station facility, and is an explanatory diagram showing the photographing situation of the surveillance area by the surveillance camera.
As shown in FIG. 1, a train vehicle 2 a and a train vehicle 2 b connected to the train vehicle 2 a are entered in the home 1. As seen from the left side of FIG. 1 (the position of a conductor (monitor observer) 6 and display unit 5 described later), the train vehicle 2a is stopped on the front side of the platform 1 and the train vehicle 2b is stopped on the depth side. The area of the platform 1 where passengers get on and off from these train vehicles 2a and 2b and the train entrance / exit is the monitoring target area. The surveillance camera 3a arranged on the front side of the home 1 mainly captures the side of the train vehicle 2a where the entrance / exit is provided and the depth side farther away from the side, and the home 1 facing the depth side of the home 1 The monitoring camera 3b arranged in the center mainly images the home 1 facing the side surface where the entrance / exit of the train vehicle 2b is provided. Thus, when image | photographing with the monitoring cameras 3a and 3b, the area | region (1st monitoring area) image | photographed with the monitoring camera 3a and the area | region (overlapping area | region) overlapping with the area | region (2nd monitoring area) image | photographed with the monitoring camera 3b ) Occurs. When the first and second images obtained by photographing the first and second monitoring areas are synthesized at a predetermined position on the overlapping area, image conversion processing is performed so that the size of the subject is made uniform. is there.

  FIG. 2 shows a schematic configuration diagram of the video monitoring apparatus of the present invention in relation to the arrangement of station facilities, and FIG. 3 shows a configuration block diagram of the video monitoring apparatus. The video monitoring apparatus 100 monitors the trains 2a and 2b of the train 1 that has entered the platform 1 and the platform 1 and the area of the platform 1 for getting on and off the vehicles, and imaging for obtaining an image of the monitored area. The unit 3 includes surveillance cameras 3a and 3b. More specifically, the front side monitoring camera (first monitoring area) that captures the train vehicle 2a, which is the monitoring target area, and the home 1 and its surroundings for getting on and off the same vehicle as well as the depth area (first monitoring area) beyond it. Surveillance camera) 3a, train vehicle 2b, and the platform 1 for getting on and off the vehicle, its surroundings, and a region including a depth beyond it, a region having an overlapping area with the first monitoring area (second monitoring area) The imaging unit 3 is configured by the depth-side monitoring camera 3b (second monitoring camera).

Next, the image processing unit 4 that captures and converts image data obtained by the monitoring cameras (first and second monitoring cameras) 3a and 3b will be described. The image processing unit 4 captures the first image (front side image) of the first monitoring area captured by the monitoring camera 3a, and the overlapping unit includes an area that overlaps the second monitoring area. The first image conversion to obtain a first converted image (front side converted image) by dividing the non-overlapping part other than the overlapping part, enlarging the overlapping part, and reducing the non-overlapping part to perform image conversion Display the second image so that the second image (depth side image) in the second monitoring area (front side image conversion unit) 4a is captured and aligned with the subject size of the first converted image. A second image conversion unit (depth side image conversion unit) 4b that obtains a second converted image (depth-side converted image) by changing the magnification and captures the first and second converted images, An image that obtains a composite image in which the second converted image is superimposed on the overlap of one converted image It constituted by the deposition processing unit 4c.
Further, a composite image display unit (display unit) 5 for displaying a composite image is disposed at a position where the composite image display unit (display unit) 5 can be seen from the conductor 6 standing on the other end side (that is, the front side) on the platform 1 and confirming the safety of passengers. The In the case where the near side image is captured and converted by the near side image conversion unit 4a and the near side image includes an area that does not need to be displayed as a composite image, only the area that needs to be displayed is included. It is possible to leave the display selection area and delete the unnecessary area. Similarly, in the case of the depth-side image conversion unit 4b, the display area of the depth-side image can be selected.

  Here, considerations by the inventors until reaching the present invention will be described. FIG. 4 is a diagram showing a conventional technique related to image composition necessary for explaining the present invention. 4A shows a depth side image 32 taken by the surveillance camera 3b, FIG. 4B shows a near side image 31 taken by the surveillance camera 3a, and FIG. 4C shows the near side image. 31 is displayed at the same magnification, and the depth-side converted image 32a obtained by reducing the depth-side image 32 is combined and displayed with the subject size and position (predetermined positions in the overlapping area) overlapped. An image 33 is shown. As can be seen from the composite image 33, the near-side image 31 is displayed at the actual magnification. Therefore, the person 20a standing on the platform 1 in front of the train vehicle 2a on the near side and raising both hands up is compared in the composite image. Since it is displayed in a large size, visual confirmation by the conductor 6 is easy, but for the person 20b standing on the platform 1 in front of the train vehicle 2b on the depth side and dropping his hand, the depth-side converted image 32a is displayed in a reduced size. Therefore, since it is displayed in a small size according to the reduction magnification, it can be seen that visual confirmation by the conductor 6 is difficult.

Therefore, the inventors expand the area in which the depth side image 32 of the near side image 31 is inserted, so that the depth side image 32 is larger than the depth side image 32a shown by the composite image 33 in FIG. A study was made to make it possible to display on the display unit 5 (on the composite image) so as to have a ratio and to facilitate visual observation in the depth side region of the monitoring target region.
FIG. 5 is a diagram relating to image composition necessary for the description of the present invention, and is an image conversion explanatory diagram for enlarging the depth side region of the home 1 in the near side image 31.
In order to enlarge and display the depth side area of the home 1, as shown in FIG. 5A, the lower end portion (point B) that intersects the home 1 at the depth side end portion of the train vehicle 2 a in the near side image 31 is illustrated. 5 (b), the more by moving the point B ₁ of the lower, it is possible to extend the vertical depth end of the train car 2a, machining depth side area of the home 1 (conversion) It becomes possible to display wider than before.

However, the rectangular area surrounded by the points A, B, C, and D in FIG. 5A is changed to a rectangular area surrounded by the points A, B ₁ , C, and D in FIG. When planar projection conversion is performed on the entire front side image 31 so as to move, the front side of the home 1 is moved in accordance with the movement of the point B to the point B ₁ as shown in the front side conversion image 40 of FIG. The side is greatly distorted with rotation. Therefore, the missing area 31c of the home 1 on the near side from the person 20a displayed in FIG. 5A is not displayed on the converted near side converted image 40. Note that the three circles on the home 1 are marks for facilitating the understanding of the transformation of the image after conversion, and on the line perpendicular to the length direction of the home 1 (in the horizontal direction in the figure) Each is a mark arranged at a distance. The circles are displayed in a horizontal row on the home 1 on the near side image 31 before conversion, but on the near side converted image 40 after conversion, the home 1 is deformed so as to be rotated. As a result, although the arrangement of the three circles is inclined to the lower left side, and as a result, the area where the depth image is arranged can be enlarged, as described above, the missing area 31c in front of the person 20a is It has been found that there is a problem that it is not displayed on the near side converted image 34.

Therefore, the inventors adjust the display area of the depth side image 32 so as to be larger than the display area corresponding to the depth side image 32 on the near side image 31, and also perform the omission as shown in FIG. An image conversion technique that does not generate the area 31c was examined, and it was found that these can be achieved by using the video monitoring apparatus 100 shown in FIGS.
The processing method in the image processing unit 4 that is the configuration of the video monitoring apparatus 100 in FIG. 3 will be described in more detail.

  In the depth side processing unit 4b that captures the depth side image 32 captured by the monitoring camera 3b, an area to be displayed in the composite image is selected from the depth side images 32 shown in FIG. 4A (display area selection). The shape of the depth side image is determined. By this process, for example, the area without the home, which is further on the back side than the terminal end of the home 1, can be cut, and the display area can be limited to only the area where passengers can get on and off. By performing such display area selection processing, the display area on the composite image can be used effectively. Needless to say, the entire region of the depth-side image 32 can be selected. Then, the size of the selected depth side image 32 is changed, and the arrangement on the composite image is determined. Information on the size of the depth-side image 32 processed in this way and the arrangement information (video processing information) on the composite image is output to the near-side image conversion unit 4b.

Next, the near-side image conversion unit 4a that captures the near-side image 31 captured by the monitoring camera 3a selects a necessary area as a monitoring area for the near-side image 31 (display area selection). It is also possible to select the entire area of the near-side image 31.
Next, image division (area division) processing of the near side image 31 is performed. In this area division processing, the near side image 31 is obtained based on the arrangement data on the composite image of the processed depth side image (depth side converted image 32b described later) obtained from the depth side image converting unit 4b. First, it is divided into an overlapping part and a non-overlapping part. The overlapping portion includes an overlapping area of the near side image 31 and the depth side image 32 and is an area where the processed depth side image is superimposed. A region other than the overlapping portion is a non-overlapping portion. Note that the non-overlapping portion is divided into, for example, a plurality of rectangular areas in order to facilitate image enlargement or reduction processing.
In addition, since the display area of the composite image is constant and the overlapped portion is displayed larger than the original image by conversion in the composite image, the non-overlapping portion on the composite image is reduced. .

The area division of the near-side image 31 will be described with reference to the image division explanatory diagram of FIG. In the example of FIG. 6, the non-overlapping part is divided into two rectangular areas. It is assumed that the entire area of the near-side image 31 is selected when the display area is selected.
In FIG. 6 (a), a divided area 1 (41) which is an overlapping region including the depth portion of the home 1 and the train vehicle 2b and is an overlapping portion with the depth side image 32 is indicated by a hatched portion. The divided area 1 (41) is a quadrangular area surrounded by four points (A, B, F, E) in the figure.
In FIG. 6 (b), one area of the non-overlapping portion divided into two, including the person 20a on the near side of the home 1, and the divided area 2 (42) in front of the entrance / exit of the train vehicle 2a is shaded Is shown in the section. The divided area 2 (42) is a rectangular area surrounded by four points (B, C, G, F) in the figure.
Moreover, in FIG.6 (c), it is the other area of the non-overlapping part divided | segmented into two, and the divided area 3 (43) including the side surface by the side of the entrance / exit of the train vehicle 2a is shown by the shaded part. The divided area 3 (43) is a quadrangular area surrounded by four points (A, B, C, D) in the figure.

The present invention is characterized in that the front side image conversion unit 4a performs image processing for each divided area. FIG. 7 illustrates an image conversion explanatory diagram of the home depth side area. As shown in FIG. 7A, when the divided area 1 (41) is surrounded by four points A, B, F, and E, a depth-side image conversion unit as shown in FIG. 7B. The point B is moved to B ₁ and the point F is moved to F ₁ so as to be the size of the enlarged divided area 1 determined in 4a, and image conversion is performed so as to enlarge the display area in the vertical direction. The rear divided area 1 (41a) is obtained. The post-processing divided area 1 (41a) is an area obtained by adding a rectangle surrounded by four points B, B ₁ , F ₁ , and F to a rectangle surrounded by four points A, B, F, and E. .

Next, FIG. 8 shows an image conversion explanatory diagram of the area in front of the home. In response to the fact that the post-processing divided area 1 (41a) has been converted to expand in the vertical direction, in the divided area 2 (42), as shown in FIG. 8A, four points B, C, As shown in FIG. 8B, the region surrounded by G and F is such that point B is B ₁ (matches point B ₁ in FIG. 7) and point F is F ₁ (point F ₁ in FIG. 7). The image is converted so as to reduce the display area in the vertical direction, and a post-processed divided area 2 (42a) is obtained. The post-processing divided area 2 (42a) is surrounded by four points B, B ₁ , F ₁ and F from a trapezoidal region surrounded by four points B, C, G and F in FIG. This is an area obtained by subtracting a square area and a triangular area surrounded by three points B, B ₁ and C.

Moreover, the image conversion explanatory drawing of the train vehicle side area is shown in FIG. As with the above, the divided area 3 (43) but, with the fact that the point B shown in FIG. 8 (a) moves to _{B 1,} as shown in FIG. 9 (a), _{B, B} 1, C A trapezoid surrounded by points A, B ₁ , C, and D by adding a triangular region surrounded by three points to the square region surrounded by the original four points A, B, C, and D The post-processing divided area 3 (43a) is obtained.
As described above, when the non-overlapping portion is divided into two quadrangular (trapezoidal) areas and image conversion is performed, the divided area after processing may be enlarged in one area. However, when all the non-overlapping parts are added up, processing is performed so that the non-overlapping parts are reduced by the amount of increase in the display area in the overlapping parts.
Note that the present invention does not limit the image conversion method, but the above-described conversion can be realized by using, for example, planar projection conversion.

Next, the post-processing divided areas 1, 2, and 3 (41a, 42a, and 43a) shown in FIGS. 7, 8, and 9 are shown as one near side converted image 31a in FIG.
As shown in FIG. 5 described above, in the case where all the image areas are similarly converted in accordance with the conversion processing of one divided area in the image, the missing area 31c occurs after the conversion. As shown, in the present invention, by performing conversion for each area after division, all four points surrounding each area are all necessary as the original monitoring area so that they are not removed from the display image after conversion. Processing is performed so that the area can be displayed. That is, the four points E, G, C, and D surrounding the near-side image 31 do not move even if the near-side converted image 31a is processed, and the boundary line that delimits the inside of the image is changed. Thus, since each area is displayed after being enlarged or compressed in a predetermined direction, it is possible to eliminate omission (leakage) of image information after conversion.

In the image composition processing unit 4c of the image processing unit 4, data of a depth side converted image (indicated by reference numeral 32b in FIG. 12 described later) obtained by processing in the depth side image converting unit 4b and the near side image converting unit 4a. If, captures the data of the front side transformation image 31a, 11 a (a) point a (train car 2a depth side upper portion) and the point _{B 1} (train car 2a depth side lower portion), FIG. 11 (b) point a on the front side converted image 31a, so as to match the point B _1, by aligning the size of the subject, by fitting the depth side converted image 32b on the front side transformation image 31a, the synthetic image as shown in FIG. 12 50 data are obtained. In the region where the two images overlap on the composite image 50, the image is processed so that the near side image is superimposed on the depth side image with a transmittance of 50%.

Data of the composite image 50 is read by the composite image display unit (display unit) 5 and output and displayed on the monitor screen, so that the monitor can be easily viewed, and the conductor 6 can accurately confirm the safety of the depth area of the home 1. Can be executed. In addition, the non-selection area 32c (shaded part) located in the upper part of the depth side image 32 of Fig.11 (a) is an area | region where the passenger's movement is not assumed, and the display area selection in the depth image conversion part 4b It shows a portion that is sometimes recognized as a non-display area.

In the present invention, the information displayed on the display unit 5 is converted by converting the divided area 1 (41a) where the depth-side converted image 32b is displayed so that a larger display area than before the image conversion can be obtained. It is easy to confirm. In other words, the effect of facilitating visual observation can be obtained by the generation of a region surrounded by four points B, B ₁ , F ₁ , and F shown in FIG. 7B or a region corresponding thereto. Here, as an indication of the size of the enlarged display, for example, when the monitor of the display unit 5 has a rectangular display space that extends vertically and horizontally, the size of the post-processed divided area 1 (41a) is set on the display screen. By setting the area to occupy the vertical ½ and horizontal ½ areas, it is possible to eliminate the difficulty of seeing distant objects even in images with a strong perspective. Needless to say, the ease of visual recognition is improved if at least the portion of the distant area displayed in the entire area image obtained by photographing the entire area of the monitoring target area is displayed larger than the display size on the entire area image. As described above, it is possible to prevent the loss of information necessary for video monitoring.

  In addition, according to the present invention, the blind spot of the monitoring target area is obtained by using a plurality of monitoring cameras 3a and 3b that capture a place where it is difficult to check getting on and off the home 1 (for example, a shadow of a pillar or a depth area of the curved home 1). By creating and displaying a composite image in real time like a single camera image, the number of monitors constituting the display unit 5 can be reduced, the number of confirmation points by the conductor 6 can be reduced, and boarding / exiting confirmation work It is possible to improve the visibility and safety.

1 Home 2 Train 2a Train vehicle (front side) 2b Train vehicle (depth side)
3 Imaging unit 3a Surveillance camera (front side)
3b Surveillance camera (depth side) 4 Image processing unit 4a Depth side image conversion unit 4b Front side image conversion unit 4c Image composition processing unit 5 Composite image display unit 6 Conductor 20a Person (front side)
20b Person (depth side) 31 Front side image 31a, 40 Front side converted image 31c Missing area 32 Depth side image 32a, 32b Depth side converted image 32c Non-selected area 33, 50 Composite image 41 Division area 1 41a Post-processing division area 1
42 Division area 2 42a Division area 2 after processing
43 Divided area 3 43a Processed divided area 100 Video monitoring device.

Claims (6)

  A first monitoring camera that images the first monitoring area that is the monitoring target area, and a second monitoring that images the second monitoring area that is the monitoring target area and has an overlapping area with the first monitoring area. The camera, the first image of the first monitoring area taken by the first monitoring camera is captured, the overlapping portion including an area overlapping the second monitoring area, and the overlapping A first image conversion unit that obtains a first converted image by dividing the non-overlapping part other than the part, enlarges the overlapping part, reduces the non-overlapping part, and performs image conversion, and the second monitoring The second image of the area is captured and the second converted image is obtained by changing the display magnification of the second image so as to match the size of the subject of the first converted image. Second image conversion unit, capturing the first and second converted images A video surveillance comprising: an image composition processing unit that obtains a composite image obtained by superimposing the second conversion image on the overlapping portion of the first conversion image; and a display unit that displays the composite image apparatus.   The first monitoring area corresponds to a portion extending from the near side to the depth side in the monitoring target region when viewed from the position where the display unit is disposed, and the second monitoring area is the depth side The video surveillance apparatus according to claim 1, wherein the video surveillance apparatus corresponds to a portion.   The first image conversion unit divides the first image so that each of the overlapping portion and the non-overlapping portion has a quadrangular shape surrounded by four points. The video monitoring apparatus according to claim 1, wherein the image conversion process is performed so that four points surrounding the overlapping portion are arranged within a range displayed on the display unit.   2. The video surveillance according to claim 1, wherein the first image conversion unit divides the non-overlapping portion of the first image into a plurality of areas, and performs image conversion processing for each of the divided areas. apparatus.   The first image conversion unit divides the non-overlapping portion so that the area has a quadrangular shape surrounded by four points, and the four points surrounding the area are displayed on the display unit. 5. The video monitoring apparatus according to claim 4, wherein the image conversion process is performed so that the image conversion process is performed within a predetermined range.   The video monitoring apparatus according to claim 1, wherein the first and second monitoring areas target a platform of a station facility and a train vehicle entering the platform.

Images