JPH11184526A - Three-dimensional position correcting method and remote manipulator system using the method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To unnecessitate complicated image processing calculation and to inexpensively and easily improve accuracy in driving positioning of a remote manipulator. SOLUTION: A television camera 4 is installed on the manipulator, a symbol 6 installed near the moving target of the manipulator is photographed by this television camera 4, an image photographed in this case is displaied on a display screen 8, a cursor is moved and positioned near the feature point of the symbol displayed on the display screen 8, and its coordinate value is read. Based on the value read in this case, the position relation between the television camera 4 and the symbol 6 is operated and based on the arithmetic result, the driving target position of the manipulator is corrected once at least (main body 7 of computer).

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a three-dimensional position correction method suitable for positioning when driving a remote manipulator, and a remote manipulator system using the method.

[0002]

2. Description of the Related Art Positioning at the time of driving a remote manipulator has conventionally used the principle of position measurement by a television camera. As a method of measuring the three-dimensional position and orientation, based on the principle of three-dimensional photography applying single-photography evaluation, two television cameras take images of signs attached to the object to be measured, and each camera's coordinate system A method has been proposed for calculating the position and orientation of the measurement target in the ground coordinate system by obtaining the position of the image. This is described in detail in Aerospace Research Institute Material No. 681, "Positional Measurement of Dynamic Model in Wind Tunnel Using Optical Three-Dimensional Position Measurement Device".

[0003] In this method, two television cameras are required, and the camera positions must be accurately aligned and camera parameters must be calibrated in advance.

[0004] In addition, a sign attached to an object to be measured is photographed by using one single-lens camera, and characteristic points on the sign are extracted by a plurality of image processing to obtain positional information of a camera coordinate system. A method has also been proposed in which the Jacobian of the remote manipulator is corrected and calculated from the difference from the position on the camera coordinate system to correct the position and orientation relationship. Also in this case, processing a camera image to extract a feature point requires complicated calculation processing, software for executing the processing, and a high-performance computer.

[0005]

As described above, in the conventional method, it is necessary to perform accurate alignment with a plurality of television cameras to measure the position and orientation, or to perform image processing calculations on camera images. there were.

SUMMARY OF THE INVENTION The present invention has been made in view of the above circumstances, and a sign installed near a target position is photographed by a single monocular camera, and an operator extracts and positions the image of the feature point from the image information. By calculating using the position information of the target position and the position of the manipulator,
It is an object of the present invention to provide a three-dimensional position correction method that does not require complicated image processing calculations, and that can easily and inexpensively improve the drive positioning accuracy of a remote manipulator, and a remote manipulator system using the method.

[0007]

According to the three-dimensional position correcting method of the present invention, a television camera is set on a manipulator, and a sign placed near a moving target of the manipulator is photographed by the television camera. Display the image on the display screen, move the cursor near the characteristic point of the sign displayed on the display screen, position it, read its coordinate value, and based on the value read here, the TV camera and the sign And calculating the target position of the manipulator based on the calculation result at least once.

[0008] A remote manipulator system according to the present invention comprises a manipulator, a sign installed near a moving position of the manipulator, an input / output control device for instructing photographing of the sign via a television camera, and taking in the photographed image;
An arithmetic unit that displays an image captured via an input / output control device on a display monitor, reads the coordinates of the characteristic points via a display cursor, and calculates the positional relationship between the television camera and the sign based on the coordinate values. And a three-dimensional position correcting device for correcting the drive target position of the manipulator according to the calculation result at least once based on a predetermined correction formula.

As a result, it is possible to easily and inexpensively and easily improve the drive positioning accuracy of the remote manipulator without requiring complicated image processing and calculation as compared with the related art.

[0010]

FIG. 1 is a diagram showing an embodiment of the present invention.

In the figure, reference numeral 10 denotes a manipulator, which is constituted by six rotating joints (joint 1) and a beam (arm 2) and has six degrees of freedom. The manipulator 10 is connected to a computer main body 7 that calculates each joint angle according to a drive target position, and the computer main body 7 is remotely operated by controlling each joint 1.

A monocular television camera 4 capable of photographing the vicinity of the drive target position is provided near the distal end 3 of the manipulator 10. The video signal from the television camera 4 is transmitted to a display monitor 8 connected to the computer main body 7 and displayed. As shown in FIG.
It has a function of generating a cursor to be displayed on the screen superimposed on the image of the television camera 4 and moving or positioning the cursor on the screen according to the input of the operator. That is, the cursor is moved to the center of the characteristic point of the sign on the screen to position the marker, and the details will be described later.

In the figure, a, b, c, d denote feature points, and + denotes a cursor. Reference numeral 5 denotes an object, and reference numeral 6 denotes a sign placed near the object. A marker having three or more feature points for positioning shown in FIG. 4 is provided near a target point in a region where the manipulator 10 is driven. (A) is the plane of the sign,
(B) is a view from the side, in which the hatched area is black and the outside is white. In particular,
This is a part that leaves a white spot on a black dyed board.

FIG. 2 is a block diagram showing the internal configuration of the computer main unit in FIG. The computer main body 7 includes a CPU 71 serving as a control center, a main storage device (MMU) 72 in which programs and data are stored, and an input / output control device (IOC7).
3, 74) and a display controller (GDC 75), which are commonly connected via a system bus 76 composed of a plurality of signal lines for address, data, and control.

A flowchart shown in FIG. 5 is stored in the MMU 72 in the form of a program, and the CPU 71 reads out the program and sequentially executes it, that is, sequentially executes the arithmetic expression shown in FIG. Each of the input / output control devices 73 and 74 has a plurality of input / output ports. A port 731 is an input port for controlling an interface with the computer main unit 7 to capture a video signal captured by the television camera 4. A port 732 is a manipulator. 10 is connected, and in order to exchange data with the joint 1 or the arm 2, data is taken in or the CPU 7 is exchanged.
1 is a bi-directional port that outputs commands issued by

A port 741 of the input / output control device 74 is connected to a pointing device used for determining a coordinate position, which will be described later, and other ports are connected to various input / output devices such as an external file device (not shown). Connected. The display controller 75 writes the data processed by the CPU 71 including the image data of the sign taken in by the television camera 4 into the display memory 77, reads out the data in synchronization with the display raster, and displays it on the display monitor 8.

Hereinafter, the operation of the embodiment of the present invention will be described in detail with reference to FIGS. First, the remote manipulator 10 is roughly moved to the vicinity of the target position, and the sign 6 provided near the target location is photographed by the television camera 4 attached to the tip of the manipulator 10, and the computer body 7 (IOC73-CPU71-GDC) is taken.
75) is displayed on the display monitor 8. Next, the operator moves and positions the cursor generated on the screen to the center of the feature point of the sign image via the pointing device. The computer main body 7 determines the position of the image coordinate system by an operator's instruction, for example, a click of a pointing device such as a mouse, and the relative position with respect to the target object predicted when the manipulator 10 is at the original target position. The difference between the center of the marker feature point assumed from the posture and the image coordinate system position is calculated.

The computer main body 7 (CPU 71) further calculates the difference between the target position and attitude and the current position and attitude by performing a correction calculation of the Jacobian indicating the position and attitude of the manipulator 10 using the values. be able to.
FIG. 5 is a flowchart showing the processing flow at this time.
FIG. 6 shows a calculation formula corresponding to this flowchart.

That is, the operator moves the cursor by operating the pointing device at the center of the characteristic point position of the marker 6 displayed on the display monitor 8, and further clicks the pointing device, so that the computer main body 7 is displayed. Then, the coordinate position on the screen is recognized (step S1). After initializing the predicted position and orientation values (step S2), the computer main body 7 calculates the rotation matrix R based on the predicted position and orientation values (step S3 /
Equation in FIG. 6). Next, the computer main body 7 (CPU 71)
Calculation of the image coordinate system position (Equation in FIG. 6: P1x, P1y) of the marker feature point assumed from the position and orientation predicted values (step S
Perform 4). Further, the computer main body 7 calculates an error matrix N (expression in FIG. 6) between the measured value of the marker feature point position and the calculated value (step S5).

Then, calculation of the Jacobian J (formula in FIG. 6) and the correction direction matrix D (formula in FIG. 6) (steps S6 and S6)
7) is performed. Further, it is checked whether or not each element of the correction direction matrix D is sufficiently small (Step S8), the predicted position and orientation values are corrected (Step S9), and the result is fed back to the processing of Step S3, and the above processing is repeated. .

The arithmetic expressions shown in FIG. 6 are calculated by the computer main body 7, and POSX, POSY, and POSZ are
The relative position predicted value of the manipulator and the sign at the target position, θX, θY, θZ are the predicted relative posture values of the manipulator and the sign at the target position, P1X, P1Y are the image coordinate system positions of the sign feature points, CX , CY are constants determined by the resolution of the TV camera or display monitor, F1X, F1
1Y indicates each of the image coordinate system positions of the marker feature points measured by cursor positioning from the television camera image.

Based on this result, the obtained difference is directly driven as the manipulator position and posture correction amount, or is corrected as the manipulator origin position and posture data correction amount, and then the next operation is performed. An error that has been added to the position and orientation of the manipulator until then can be removed, and the positioning accuracy can be improved. The number of times of this correction may be one, but it is possible to further improve the positioning accuracy by sequentially performing the correction while moving.

[0023]

As described above, according to the present invention, a television camera is installed on a manipulator, a sign placed near a moving target of the manipulator is photographed by the television camera, and an image photographed here is displayed on a display screen. The cursor is moved to the vicinity of the characteristic point of the sign displayed on the display screen and positioned to read the coordinate value, and the positional relationship between the TV camera and the sign is determined based on the read value. The method is characterized in that the calculation is performed and the target position of the manipulator is corrected at least once based on the calculation result. By constructing a remote manipulator system that realizes this method, a single-lens camera becomes complicated. The positioning accuracy of the remote manipulator can be easily improved at low cost without performing complicated image processing calculations.

[Brief description of the drawings]

FIG. 1 is a diagram showing an embodiment of the present invention;

FIG. 2 is a block diagram showing an internal configuration of a computer main body shown in FIG. 1;

FIG. 3 is a conceptual diagram showing a method of positioning feature points on a television camera image, which is adopted in the present invention;

FIG. 4 is a diagram showing the shape of a sign used as a reference for positioning in the present invention;

FIG. 5 is a diagram showing a flow of processing for calculating a position / posture relation correction amount used in the present invention;

FIG. 6 is a diagram showing an arithmetic expression used in the present invention;

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 ... Joint (joint), 2 ... Beam part (arm), 3 ... Manipulator tip part, 4 ... TV camera, 5 ... Measurement object, 6 ... Sign, 7 ... Computer body, 8 ... Display monitor, 10 ... Manipulator, 71 CPU (Central Processing Unit), 72 Main Memory Unit (MMU), 73, 74 Input / Output Control Unit (IOC), 75 Display Controller (GD)
C), 76: system bus, 77: display memory, 73
1, 741 ... input port, 732 ... input / output port.

Claims (4)

[Claims] 1. A television camera is set on a manipulator, and a sign used as a reference for positioning, which is set near a movement target of the manipulator, is photographed by the television camera, and the photographed image is displayed on a display screen. The cursor is moved to the vicinity of the characteristic point of the sign displayed on the display screen and positioned to read the coordinate value, and the positional relationship between the TV camera and the sign is determined based on the read value. A three-dimensional position correction method, comprising: performing a calculation; and correcting the drive target position of the manipulator based on the calculation result at least once. 2. The three-dimensional position correction method according to claim 1, wherein the position correction includes a correction of a posture relationship. 3. An input / output control for instructing photographing of a manipulator, a sign installed near a moving position of the manipulator, and used as a reference for positioning, and a sign via a television camera, and taking in the photographed image. Displays the image captured via the device and the input / output control device on the display monitor, reads the coordinates of the characteristic points via the display cursor, and calculates the positional relationship between the TV camera and the sign based on the coordinate values. A remote manipulator system comprising: an arithmetic device that performs a correction of a drive target position of a manipulator according to a calculation result at least once based on a predetermined correction formula. 4. The three-dimensional position correction device corrects and calculates a Jacobian of a remote manipulator from a difference between position information on a television camera coordinate system and a position on the television camera coordinate system assumed from a target position. Claim 3
The remote manipulator system as described.