JPS61265510A - 3D measurement method for objects - Google Patents

Abstract

PURPOSE:To make it possible to measure the spatial arrangement of matter at a high speed by one camera, by irradiating a plurality of slit lights preset in arrangement according to predetermined order. CONSTITUTION:The image of the matter irradiated with an illuminator 12 is picked by a camera 1 and the output thereof is converted to a digital signal which is, in turn, inputted to an image processor 15. This input signal is processed by a density processing part 16 to be stored in an image memory 21 as an image A wherein the contour of the matter was extracted. Next, a scanning region is calculated from the obtained data and the selection of a slit light source and the determination of the scanning region are subsequently performed on the basis of the calculation result. Thereafter, the illuminator 12 is turned OFF and a selected slit light source is lighted to irradiate slit light and the slit light image on the matter is picked up to extract the center line thereof and stored in an image memory 22 as an image B. Succeedingly, the data of the images A, B are read from the image memories 21, 22 and the slit image corresponding to the slit light image is extracted by an inter- image AND operation part 20 and the coordinates thereof are stored in a buffer 26. Subsequently, a distance is calculated by a distance calculation part 25.

Description

[Detailed Description of the Invention] [Table of Contents] Overview
・・・・・・・・・・・・・・・・・・・・・(1) Industrial application field・・・・・・・・・・・・・・・・・・・・・
・・・(2) Conventional technology・・・・・・・・・・・・
・・・・・・・・・・・・・・・・・・・・・(2) Problem to be solved by the invention ・・・(3) Means for solving the problem...・・・・・・・・・(4) Basic principle of the invention・・・・・・・・・・・・・・・・・・・・・
...(4) Example (i) Configuration of 414 devices in total ......
...(8)(ii) Total of 1111 processes...
・・・・・・・・・・・・・・・・・・・・・(10)
Effect of the invention····················
... This is a method in which a plurality of slit lights whose arrangement is set is irradiated in a predetermined order, and the spatial arrangement of an object is measured at high speed with a single camera.

[Industrial application field]

The present invention relates to a method for measuring the spatial arrangement of an object, which is necessary when an automatic machine such as a robot performs work. This relates to a third-order observation a+1 method of an object in which a total of ΔIII is obtained.

[Conventional technology]

Conventionally, there is a so-called binocular stereoscopic viewing method as a method of viewing an object three-dimensionally by using visual information. In this method, first, as shown in Figure 9, two cameras 1 and 2 are fixedly placed near the target object 4, and the object is simultaneously photographed and the two-dimensional position of the object is determined based on the principle of triangulation. This is a method of measuring.

[Problem that the invention seeks to solve]

By the way, in this conventional method, the correspondence between the image taken by camera 1 and the image taken by camera 2 (9th
In the figure, it is difficult to find a point B corresponding to point A on the right screen for a point of interest A on the left screen, and the process is complicated and takes a long time. Further, in the case of a wire or the like, if it extends parallel to the arrangement of the two cameras, there is a problem in that corresponding points cannot be uniquely determined and measurement cannot be performed.

Therefore, the present invention makes it possible to perform three-dimensional measurements of objects such as wire rods without the complicated process of associating slit images taken with two cameras in the so-called binocular stereoscopic viewing method. This was done for the purpose of providing a method.

[Means for solving problems]

The present invention irradiates an object surface with a plurality of slit lights each having a slit light plane whose arrangement is set in advance in a predetermined coordinate system, and creates a slit light image of the object surface by the slit light. One step is to obtain a corresponding slit image on a predetermined image plane, and determine the spatial arrangement of the object in the coordinate system based on the position of the image plane of each slit image and the slit light plane corresponding to each slit image. This is a three-dimensional measurement method for objects to be measured.

[Basic principle of the invention]

Next, the basic principle of the present invention will be explained based on FIGS. 1 and 2. As a predetermined coordinate system, a camera orthogonal coordinate system QC-XCyCzC is set, with the focal point QC of the camera l as the origin and the optical axis direction of the camera l as two axes. Furthermore, the slit orthogonal coordinate system with the slit light source 6 as the origin is OS
Set yS zS. The relationship between the camera orthogonal coordinate system and the slit orthogonal coordinate system is generally expressed by the following relational expression.

However, hii (i, j=1.2.3) is i.

j represents the direction cosine of the angle between the coordinate axes, hi4(i
=1.2.3) represents the parallel movement distance. Now, the slit light surface is obtained by rotating the slit light source 6 by θJ around the mo surface xs'=o using the yS axis of the slit orthogonal coordinate system as the rotation axis.
Then, S'' is S'' in the slit orthogonal coordinate system: xscasO-zSsinO= 0...
It is expressed as (2). If this is expressed in the camera orthogonal coordinate system, using equation (1), S;: (b++ xC+h17 YC+h13 zC+
b+4) cos 01(h31 zc+h32 yC
+h33zc+h3s) sjnθ.

=0...(3) Hereinafter, for convenience of explanation, it is assumed that the object 4 is linear. By irradiating the slit light 8, a point-like slit light image 9 is created on the surface of the object 4, and by photographing it with the camera l, a slit image corresponding to the slit light image 9 is displayed on the image plane C4m of the photographing element of the camera l. Suppose that we obtain Let the coordinates of the slit light image 9 of the object 4-H be P'' (xp+C+
yPiC, zp; t), and the coordinates of the slit image corresponding to the three-two optical image 9 are IJ (XIJC+ 7
11C+ZIJC), and here, for the sake of simplicity, the image plane C is a plane parallel to the plane of
σ了, ), then L is expressed as. Then, the point-like slit light image 9 P
The coordinates of , are the line L which is the line of sight 10 and the slit light surface SJ
It is found as the intersection with

That is, from equation (4), we obtain, substitute this into equation (3), and summarize for zC: =h+a cos O1+b3a sin O+
- From (6), here, . In this way, the three-dimensional coordinates of P'', which is the slit light image 9, are obtained. Similarly, by sequentially rotating the slit light source 6 with the yS axis as the rotation axis, the object 41
By obtaining the three-dimensional coordinates of the slit light image 9 of −, three-dimensional measurement of the object 4 can be performed.

〔Example〕

Embodiments according to the present invention will be described below based on the drawings.

(i) Configuration of measuring device The device for realizing the three-dimensional measuring method of an object according to the present invention is as follows:
An example is constructed as shown in FIG.

In the figure, 1 is a camera, 11 is an analog-to-digital converter, and 12 is a normal light that illuminates the surface of the object 4.

A two-dimensional scanner 13 is used to move the slit light 8, and is composed of a movable mirror 13a, an x-axis drive section 13b, and a y-axis drive section 13c. 14 is a slit light device, which works in conjunction with the two-dimensional scanner 13, and includes a fixed mirror 14a, a beam splitter 14b, and a slit light source 14c.
and a slit light source 14d. The slit light source 14c and the slit light source 14d are appropriately selected and used according to the arrangement of the object 4. 15 is an image processor; a grayscale image processing unit 16 for extracting the outline of the object 4;
, a ridge point processing unit 17 for extracting a line with a thickness of 1 pixel width from the center line of the linear object 4, and an extremity processing unit 18 for extracting local peak points, and converting the digital signal into a binary signal. It consists of a binarization processing section 19 for converting images into images, and an inter-image AND operation section 20 having an inter-image operation unit 3. The image processor 15 is the same as the device shown in Japanese Patent Application No. 58-091309, with a polar point processing section 18 and an inter-image AND@ calculating section 20 added thereto. 21 and 22 are image memories, which are devices that store images obtained on the image plane C of the camera 1. 23 is an object direction/scanning area calculation unit,
This is to calculate the slit image position coordinates of the image plane c-h. Reference numeral 25 denotes a distance calculation unit which calculates the position coordinates of the slit light image 9, thereby performing three-dimensional measurement of the object. Reference numeral 26 denotes a buffer. 29 is CP
U, which is connected to the above-mentioned complete system ε, and performs control and calculation of all the devices.

(ii) Measurement processing Assume a linear object as the object to be measured.

First, as shown in FIG. 1, a camera l and a slit light source 6 are placed near a target object 4. The slit light source 6 emits the slit light 8 set as described in the basic principle of the invention, and is selected from the slit light source 14c and the slit light source 14d shown in FIG. 3 depending on the position of the object 4. It is something. The process will be explained below according to the flowchart shown in FIG. The object 4 is illuminated by lighting the normal illumination 12 placed near the object 4 (41). However, object 4
There is no need to turn on the light if it is possible to clearly distinguish the background from the background. Next, the object 4 illuminated by the normal illumination 12 is photographed by the camera 1, and the outline of the object 4 is captured on the image plane C of a photographing element or the like. At that time, the photographed image is converted into a digital signal having grayscale levels by the analog e-digital converter 11 (80), and the signal is then manually inputted to the image processor 15.The signal input to the image processor 15 is a grayscale image. The processing unit 16 performs an upper smoothing process 81, a Laplacian process 82, or a combination process 87 thereof, as shown in the flowchart of FIG. Next, the signal is subjected to ridge point processing 83 in the ridge point processing section 17, and further subjected to binarization processing 84 in the binarization processing section 19. If there is a lot of whisker-like noise on the center line of the object 4, a sweat removal process 85 is performed after the binarization process 84 to remove the noise. Through the processing described above, the center line of the object 4 is extracted as a line with a thickness of 1 pixel width, and storage processing 86 is performed in the image memory 21 as the image 30 (42). From the obtained data, object direction/scanning area calculation unit 2
3 to calculate the placement direction of the object 4. The method is
First, using the camera orthogonal coordinate system (or the orthogonal coordinate system set on the image plane), start points r, s of the object 4 on the image plane CJ:
(xsC, ysC) and the end point Le (NPC, 7e
C). Assuming that the object 4 to be treated here does not have a very complicated shape, an angle θ indicating the arrangement direction of the object 4 is calculated from L and Le. That is, As a result, the direction vector 1 = (cos O, sin +)) indicating the arrangement of the object 4 is found (43
). Based on the north results, the slit light sources 14c and 14d arranged in different directions are selected and the slit scanning area is determined (44). The slit light source is selected by calculating the absolute value of the inner product intersection Sc and the inner product Sd when the normal vectors of the slit surfaces of the slit light source 14C and the slit light source 14d are respectively SC, S, and I. The slit light source with the smaller absolute value of the inner product is selected. This is to obtain more slit light images 9 for the slit light 8 to ensure measurement. However, it is not necessary to select only one of the slit light sources 14c and 14d, and measurement may be performed twice from different directions using both slit light sources 14C914d.

On the other hand, the scanning area covered by the slit light 8 is
Assuming that the object 4 is not a complicated shape, four points corresponding to a necessary and sufficient area surrounding the object 4, L = l (XSC, ys
C), La (Xsc, MI8c) +L1, (
Xe', ypC) and Lb (XPC, ysc) are within the rectangular area of the image plane C-1-. In reality, the area is set slightly larger than this, and the range of the irradiation angle of the slit light 8 is 0 and 0. and? decide. The slit light 8 has this irradiation angle range O5 and θ,
・You will scan between.

Next, the normal illumination 12 is turned off, the selected slit light source 6 is turned on, and an appropriate position 21 (4
In step 5), the slit light 8 is irradiated. A bright line 7 is drawn across the surface of the object 4 by the slit light 8, and the object 4
A slit light image 9 is formed on the surface. The bright line 7 including the slit light image 9 is photographed by the camera 1, and the image plane C of the camera 1,
1-, a line image 37 including a slit image 38 corresponding to the slit light image 9 by the slit light 8 is obtained (37). At this time, the photographed line image 37 is converted into a digital signal 5) by the analog-to-digital converter 11 shown in FIG. The signal input to the image processor 15 is sent to the grayscale image processing section 16, where it is subjected to layering processing 91. as shown in the flowchart of FIG. A love run process 92 or a combination process 98 of both is performed. Next, the signal is subjected to ridge point processing 93 in the ridge point processing section 17, and further, the signal is subjected to ridge point processing 93 in the ridge point processing section 17.
At step 9, binarization processing 95 is performed, and the center line of the line image 37 produced by the slit light 8 is extracted and stored as an image 31 in the image memory 2.
2 (46). At this time, instead of the ridge point processing 93 performed by the ridge point processing section 17, the polar point processing section 1
Pole processing 94 according to 8 may also be applied. in this case,
Rather than the center line of the line image 37 produced by the slit light 8, points having local peaks are extracted.

In particular, when a slint image 38, which will be described later, is extracted as one pixel, the integration process shown in FIG. 8 is not necessary, which is efficient.

Next, as shown in FIG. 7, image data is read out simultaneously from the image memory 21 and the image memory 22 (100).
, A between image 30 and image 31 in inter-image ANDS calculating section 20
An ND operation 'H (101) is performed to extract a slit image 38 corresponding to the slit light image 9 located on the hill of the surface of the object 4 from the line image 37. Slit image 38 of the image plane Chi
The coordinate values C1 (xC, yc) of are stored in the eight buffer 26 (107.42). After the above processing is completed, while confirming the irradiation of the three-lit light 8 within the scanning area (48), use the two-dimensional scanner 13 to identify the three-light beam whose arrangement of the slit light surface S is different from that described in L. The slit light source 6 is moved so as to emit the slit light 8 having S (45
), and the series of processes (46, 47) described above are repeated in a predetermined order, such as in the order of slits and the inclination angle of the optical surface S. The coordinates I [ClCl In this case, for each irradiation of the slit light 8, the irradiation direction, that is, y in the slint orthogonal coordinate system, and the rotation angle θ with the axis as the rotation axis, are CP
Store in U29. In that case, if the two-dimensional scanner 13 and the image processor 15 are synchronized, the coordinate value C4 (xC) of the slit image 38 on the image plane Cl- can be obtained in real time.

yC) can be extracted. By the way, the coordinate value CI (xC.

yC) is not necessarily one point, but may appear several times. Therefore, in this embodiment, the coordinate value C of the slit image 38 is
After extracting j (XC, yC), the distance calculation unit 25
It was decided that the distance calculation process 49 would be performed. Distance calculation process 4
9, as shown in the flowchart of FIG. 8, first C; (x
', y') is performed. Integrated processing 1
10, the coordinate values CI (xC, yC) of the slit image 38 are classified into a plurality of appropriate classes by the process shown in Japanese Patent Application No. 58-235898, and each class is determined by the average of each coordinate position within the same class. The I+ target values of several points in each class are represented by one slit image 38 in each class.
The coordinate value of woCj (XiC, Vjc).

This is a process of integrating j=1 to n. When the coordinate values Cj (XjC+VjC) of the integrated slit image 38 are expressed not in the camera orthogonal coordinate system but in another image plane orthogonal coordinate system set on the image plane, the coordinate values of the slit image 38 are expressed in the camera orthogonal coordinate system. A transformation process 111 expressed in an orthogonal coordinate system is added. When the image plane C is expressed in the camera orthogonal coordinate system as ZC=f (camera focal length 5t), the image plane orthogonal coordinate system Q! ziylzi and camera orthogonal coordinate system QC-
The relationship between Xcyczc and Xcyczc is expressed as follows using equation (1).

Using this relational expression, Cr (XiC + YjC) is converted into Di (xlJc, yNc) and t
6゜I, here the coordinates are expressed in the camera orthogonal coordinate system.
(XiC, Yj') is consistent. 0 that uniquely determines the slit light surface S1 of the slit light 8 obtained in this way.
, and the corresponding coordinate values of the slit image 38, Dj (Xi□C, yI, C) (j = 1-n), are substituted into equations (7) and (8) to obtain the object 4. Since the position of the slit light image 9 can be expressed in the camera orthogonal coordinate system, it becomes possible to know the spatial arrangement of the object 4, that is, to measure the object 4 in three dimensions (112). still,
In this embodiment, the object 4 is considered to be linear, but the object 4 is not limited to a linear case, but can also be applied to a general three-dimensional case. At this time, the slit light beam on the object is processed as a linear beam (in this embodiment, it is processed as a point beam).

[Effects of the Invention] As described above, according to the present invention, the arrangement 4 is obtained by irradiating the object surface L with a plurality of slit lights each having a preset slit light surface in a predetermined order. Three-dimensional measurement of the object can be performed based on the relationship between the image and each slit light surface. Therefore, in the present invention, it is possible to measure an object with a single camera,
Since it does not require the complicated process of associating each image with each other, which is required when measuring with two cameras, it is possible to measure objects quickly in real time.

[Brief explanation of drawings]

Fig. 1 is an explanatory diagram of the principle of the present invention, Fig. 2 is a diagram of the basic principle of the present invention, Fig. 3 is an overall configuration diagram of an embodiment according to the present invention, and Fig. 4 is a diagram illustrating the principle of the present invention.
5 is a flowchart of the process of extracting an object contour, FIG. 6 is a flowchart of the line image extraction process using slit light, FIG. 7 is a flowchart of the slit image extraction process, and FIG. FIG. 8 is a flowchart of distance calculation processing, and FIG. 9 is a conventional three-dimensional measurement method of an object. l...Camera 4...Object 6...
・Slit light source 8...Slit Yumei Mitsumoto's body, Buried VL light Figure 1 Figure 2 / Il-Invention 1z and 1 Tadasugi

[Processing Seeds Figure 4] Monoine outline raffle drawing Figure 5 Also, Figure 6 is a diagram of the line drawing East extraction process by lit conversion 1z.

Claims (1)

[Claims] A plurality of slit lights each having a slit light surface whose arrangement is set in advance in a predetermined coordinate system is irradiated onto the object surface in a predetermined order, and a slit formed by the slit light on the object surface is While obtaining a slit image corresponding to the optical image on a predetermined image plane, the spatial image of the object in the above coordinate system is calculated based on the position of each slit image on the image plane and the slit light plane corresponding to each slit image. A three-dimensional measurement method for objects that measures their placement.