WO2020173240A1

WO2020173240A1 - Image acquisition apparatus calibration method and apparatus, computer device, and storage medium

Info

Publication number: WO2020173240A1
Application number: PCT/CN2020/072491
Authority: WO
Inventors: 孙高磊; 张文刚; 梅能华; 李相前; 罗小军; 吴丰礼
Original assignee: 广东拓斯达科技股份有限公司
Priority date: 2019-02-27
Filing date: 2020-01-16
Publication date: 2020-09-03
Also published as: CN109829953B; CN109829953A

Abstract

Disclosed in the present application are an image acquisition apparatus calibration method and apparatus, a computer device, and a storage medium. A method in an embodiment comprises: acquiring pixel coordinates of a preset feature point and corresponding robot coordinates when a robot translates to different positions in an area to be calibrated; according to the pixel coordinates of the preset feature point and the corresponding robot coordinates when the robot translates to different positions in the area to be calibrated, obtaining a translation position relationship between the pixel coordinates of the preset feature point and the robot coordinates; according to the pixel coordinates of the preset feature point and the corresponding robot coordinates when the robot rotates to different angles, acquiring pixel coordinates of the preset feature point and corresponding robot coordinates when the robot rotates to different angles, obtaining the position of the center of rotation; and according to the translation position relationship and the position of the center of rotation, obtaining a plane conversion relationship between the robot and an image acquisition apparatus so as to calibrate the image acquisition apparatus.

Description

Image acquisition device calibration method, device, computer equipment and storage medium. This application requires the priority of the Chinese patent application filed with the Chinese Patent Office with application number 201910146310.7 to be submitted to the Chinese Patent Office on February 27, 2019. The entire content of the application is incorporated into this by reference Applying. Technical field

This application relates to the field of machine vision technology, for example, to an image acquisition device calibration method, device, computer equipment, and storage medium. Background technique

In the process of grasping products by the robot's visual positioning equipment, for the hand-eye calibration method where the camera is installed at the end of the robot, that is, the camera moves with the robot, a tool point is usually corrected at the end of the robot. When the robot grabs products with inconsistent positions, the robot drives the camera to move so that the product is in the camera's field of view. Calculate the coordinates of the point on the product image pixel in the robot tool coordinate system, guide the robot to grab the product at the specified position, and ensure the consistency of the product grabbing position.

In the specific operation process, the robot tool points are usually displayed and calibrated by 3 or 4 points, which are realized by human observation. Many tool points displayed and calibrated have errors, and it is difficult to ensure the accuracy of the tool points. The accuracy of the calibration affects, resulting in low accuracy of the final robot vision positioning equipment. Summary of the invention

The present application provides a camera calibration method, device, computer equipment, and storage medium that can improve the calibration accuracy, so as to improve the robot vision positioning accuracy.

An embodiment of the application provides a method for calibrating an image acquisition device, where the image acquisition device is applied to a robot visual positioning device, and the method includes:

Acquiring the pixel coordinates of the preset feature points and the robot coordinates corresponding to the pixel coordinates of the preset feature points when the robot moves to different positions in the area to be calibrated;

According to the pixel coordinates of the preset feature points and the robot coordinates corresponding to the pixel coordinates of the preset feature points, obtain the translational position relationship between the pixel coordinates of the preset feature points and the robot coordinates; obtain when the robot rotates to different angles The pixel coordinates of the preset feature point and the robot coordinates corresponding to the pixel coordinates of the preset feature point;

Obtaining the rotation center position according to the pixel coordinates of the preset feature point at the different angles and the robot coordinates corresponding to the pixel coordinates of the preset feature point; According to the translational position relationship and the rotation center position, the plane conversion relationship between the robot and the image acquisition device is obtained to calibrate the image acquisition device.

An embodiment of the present application also provides a calibration device for an image acquisition device, the image acquisition device is applied to a robot vision positioning device, and the device includes:

The first information acquisition module is configured to acquire the pixel coordinates of the preset feature points and the robot coordinates corresponding to the pixel coordinates of the preset feature points when the robot moves to different positions of the area to be calibrated;

The translational relationship acquisition module is configured to obtain the translational position relationship between the pixel coordinates of the preset feature points and the robot coordinates according to the pixel coordinates of the preset feature points and the robot coordinates corresponding to the pixel coordinates of the preset feature points;

The second information acquiring module is configured to acquire the pixel coordinates of the preset feature points and the robot coordinates corresponding to the pixel coordinates of the preset feature points when the robot rotates to different angles;

A rotation center obtaining module, configured to obtain the rotation center position according to the pixel coordinates of the preset feature points at the different angles and the robot coordinates corresponding to the pixel coordinates of the preset feature points;

The calibration information acquisition module is configured to obtain the plane conversion relationship between the robot and the image acquisition device according to the translational position relationship and the rotation center position, so as to calibrate the image acquisition device.

An embodiment of the present application further provides a computer device, which includes a memory and a processor, the memory stores a computer program, and the processor implements the following steps when executing the computer program:

Obtaining the rotation center position according to the pixel coordinates of the preset feature point at the different angles and the robot coordinates corresponding to the pixel coordinates of the preset feature point;

According to the translational position relationship and the rotation center position, the plane conversion relationship between the robot and the image acquisition device is obtained, so as to calibrate the image acquisition device.

The embodiment of the present application also provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the following steps are implemented:

Acquiring the pixel coordinates of the preset feature points and the robot coordinates corresponding to the pixel coordinates of the preset feature points when the robot moves to different positions of the area to be calibrated; According to the pixel coordinates of the preset feature points and the robot coordinates corresponding to the pixel coordinates of the preset feature points, obtain the translational position relationship between the pixel coordinates of the preset feature points and the robot coordinates; obtain when the robot rotates to different angles The pixel coordinates of the preset feature point and the robot coordinates corresponding to the pixel coordinates of the preset feature point;

According to the translational position relationship and the rotation center position, the plane conversion relationship between the robot and the image acquisition device is obtained, so as to calibrate the image acquisition device. Description of the drawings

Figure 1 is an application scenario diagram of an image acquisition device calibration method in an embodiment of this application; Figure 2 is a schematic flow chart of an image acquisition device calibration method in an embodiment of this application; Figure 3 is a translational position relationship acquisition in an embodiment of this application A schematic flow diagram of the steps; FIG. 4 is a schematic diagram of the robot position translation in the XY axis direction in an embodiment of the application; FIG. 5 is a schematic diagram of the area to be calibrated equally divided in an embodiment of the application;

6 is a schematic flow chart of the step of obtaining the position of the rotation center in an embodiment of the application; FIG. 7 is a schematic flow chart of the step of obtaining the rotation center conversion relationship in an embodiment of the application; A schematic diagram of a circular feature point at an angle; FIG. 9 is a schematic diagram of a position of the robot during rotation in an embodiment of the application;

Fig. 10 is a schematic flow chart of a method for calibrating an image capture device in another embodiment of this application; Fig. 11 is a schematic diagram of the application of the method for calibrating an image capture device in an embodiment of this application; Fig. 12 is a schematic diagram of an image capture device calibration in an embodiment of this application Block diagram of the device;

Fig. 13 is an internal structure diagram of a computer device in an embodiment of the application. DETAILED DESCRIPTION OF THE EMBODIMENTS The following describes the application with reference to the drawings and embodiments. It should be understood that the specific embodiments described here are only used to explain the application, and are not used to limit the application.

The image acquisition device calibration method provided in this application can be applied to the application environment as shown in FIG. 1. In the process of grasping objects by the robot visual positioning equipment, the end of the robot 10 includes a movable tool arm 11, through which an image acquisition device 20 and a standard fixture 30 can be installed at the end of the robot 10 through the tool arm 11. A fixed feature point 40 is placed below the. In In an embodiment, the image acquisition device may be a camera.

In one embodiment, as shown in FIG. 2, a method for calibrating an image acquisition device is provided. The method is applied to the image acquisition device in the robot visual positioning equipment in FIG. 1 as an example for description. This embodiment provides The method includes the following steps:

Step 202: Obtain the pixel coordinates of the preset feature points and the robot coordinates corresponding to the pixel coordinates of the preset feature points when the robot moves to different positions of the area to be calibrated.

Step 204: Obtain the translational position relationship between the pixel coordinates of the preset feature point and the robot coordinates according to the pixel coordinates of the preset feature point and the robot coordinates corresponding to the pixel coordinates of the preset feature point.

Move a relative position toward the X axis at the current coordinate position of the robot, search for a given circular feature point in the camera, and record the coordinates of the robot and the center image coordinates of the circular feature point at this time. In the same way, the robot moves a relative position toward the Y axis at the current coordinate position of the robot, searches the camera for a given circular feature point, and records the coordinates of the robot and the center image coordinates of the circular feature point at this time. The rough translation position conversion relationship between the pixel coordinates of the circular feature points and the robot coordinates is obtained by the above method.

The area to be calibrated is equally divided into a preset number of sub-areas, and the pixel coordinates of the center of the preset number of sub-regions are extracted respectively, and the rough movement of the robot is calculated through the rough translation position conversion relationship and the pixel coordinates of the center of the preset number of sub-regions. position. Move the robot to the calculated rough positions, and recalculate the pixel coordinates of the circular feature points when the robot reaches different rough positions through image recognition, thereby obtaining the precise translational position between the pixel coordinates of the feature points and the robot coordinates relationship.

In this embodiment, the rough translation position conversion relationship may also be referred to as the first translation position conversion relationship, the precise translation position relationship may also be referred to as the second translation position conversion relationship, and the rough position may also be referred to as the target position.

Step 206: Obtain the pixel coordinates of the preset feature points when the robot rotates to different angles and the robot coordinates corresponding to the pixel coordinates of the preset feature points.

Step 208: Obtain the rotation center position according to the pixel coordinates of the preset feature points at different angles and the robot coordinates corresponding to the pixel coordinates of the preset feature points.

The camera is driven by the robot to rotate with a preset step length to obtain an arc composed of the center of the circular feature points when rotated to different angles, and the center coordinates of the arc are fitted to obtain the rough rotation center of the robot. Calculate the difference between the rough rotation center and the initial rotation position of the robot, and find the rotation center conversion relationship between the pixel coordinates of the circular feature point and the robot coordinates when the robot rotates.

The camera is driven by the robot to rotate in another fixed step. Every time the robot rotates, the robot coordinates corresponding to the pixel coordinates of the circular feature points at this time are obtained through the rotation center conversion relationship. Calculate the difference between the coordinates and the actual coordinates of the robot at this time, and compensate the calculated difference to the robot coordinates, and the robot moves to the specified position. In an embodiment, after the robot moves to the corresponding robot coordinates, according to the pixel coordinates of the circular feature point at this time and the pixel coordinates of the circular feature point when the robot is at the initial position, Obtain the robot coordinate error, and compensate the error to the robot coordinate. After the robot completes its rotation within 360 degrees, it fits all coordinate errors, and compensates the fitted coordinate errors to the rough rotation center of the robot coordinates, and calculates the difference between the compensated rotation center coordinates of the robot and the initial position coordinates , So that the robot can accurately rotate around the initial position of the circular feature point through the calculated difference.

Step 210: Obtain the plane conversion relationship between the robot and the image acquisition device according to the translational position relationship and the rotation center position, so as to calibrate the image acquisition device.

In the calibration method of the image acquisition device, the pixel coordinates of the feature points and the corresponding robot coordinates are preset according to the robot translation to different positions in the area to be calibrated to obtain the translational position relationship between the pixel coordinates of the preset feature points and the robot coordinates, and then According to the pixel coordinates of the feature points and the corresponding robot coordinates when the robot rotates to different angles, the rotation center position is obtained, and the plane conversion relationship between the robot and the image acquisition device is obtained according to the translational position relationship and the rotation center position, so as to collect images. The device is calibrated without the need for manual calibration of robot tool points, only the feature points need to be placed in the field of view of the image acquisition device. The above solution can improve the calibration accuracy, thereby improving the accuracy of the robot's visual positioning equipment, and does not require additional auxiliary hardware The equipment is simple and efficient, which can greatly reduce the difficulty of the operator to adjust the machine.

In one embodiment, as shown in FIG. 3, according to the pixel coordinates of the preset feature points and the robot coordinates corresponding to the pixel coordinates of the preset feature points, the translational position relationship between the pixel coordinates of the preset feature points and the robot coordinates is obtained , Including: Step 302, according to the robot coordinates corresponding to the pixel coordinates of the preset feature points and the pixel coordinates of the preset feature points when the robot translates to different preset positions along the X-axis direction and the Y-axis direction in the area to be calibrated, to obtain Roughly translate the positional relationship, the X-axis and Y-axis are mutually perpendicular axes on the same horizontal plane in the three-dimensional space coordinates; Step 304, divide the area to be calibrated into a preset number of sub-areas, and extract the preset number of sub-areas respectively The pixel coordinates of the center; Step 306: Obtain a preset number of rough positions according to the rough translation position relationship and the preset number of pixel coordinates of the center of the sub-region; Step 308: Obtain the preset number of rough positions when the robot reaches the preset number Pixel coordinates of the feature point; Step 310, according to the pixel coordinates of the preset feature point and the coordinates of the corresponding rough position when the robot reaches the preset number of rough positions, correct the rough translation position relationship to obtain the precise translation position relationship.

In one embodiment, according to the pixel coordinates of the preset feature points and the robot coordinates corresponding to the pixel coordinates of the preset feature points when the robot translates to different preset positions along the X-axis direction and the Y-axis direction in the area to be calibrated, The rough translational position relationship includes: when the ear harvesting robot translates from the current position to the preset first position along the X-axis, the first pixel coordinates of the preset feature points and the first pixel coordinates of the preset feature points of the robot corresponding to the first pixel coordinates One coordinate, where the preset first position is obtained by the dichotomy; the second pixel coordinate of the preset feature point and the second pixel coordinate of the preset feature point are acquired when the robot translates from the current position to the preset second position along the Y-axis direction The second coordinate of the robot corresponding to the pixel coordinate, where the preset The two positions are obtained by dichotomy; according to the first pixel coordinates of the preset feature points and the first pixel coordinates of the robot corresponding to the first pixel coordinates of the preset feature points, the second pixel coordinates of the preset feature points, and the values of the preset feature points The second coordinate of the robot corresponding to the second pixel coordinate obtains the rough translational position relationship. 4, the current coordinate of the robot ₂₀ (}; ₀₎ position a relative position of the X-axis, the adjustment value by dichotomy, the camera searches a given circular feature points, recorded robot coordinate 2O () And the center image coordinate of the circular feature point is ₀ (, ). In the same way, at 2oCxo,}; ()), move a relative position to the Y axis, adjust the value through the dichotomy, search for a given circular feature point in the camera, and record the robot coordinate QoO: ₂ }; ₂ ) And the center image coordinates of the circular feature point Po(u ₂ , v ₂ ) _o Calculate the conversion relationship A ₀ between corpse ₀ and 2o, and obtain P _0* A ₀ = Qo,

As shown in Figure 5, the area to be calibrated is equally divided into 9 areas in the camera's field of view, and the center coordinates of the 9 areas are lifted separately /V(, v ₀ ), P, \u _h Vj). /VO _f , V _f ), where (^ 8, by converting the moment

In one embodiment, as shown in FIG. 6, according to the pixel coordinates of the preset feature points at different angles and the robot coordinates corresponding to the pixel coordinates of the preset feature points, obtaining the rotation center position includes: Step 602: Obtain the rotation center transformation The relationship and the rough rotation center. The rotation center conversion relationship is used to characterize the relationship between the pixel coordinates of the preset feature points and the robot coordinates when the robot rotates; step 604, preset the first step according to the rotation center conversion relationship and the initial position of the robot Default for long rotation to different angles The pixel coordinates of the feature point are obtained, and the corresponding robot coordinates are obtained; step 606, the pixel coordinates of the preset feature points are obtained when the robot reaches the position of the corresponding robot coordinates; step 608, the preset is based on the robot reaching the position of the corresponding robot coordinates The pixel coordinates of the feature points and the pixel coordinates of the feature points are preset at the initial position of the robot to obtain the robot coordinate error; Step 610: Fit the coordinate error within the full angle range, and compensate the fitted coordinate error to a rough rotation The coordinates of the center, get the precise rotation center position.

In one embodiment, as shown in FIG. 7, obtaining the rotation center conversion relationship and the rough rotation center includes: Step 702, preset the pixel of the feature point when the robot rotates to a different angle with a preset second step length at the initial position Coordinates, fit the pixel coordinates to obtain the rough rotation center; step 704, calculate the rotation error between the coordinates of the rough rotation center and the pixel coordinates of the preset feature point at the initial position; step 706, correspond to the rotation error, the pixel coordinates, and the pixel coordinates The robot coordinates of, get the rotation center conversion relationship.

In one embodiment, according to the pixel coordinates of the preset feature points when the robot rotates to a different angle with the preset second step length at the initial position, before fitting the pixel coordinates to obtain the rough rotation center, the method further includes: acquiring the robot in the area to be calibrated When rotating around the Z axis in the positive direction, the image acquisition device searches for the maximum forward angle of the preset feature point, and when the robot rotates around the Z axis in the reverse direction, the image acquisition device searches for the maximum reverse angle of the preset feature point, the Z axis Is the axis perpendicular to the horizontal plane in the three-dimensional space coordinate axis; according to the maximum forward angle and the maximum reverse angle, the preset second step length of the robot is obtained.

Where D _x ', D _y

D _x \D _y

P ₂ = P, +

D _x \D _{y The} robot returns to the initial position 2 ₀ (, ), rotates around the Z axis, the relative angle range is 0-360°, the step length de is set to a fixed value, and the corresponding circular feature point is calculated through the rotation center conversion relationship ₂ . The robot coordinates 2 ₂ O _G ,}; _G ), Q ₂ ix _h yi)..., Q ₂ (x _m ,y _m ), Q ₂ (x _m ,y _m ) when the center image coordinate ₀ (« ₀ ,) is unchanged The robot coordinates corresponding to corpse ₀ (« ₀ ,%) are (x _e ,;y _c ),

x _c = u _Q a ² n + v _Q a ² _2i + a ² _3\

_<<

y _c = u ₀ a ² i2 + v ₀ a ² 22 + a ² 32

_Xm = x _c + ((x _0- x _c ) cos(md _d )-(y ₀ y _c ) sm(md _d ))

_<<

y _m = y _c + (Oo— A) sin(md ₀ ) + (y ₀ y _c ) cos(md _e ))

Move the robot to the calculated coordinate positions Q ₂ {x ₀ ,yo) _^ Qiixuyi). Q ₂ (x _m ,y _m ) After the robot moves in place, calculate the actual and theoretical coordinates of the circular feature points. ₀ (« ₀ The robot coordinate error between,) will be compensated to the robot coordinate. In one embodiment, the robot moves to the calculated coordinate positions Q ₂ (x ₀ ,yo) _^ 22(, ), ., 220m Jm), and then recognizes the corresponding circular feature point center image coordinates A(M ₀ , v ₀ ), PsiM _hV] ), ., P ₃ (u _m , v _m ), calculate the actual circular feature point center image coordinate value and the theoretical value of the robot coordinate error of ₀ (M ₀ , V ₀ ). Calculate the robot coordinate data after the robot is rotated according to the rough translation position conversion relationship, and calculate the conversion relationship between the pixel coordinates of the circular feature point and the robot coordinates after the rotation. The 2o coordinate of the robot rotating around corpse ₀ («0,) is transformed into Q'oixojo) _^ Q '

V ₀ is the corresponding circular feature point center image coordinate and ₀ coordinate value,,

Coordinate value, d㊀ is a fixed step value, O^m^n and 11.

Where V /o is the coordinate value of the corresponding circular feature point center image coordinate, and i/ ³ m is the coordinate value of coordinate ₃ .

In this embodiment, the body ₃ is the actual value of the image coordinate of the center of the circular feature point.

In one embodiment, after the coordinate error after fitting is compensated to the coordinates of the rough rotation center to obtain the precise rotation center position, the method further includes: calculating the difference between the coordinates of the precise rotation center position and the initial position robot coordinates According to the translational position relationship and the rotation center position, obtaining the plane conversion relationship between the robot and the image acquisition device includes: obtaining the plane conversion relationship between the robot and the image acquisition device according to the translation position relationship, the rotation center position and the difference.

Calculate the difference between the arc center coordinates and the initial position to find out the conversion relationship A ₂ between the pixel coordinates of the circular feature point at the rotation center of the robot and the robot coordinates in the translation process of step (2). (4) Solve the precise rotation center position of the robot. Every time the robot rotates, the conversion relationship A ₂ between the rotation center and the robot coordinates obtained in step (3) is used to compensate the calculated difference to the robot coordinates, and the robot moves to the specified Position; After the robot moves in place, calculate the robot coordinate error between the actual coordinate of the circular feature point and the theoretical coordinate, and compensate the error to the robot coordinate; After the robot completes rotation within 360 degrees, fit all the robot after the rotation error compensation For the rotation center of the coordinates, calculate the difference between the rotation center coordinates and the initial position of the robot after fitting, so that the robot can accurately rotate around the initial position of the circular feature point through the calculated value. (5) Solve the precise plane conversion relationship between the robot and the camera, and calculate the plane conversion relationship between the robot and the camera according to the step (2) the precise translation position relationship between the robot and the camera and the step (4) the precise rotation center position of the robot, and the circular feature point center image Coordinates /MM ₀ , V ₀ ), PAU _UV! ). Pi{u _h v^

In one embodiment, as shown in FIG. 11, it is assumed that the coordinates of the position where the robot grabs the product at the camera are (5^:, 5^; y, 5^r), and the feature point coordinates of the template product image in the camera field of view Is (M/w, M/?};, M/?r), the robot grabs the template product position coordinates as (Tqx, Tqy, rgr), then when the robot takes a photo at the position (V^x, Vqy, Vgr) , The coordinates (Rqx, Rqy, Rqr) of any product grabbing position can be calculated by identifying the coordinates of any product image feature point as (A//w, A/p};, A/pr) in the camera field of view.

turn off

Among them, Q Vqr Sqr Mqx M^; y represents the robot coordinate value corresponding to the feature point of the template image, and Nqx Nqy represents the coordinate value of the robot rotation center corresponding to any product image feature point.

It should be understood that although the multiple steps in the flowcharts of FIGS. 2-3 and 6-7 are displayed in sequence as indicated by the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated in this article, there is no strict order for the execution of these steps, and these steps can be executed in other orders. Moreover, at least part of the steps in FIGS. 2-3 and 6-7 may include multiple sub-steps or multiple stages, and these sub-steps or stages are not necessarily executed at the same time, but may be executed at different times. The order of execution of these sub-steps or stages is not necessarily performed sequentially, but may be performed alternately or alternately with other steps or at least part of the sub-steps or stages of other steps.

In one embodiment, as shown in FIG. 12, an image acquisition device calibration device is provided, including: a first information acquisition module 1202, a translation relationship acquisition module 1204, a second information acquisition module 1206, a rotation center acquisition module 1208, and Calibration information acquisition module 1210. Among them, the first information acquisition module is set Set to obtain the pixel coordinates of the preset feature points and the robot coordinates corresponding to the pixel coordinates of the preset feature points when the robot moves to different positions of the area to be calibrated. The translational relationship acquisition module is configured to obtain the translational position relationship between the pixel coordinates of the preset feature point and the robot coordinates according to the pixel coordinates of the preset feature point and the robot coordinates corresponding to the pixel coordinates of the preset feature point. The second information acquiring module is configured to acquire the pixel coordinates of the preset feature points when the robot rotates to different angles and the robot coordinates corresponding to the pixel coordinates of the preset feature points. The rotation center obtaining module is set to obtain the position of the rotation center according to the pixel coordinates of the preset feature points at different angles and the corresponding robot coordinates. The calibration information acquisition module is set to obtain the plane conversion relationship between the robot and the image acquisition device according to the translational position relationship and the rotation center position, so as to calibrate the image acquisition device.

In one embodiment, the translational relationship acquisition module is configured to obtain the pixel coordinates of the preset feature points and the robot coordinates according to the pixel coordinates of the preset feature points and the robot coordinates corresponding to the pixel coordinates of the preset feature points in the following manner The translational position relationship: According to the robot in the area to be calibrated along the X-axis direction and the Y-axis direction to preset different positions, the pixel coordinates of the preset feature points and the corresponding robot coordinates are obtained to obtain the first translational position relationship. The and Y-axis are mutually perpendicular axes on the same horizontal plane in the three-dimensional space coordinates; divide the area to be calibrated into a preset number of sub-areas, and extract the pixel coordinates of the center of the preset number of sub-areas respectively; And the pixel coordinates of the center of the preset number of sub-regions to obtain the preset number of corresponding rough positions; obtain the pixel coordinates of the preset feature points when the robot reaches the preset number of rough positions; according to when the robot reaches the preset number of rough positions The pixel coordinates of the feature point and the coordinates of the corresponding rough position are preset, and the first translational position relationship is corrected to obtain the second translational position relationship.

In one embodiment, the translational relationship acquisition module is configured to preset the pixel coordinates of the feature points and the corresponding robot when the robot translates in the X-axis direction and the Y-axis direction to preset different positions in the area to be calibrated in the following manner Coordinates to obtain the first translational position relationship: obtain the first pixel coordinates of the preset feature point and the first pixel coordinates of the preset feature point when the robot translates from the current position along the X-axis direction to the preset first position One coordinate, where the preset first position is obtained by the dichotomy; the second pixel coordinate of the preset feature point and the second pixel coordinate of the preset feature point are acquired when the robot translates from the current position to the preset second position along the Y-axis direction The second coordinate of the robot corresponding to the pixel coordinates, where the preset second position is obtained by dichotomy; the first pixel coordinates of the preset feature point and the first pixel coordinates of the preset feature point corresponding to the robot first coordinate and the preset The second pixel coordinates of the feature point and the second pixel coordinates of the robot corresponding to the second pixel coordinates of the preset feature point obtain the first translational position relationship.

In one embodiment, the rotation center obtaining module is configured to obtain the rotation center conversion relationship and the first rotation center, and the rotation center conversion relationship is used to characterize the relationship between the pixel coordinates of the preset feature point and the robot coordinates when the robot rotates; The rotation center conversion relationship and the pixel coordinates of the preset feature points when the robot rotates to different angles with the preset first step at the initial position, to obtain the corresponding robot sitting Obtain the pixel coordinates of the preset feature point when the robot reaches the position of the corresponding robot coordinate; According to the pixel coordinates of the preset feature point when the robot reaches the position of the corresponding robot coordinate and the preset feature point of the robot at the initial position The pixel coordinates are used to obtain the robot coordinate error; the coordinate error in the full angle range is fitted, and the fitted coordinate error is compensated to the coordinates of the first rotation center to obtain the position of the second rotation center.

In one embodiment, the rotation center obtaining module is configured to obtain the rotation center conversion relationship and the first rotation center in the following manner: preset the pixels of the feature points when the robot rotates to a different angle with a preset second step length at the initial position Coordinates, fit the pixel coordinates to obtain the first rotation center; calculate the rotation error between the coordinates of the first rotation center and the pixel coordinates of the preset feature point at the initial position; obtain the rotation according to the rotation error, the pixel coordinates and the corresponding robot coordinates Center conversion relationship.

In one embodiment, the rotation center obtaining module is further configured to obtain the maximum forward angle of the preset feature point searched by the image acquisition device when the robot rotates around the Z axis in the positive direction in the area to be calibrated, and the robot rotates around the Z axis in the reverse direction When the image acquisition device searches for the maximum reverse angle of the preset feature point, the Z axis is the axis perpendicular to the horizontal plane in the three-dimensional space coordinate axis; according to the maximum forward angle and the maximum reverse angle, the robot’s preset second step is obtained long.

In one embodiment, the rotation center obtaining module is further configured to calculate the difference between the coordinates of the precise rotation center position and the coordinates of the initial position, and the calibration information obtaining module is configured to calculate the difference between the positional relationship of the translation, the position of the rotation center, and the difference. , Obtain the plane conversion relationship between the robot and the image acquisition device.

Regarding the specific definition of the image acquisition device calibration device, please refer to the above definition of the image acquisition device calibration method, which will not be repeated here. The various modules in the calibration device of the image acquisition device described above can be implemented in whole or in part by software, hardware, and combinations thereof. The foregoing multiple modules may be embedded in the form of hardware or independent of the processor in the computer device, or may be stored in the memory of the computer device in the form of software, so that the processor can invoke and execute the operations corresponding to the foregoing multiple modules.

In an embodiment, a computer device is provided. The computer device may be a terminal, and its internal structure diagram may be as shown in FIG. 13. The computer equipment includes a processor, a memory, a network interface, a display screen and an input device connected through a system bus. In this embodiment, the processor of the computer device is used to provide calculation and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The network interface of the computer device is used to communicate with an external terminal through a network connection. The computer program is executed by the processor to realize a calibration method of the image acquisition device. The display screen of the computer device may be a liquid crystal display or an electronic ink display screen, and the input device of the computer device may be a touch layer covered on the display screen, or it may be a button, a trackball or a touch pad provided on the shell of the computer device , It can also be an external keyboard, touchpad or mouse. Those skilled in the art can understand that the structure shown in FIG. 13 is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. In an embodiment, the computer device may include more or fewer components than shown in the figure, or combine certain components, or have a different component arrangement.

In one embodiment, a computer device is provided, including a memory and a processor, the memory stores a computer program, and the processor executes the computer program to implement the image acquisition device calibration method in any embodiment.

In one embodiment, a computer-readable storage medium is provided, and a computer program is stored thereon, and the computer program is executed by a processor to implement the image acquisition device calibration method in any embodiment.

A person of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be implemented by instructing relevant hardware through a computer program. The computer program can be stored in a non-volatile computer readable storage. In the medium, when the computer program is executed, it may include the process of the embodiment of the foregoing method. In an embodiment, any reference to memory, storage, database, or other media used in the multiple embodiments provided in this application may include non-volatile and/or volatile memory. Non-volatile memory may include read-only memory (Read-Only Memory, ROM), programmable ROM (Programmable Read-Only Memory, PROM), electrically programmable ROM (Erasable Programmable Read-Only Memory, EPROM), and electrically erasable Except for programmable ROM (Electrically Erasable Programmable Read-Only Memory, EEPROM) or flash memory. Volatile memory can include random access memory (Random Access Memory, RAM) or external cache memory. As an illustration and not a limitation, RAM is available in various forms, such as static RAM (Static Random-Access Memory, SRAM), dynamic RAM (Dynamic Random Access Memory, DRAM), synchronous DRAM (Synchronous Dynamic Random Access Memory, SDRAM), Double Data Rate SDRAM (Double Data Rate Synchronous Dynamic Random Access Memory, DDRSDRAM), Enhanced SDRAM (Enhanced Synchronous Dynamic Random Access Memory, ESDRAM), Synchronous Link DRAM (Synchlink Dynamic Random Access Memory, SLDRAM), Memory Bus Direct RAM ( Rambus Direct Dynamic Random Access Memory (RDRAM), Direct Rambus Dynamic Random Access Memory (DRDRAM), and Rambus Dynamic Random Access Memory (RDRAM), etc.

Claims

Rights request

1. A method for calibrating an image acquisition device, wherein the image acquisition device is applied to a robot visual positioning device, the method comprising:

Acquiring the pixel coordinates of the preset feature point and the robot coordinates corresponding to the pixel coordinates of the preset feature point when the robot moves to different positions of the area to be calibrated;

Obtain the translational position relationship between the pixel coordinates of the preset feature point and the robot coordinates according to the pixel coordinates of the preset feature point and the robot coordinates corresponding to the pixel coordinates of the preset feature point; obtain the robot rotation The pixel coordinates of the preset feature point and the robot coordinates corresponding to the pixel coordinates of the preset feature point at different angles;

1. The method according to claim 1, wherein the pixel coordinates of the preset feature point are obtained according to the pixel coordinates of the preset feature point and the robot coordinates corresponding to the pixel coordinates of the preset feature point The translational positional relationship with robot coordinates, including:

According to the pixel coordinates of the preset feature point and the robot coordinates corresponding to the pixel coordinates of the preset feature point when the robot translates in the X-axis direction and the Y-axis direction to preset different positions in the area to be calibrated, In the first translational position relationship, the X axis and the Y axis are mutually perpendicular axes on the same horizontal plane in three-dimensional space coordinates;

Dividing the area to be calibrated into a preset number of sub-areas, and respectively extracting the pixel coordinates of the center of the preset number of sub-areas;

Obtaining a preset number of target positions according to the first translational position relationship and the pixel coordinates of the center of the preset number of sub-regions;

Respectively acquiring the pixel coordinates of the preset feature points when the robot reaches the preset number of target positions;

According to the pixel coordinates of the preset feature point and the coordinates of the corresponding target position when the robot reaches the preset number of target positions, the first translational position relationship is corrected to obtain a second translational position relationship.

3. The method according to claim 2, wherein the pixel coordinates of the preset feature points and the pixel coordinates of the preset feature points when the robot translates to preset different positions in the X-axis direction and the Y-axis direction in the area to be calibrated respectively The robot coordinates corresponding to the pixel coordinates of the preset feature points to obtain the first translational position relationship includes: Acquiring the first pixel coordinate of the preset feature point and the first pixel coordinate of the robot corresponding to the first pixel coordinate of the preset feature point when the robot translates from the current position along the X-axis direction to the preset first position, Wherein, the preset first position is obtained by dichotomy;

Acquiring the second pixel coordinates of the preset feature point and the second pixel coordinates of the robot corresponding to the second pixel coordinates of the preset feature point when the robot translates from the current position along the Y-axis direction to the preset second position, Wherein, the preset second position is obtained by dichotomy;

According to the first pixel coordinates of the preset feature point, the first coordinate of the robot corresponding to the first pixel coordinates of the preset feature point, the second pixel coordinates of the preset feature point, and the value of the preset feature point The second coordinate of the robot corresponding to the second pixel coordinate obtains the first translational position relationship.

4. The method according to claim 1, wherein said obtaining the rotation center position according to the pixel coordinates of the preset feature point at the different angles and the robot coordinates corresponding to the pixel coordinates of the preset feature point, Include:

Obtaining a rotation center conversion relationship and a first rotation center, where the rotation center conversion relationship is used to characterize the relationship between the pixel coordinates of the preset feature point and the robot coordinates when the robot rotates; the conversion relationship is based on the rotation center; The pixel coordinates of the preset feature points when the robot rotates to a different angle with a preset first step in the initial position, to obtain the corresponding robot coordinates;

Acquiring the pixel coordinates of the preset feature point when the robot reaches the position of the corresponding robot coordinate;

Obtain the robot coordinate error according to the pixel coordinates of the preset feature point when the robot reaches the position where the corresponding robot coordinate is located and the pixel coordinates of the preset feature point when the robot is in the initial position;

Fitting the coordinate errors in the full angle range, and compensating the fitted coordinate errors to the coordinates of the first rotation center to obtain the position of the second rotation center.

5. The method according to claim 4, after said compensating the fitted coordinate error to the coordinate of the first rotation center and obtaining the position of the second rotation center, the method further comprises:

Calculating the difference between the coordinates of the position of the second rotation center and the coordinates of the initial position;

The obtaining the plane conversion relationship between the robot and the image acquisition device according to the translational position relationship and the rotation center position includes:

According to the translational position relationship, the rotation center position, and the difference value, the plane conversion relationship between the robot and the image acquisition device is obtained.

6. The method according to claim 4, wherein said obtaining the rotation center conversion relationship and the second A center of rotation, including:

Fitting the pixel coordinates to obtain the first rotation center according to the pixel coordinates of the preset feature points when the robot rotates to a different angle with a preset second step length at the initial position;

Calculating the rotation error between the coordinates of the first rotation center and the pixel coordinates of the preset feature point at the initial position;

According to the rotation error, the pixel coordinates of the preset feature point when the robot reaches the preset number of target positions, and the robot coordinates corresponding to the pixel coordinates, a rotation center conversion relationship is obtained.

7. The method according to claim 6, said fitting the pixel coordinates according to the pixel coordinates of the preset feature point when the robot rotates to a different angle with a preset second step length at the initial position Before getting the rough rotation center, it also includes:

Obtain the maximum positive angle of the preset feature point when the robot rotates around the Z-axis in the area to be calibrated, the image acquisition device searches for the maximum positive angle of the preset feature point, and the robot in the area to be calibrated around the Z-axis in the opposite direction When rotating, the image acquisition device searches for the maximum reverse angle of the preset feature point, and the Z axis is the axis perpendicular to the horizontal plane in the three-dimensional space coordinate axis;

Obtain the preset second step length according to the maximum forward angle and the maximum reverse angle.

8. A calibration device for an image acquisition device, wherein the image acquisition device is applied to a robot visual positioning equipment, and the device includes:

The first information acquisition module is configured to acquire the pixel coordinates of the preset feature points and the corresponding robot coordinates when the robot moves to different positions in the area to be calibrated;

The translational relationship acquisition module is configured to obtain the translational position between the pixel coordinates of the preset feature point and the robot coordinates according to the pixel coordinates of the preset feature point and the robot coordinates corresponding to the pixel coordinates of the preset feature point Relationship

The second information acquisition module is configured to acquire the pixel coordinates of the preset feature point and the robot coordinates corresponding to the pixel coordinates of the preset feature point when the robot rotates to different angles;

The rotation center obtaining module is set to obtain the rotation center position according to the pixel coordinates of the preset feature points at the different angles and the robot coordinates corresponding to the pixel coordinates of the preset feature points; the calibration information obtaining module is set to be based on The translational position relationship and the rotation center position are used to obtain a plane conversion relationship between the robot and the image acquisition device, so as to calibrate the image acquisition device.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, and the processor implements the method according to any one of claims 1 to 7 when the computer program is executed. 10. A computer-readable storage medium storing a computer program, which when executed by a processor implements the method according to any one of claims 1 to 7.