JPH03131706A - Three-dimensional position measuring instrument - Google Patents

Abstract

PURPOSE:To speed up processing by providing a slit light projection device which projects slit light in a space wherein an object is present and measuring the three-dimensional position of the object according to images obtained from two image pickup devices. CONSTITUTION:Images of the object 3 are picked up by TV cameras 1 and 2 and the slit projection device 6 projects the horizontal slit light 7 to irradiate a part 8 on the object 3 with the light. In a slit-light OFF state, images are picked up by the TV cameras 1 and 2 and in a slit-light ON state, an image is picked up by the TV camera 1. A difference circuit 21 extracts a slit image and a maximum and minimum value detecting circuit 23 finds the maximum value ymax and minimum value ymin of the depth (y) of the slit image. A correspondence retrieval circuit 16 determines the range of a corresponding search on a horizontal scanning line from data on the depth ymax and ymin and a three-dimensional coordinate calculating circuit 17 calculates the coordinates.

Description

DETAILED DESCRIPTION OF THE INVENTION [Industrial Application Field] The present invention is a method for detecting objects in three dimensions using binocular stereoscopic viewing, which is applied to, for example, visual sensors for mobile robots and obstacle detection sensors for automatic guided vehicles. This relates to a three-dimensional position measuring device J that measures position.

[Conventional technology]

Binocular stereopsis is known as a device for measuring the three-dimensional position of an object in a non-contact manner. As shown in Fig. 4, two television cameras 1.2 are installed so that their optical axes are on the xy plane and parallel to the y axis, and the object 3 is imaged. This is a device that determines the three-dimensional position of an object 3 based on an image signal obtained from a television camera 1.2.

That is, when the object 3 is imaged with the configuration shown in FIG. 4, the point P on the object 3 becomes a point P on the image IL e IR obtained by each television camera 1. p
L(xL, yl), the image is formed on the point pR(χn e yR). Therefore, among the coordinates of point P, the y coordinate, that is, the depth LA=d, is given by the following equation.

Furthermore, if necessary, the x and z coordinates can also be calculated from simple geometric properties ζ. Here, 4 is the y-axis and the optical axis of the television camera l, and 5 is the optical axis of the television camera 2, L,
R is the center of the imaging lenses of the television camera 1.2, and f is the focal length of those lenses. In addition, CRICL is the center position of the image 'L * R obtained by the TV camera 1.2 (xL * yt,) 1 (xn + yn)
is the coordinate axis ζ on those images.

Image J) of point P on image IL obtained from television camera I
As a method for searching for the image pR on the image IR obtained from the television camera 2 corresponding to LJ, there are correlation methods, coarse-grained search,
Relaxation methods, dynamic programming methods, etc. are used, but in either method, as shown in FIG. Point a on the horizontal scanning line IL.

I try to search for one corresponding point.

Therefore, the point corresponding to Dotcho in Figure 6 is point a.

A point is selected as a point on an edge that has similar characteristics to the edge on which it exists. Here, an edge is a part on an image where the brightness changes suddenly, and corresponds to the outline or ridge of an object, and serves as a feature point for association.

[Problem to be solved by the invention]

The problem with the conventional three-dimensional position measuring device explained in FIGS. 4 to 6 is that the point corresponding to point am in FIG.
When searching on R, it is necessary to examine all feature points included from the left end to the right end of the horizontal scanning line R.

This is due to the fact that the distance from the object to the television camera is unknown.

As described above, in the conventional method, it is necessary to check all the feature points on the horizontal scanning line IL, which results in an extremely long processing time and is impractical.

The present invention has been made to solve the above-mentioned problems, and an object of the present invention is to limit the space in which the object exists using another means, so that the point a,
When searching for a point corresponding to the horizontal scanning line IR on the horizontal scanning line IR, the search range can be limited, and as a result, it is possible to provide a three-dimensional position measuring device that can speed up the processing.

[Means to solve the problem]

A three-dimensional position measuring device according to the present invention includes two imaging devices, a first and a second imaging device, that capture images of a target object, and calculates the three-dimensional position of the target object based on two images obtained from the two imaging devices. The apparatus for measuring a position is characterized in that it includes a slit light projecting device that projects a slit light toward a space 1 containing the object.

That is, in the present invention, the search range for associating mutually corresponding feature points between the two images is limited by projecting slit light into a space containing the object.

[Effect]

According to the present invention, when the slit light projected from the slit light projector S into the space containing the object is captured by one of the television cameras, the slit image on the object surface is captured by the principle of triangulation. The depth of each point is determined. Among these depths, by checking the closest depth and the farthest depth to the television camera, it is possible to limit the search range on the horizontal scanning line in the correspondence search using binocular stereoscopic vision.

〔Example〕

FIG. 1 is a configuration diagram of a three-dimensional position measuring device according to an embodiment of the present invention, in which an object 3 is imaged by a television camera 1.2 which is an imaging means, and a slit light projector 6.
A horizontal slit light 7 is projected from the horizontal slit light 7, and an irradiation portion 8 of the slit light is generated on the object 3.

Here, the television cameras 1.2 are arranged so that their respective optical axes are on the same horizontal plane and parallel to each other, as explained in FIGS. 4 and 5 as a conventional technique. But, IP! Even if the elements are not arranged in this way, the same result can be obtained by correcting the captured image, so there is no particular restriction on the arrangement. Moreover, the same can be said about the horizontal slit light 7.

The slit light projector #6 can be turned on and off through the slit light 0N10FF circuit I9 according to commands from the host computer 18.

The image signal obtained from the television camera 1.2 is converted into a digital signal (ζ) by the ~Φ conversion circuit 10i11 and stored in the image memory IO,11.

The feature extraction circuits 12 and 13 detect changes in brightness from the image signal, and extract edges that correspond to the outline and edge 1ζ of the object 3 as feature points.

The correspondence search circuit z6 is a circuit that associates the feature points on each image obtained by the feature extraction circuits 14 and 15 for each horizontal scanning line, and the correspondence search circuit I7 searches the matched feature points using equation (1). This is a three-dimensional coordinate calculation circuit that calculates depth.

These processes are performed when the slit light projection device 6 is off. On the other hand, when the slit light is on, the converter I
The digitized image signal which is the output of Q is stored in the image memory 20, and the data is stored in the image memory 12 and the image obtained when the slit light is off, and the difference circuit 2! The difference between the two is calculated.

The center detection circuit 22 calculates the center position of the slit image for each vertical scanning line in the image in which only the slit image remains, which is the output of the difference circuit 2I.

Furthermore, 23 calculates the depth of each point of the slit image from the center position of each vertical scanning line based on the principle of triangulation, then calculates the maximum and minimum values of the depth, and sends the horizontal This is a maximum value/minimum value detection circuit that outputs a search range on a scanning line.

Next I will describe the operation of the three-dimensional measuring device configured as shown below. First, with the slit light turned off, a television camera is set.

2 to capture an image. Next, the slit light is turned on and an image is captured by the television camera l, which is sent to the differential circuit 2I via the image memory 20. Therefore, the difference between the image and the image stored in the image memory I2 when the slit light is off is taken. Then, only the slit image remains as shown in FIG. 2jζ.

Here, the slit light is on the same plane as the plane of z=-h.

The center detection circuit 22 calculates the center position q (xL, yJ) of the slit image on each vertical scanning line JV that intersects with the slit image of this image. Here, the point Q is the object corresponding to the point q. This is the point above.

The depth (y coordinate) y of point Q is determined by the following equation.

y= −f−1 ・・・・・・・・・(
2) Find the depth y at each point of the L slit image, and find the maximum value Y
max and minimum value'! Maximum value/minimum value detection circuit 2
Find it in 3.

These data are sent to the correspondence search circuit I6.

On the other hand, image memory 12.13 obtained when the slit light was off.
In the four stored images, feature extraction circuits 14 and 15 detect portions where brightness changes as feature points, and the detected portions are sent to a correspondence search circuit 16.

The correspondence search circuit 16 determines the range of correspondence search on the horizontal scanning line from the depth Ymax and ymlo data.

In this method, the feature point p on the image IL is shown in FIG.
When searching for the point corresponding to L on the artist R, the space in the depth direction to be considered is ymin≦y≦3'max, so there are points on the object between point P and point P. Therefore, on the image IR, pRl(XH≦
Somewhere in p4, there exists a corresponding point J)R.

In C(7) formula, jp * f/ymin, no b-f/y
max is unchanged at the position of point pL and is common to all correspondence searches on horizontal scanning lines ◇ In this way, the search range for point pL is pH1≦XH≦
Horizontal scanning line because it is limited to pR2! There is no need to search all ranges on R.

〔Effect of the invention〕

As described above, according to the present invention, three images of the object are obtained based on the first and second imaging means for imaging the object and the respective image signals obtained by these imaging means.
Since the device for measuring dimensional position is equipped with a device that projects slit light into the space containing the object, conventional binocular stereopsis alone cannot match the feature points of each of the two images. Although the processing time was extremely long due to the wide search range on the horizontal scanning line, measurement using slit light allows the minimum and maximum values in the depth direction of the space containing the object to be determined. A three-dimensional position measuring device can be provided in which the search M range can be limited to a small size and the processing time can be greatly shortened @IC.

[Brief explanation of the drawing]

FIG. 1 is an overall configuration diagram according to an embodiment of the present invention, and FIG.
The figure is an explanatory diagram of measurement using slit light in the same embodiment, Figure M3 is an explanatory diagram of the effect of limiting the search range in the same embodiment, Fig. 4 is a configuration diagram of a conventional three-dimensional measuring device, and Figs. FIG. 6 is an explanatory diagram of the operation of conventional three-dimensional measurement training, respectively. 1, 2...TV camera, 3...Object, 6...Slit light projector.

Claims (1)

[Claims] It has two imaging devices, a first and a second imaging device, for imaging a target object,
A device for measuring a three-dimensional position of the object based on two images obtained from the two imaging devices, comprising a slit light projection device that projects a slit light toward a space including the object. A three-dimensional position measuring device characterized by: