JPS58114172A - Three-dimensional visual input device - Google Patents

Abstract

PURPOSE:To give information required for positioning a robot to an object, by adjusting the focus of a camera device, which is arranged in the X-axis direction, by position information of the camera device which is arranged in the Z- axis direction of the object. CONSTITUTION:A camera device IV is arranged in the Z-axis direction of a set table 3 so as to be focused on the upper face of the set tble 3. A camera device 1H is arranged in the X-axis direction. In a three-dimensional visual input device having this constitution, the output of the device 1V is applied to a horizontal pattern measuring circuit 6, and the output of the circuit 6 is applied to a discriminating circuit 8. The output of the device 1H is applied to a vertical pattern measuring circuit 7, and the output of the circuit 7 is applied to the circuit 8. The device 1H is focused to an object 2 correctly through a focus controlling circuit 9 on a basis of the result of the circuit 8. The output of the circuit 8 is applied to a robot controlling circuit 10 to give information required for positioning a robot 4 for the object 2.

Description

Detailed Description of the Invention (1) Technical Field of the Invention The present invention relates to a three-dimensional visual input device, and particularly relates to a three-dimensional visual input device using two camera devices suitable for use in robot handling work, etc. .

(2) Background of the technology When recognizing an object as a three-dimensional image using a camera device, in the normally used autofocus camera device, the focus is selected by a human visually checking the viewfinder. The background, etc. other than the subject cannot be brought into focus. However, when handling an object using a robot or the like, it is necessary to correctly recognize the position of the object and provide accurate position information to the robot.

In this case, methods have been proposed in which the object is viewed overhead using a camera device and the position information is recognized using a surveying method such as triangulation, but the plane on which the object to be grasped by the robot is placed is set in advance. However, if the object is placed without any restrictions on its depth or horizontal position, or if multiple objects are placed on the same plane, the above object recognition method cannot accurately locate the object. No information was obtained. For this reason, there has been a demand for a three-dimensional visual input device that can recognize object position information.

(3) Prior art and problems Figure 1 shows a conventional method of determining the position of an object using a camera device for position recognition. If so, the position and height of the object 2 can be determined by the principle of triangulation by looking down on the object 2 with the camera device 1 placed at a predetermined height H from the mounting table 3.

However, in the above-mentioned surveying method, within the fixed field of view of the camera device, that is, at a predetermined distance of 11
It is necessary to place an object in 1L.

However, the object 2 on the mounting table is not always placed at a fixed position; for example, when the object 2 is placed at the position indicated by the broken line, the object 2 may be out of the field of view of the camera device 1 or out of focus. This has the disadvantage that position measurement is impossible.

(4) Purpose of the Invention In view of the above-mentioned conventional drawbacks, the present invention provides a first camera device disposed in the axial direction of an object placed on a mounting table;
The position information from the first camera device is given to a second camera device arranged in a direction perpendicular to the optical axis of the camera device, the focus is automatically adjusted, and high-quality three-dimensional information is transmitted to the second camera device. The purpose is to obtain data from a device and control robots, etc.

[5) Structure of the Invention According to the present invention, in order to recognize the shape of an object, a first camera device is arranged in the vertical direction of the object, and a first camera device is arranged in the horizontal direction of the object. The second camera device has a variable focus mechanism whose focus can be controlled from the outside, and the object position is measured by pattern recognition of the object image by the first camera device. To provide a three-dimensional visual input device, characterized in that data is input to the second camera device, and the three-dimensional shape position is recognized from an image obtained by moving the camera device over the object on a focusing table. This is achieved by

(6) Embodiment of the Invention An embodiment of the present invention will be described below with reference to FIGS. 2 to 5.

FIG. 2 is a perspective view showing the arrangement of the three-dimensional visual input device of the present invention, in which an object 2 placed on a mounting table 3 is grasped by a robot 4.
To describe the case of handling, the first camera device IV is arranged so as to focus on the top surface of the mounting table 3 in the X-axis direction of the mounting table 3, that is, in the axial direction of the object 2.

The second camera device IH is arranged in the X-axis direction, that is, in a horizontal plane perpendicular to the Z-axis.

In the three-dimensional visual input device configured as described above, the XY coordinates 5 are set within the field of view by image processing the plane image of the object 2 on the mounting table 3 obtained from the first camera device 1.
In order to measure the object position at , the output of the first camera device 1v and the output of the horizontal pattern measuring circuit 6 are applied to the determination circuit 8 as shown in FIG. On the other hand, the output from the second camera device IH is applied to the vertical pattern measurement circuit 7, and the output of the vertical pattern measurement circuit 7 is applied to the determination circuit 8.

Based on the result of the determination circuit 8, the second camera device IH is correctly focused on the object 2 through the focus control circuit 9.

The judgment output from the judgment circuit 8 is the horizontal position of the object 2.

The robot 4 grasps and handles the object 2 by applying the XYZ coordinates in which the robot 4 should be positioned according to the object 2 to the robot 4 through the robot control circuit 10.

Although FIG. 2 shows the case where one object 2 is placed on the mounting table 3, as shown in FIG.
c and 2d are placed on the mounting table 3, an object close to the second camera device IH, that is, an object 2d with a large X-axis coordinate.
What is necessary is to sequentially focus and measure the objects and sequentially handle them by the robot 4.

In the case shown in FIG. 5, a plurality of objects 2a, 2b, and 2C are arranged in a line in the X coordinate direction at the -yI point of the Y coordinate with respect to the second camera device IH.
X+), 2b()'l, +x2).

2c (-yl, '+X3), but in this case, after measuring and handling the object 2C close to the second camera device IH, the object 2b is sequentially arranged.

What is necessary is to measure 2a and perform handling.

(7) Effects of the Invention As explained in detail above, the three-dimensional visual input device of the present invention makes it easy to accurately measure an object image and provide information necessary for positioning the robot to the object. It has characteristics that can be achieved through configuration.

[Brief explanation of the drawing]

FIG. 1 is an explanatory diagram for explaining the conventional object recognition method,
FIG. 2 is a perspective view showing the configuration of the three-dimensional visual input device of the present invention, FIG. 3 is a system diagram of the three-dimensional visual human power device of the present invention, and FIGS. FIG. 1, LH, IV--Camera device, 2.2a. 2b, 2c, 2d...object, 3...mounting table, 4...
・Robot, 5...XY coordinates, 6...Horizontal pattern measurement circuit, 7...Vertical pattern measurement circuit, 8
... Focus control circuit, 10... Robot control circuit.

Claims (1)

[Claims] In order to recognize the shape of an object, it has a first camera device arranged vertically to the object, and a second camera device arranged horizontally to the object, and the second camera device focuses on the outside. inputting position data obtained by measuring the object position by pattern recognition of the object image by the first camera device to the second camera device; A three-dimensional visual input device, characterized in that a three-dimensional shape position is recognized by an image obtained by focusing on the object.