CN111028298A - Convergent binocular system for rigid coordinate system space transformation calibration - Google Patents

Abstract

The invention discloses a convergent binocular system used for the calibration of rigid body coordinate system space transformation, belonging to the technical field of rigid body coordinate system space transformation and positioning. The invention solves the problems of slow calibration speed and poor accuracy when the existing tool with poor rigidity performs the calibration of the calibration system space transformation. The present invention mainly adopts a convergent binocular system, two cameras are placed in the form of intersecting optical axes, and adjusting the relative poses of the two cameras can flexibly change the size of the intersecting visual space to adapt to the size and calibration resolution requirements of different rigid bodies, Through the relevant camera calibration technology, the camera space coordinate system is established. Through certain image processing algorithms, the camera space coordinates of any point in the intersecting visual measurement space can be accurately represented, and the rigid body can be controlled to perform certain steps of translation and rotation. The pose of the keypoint coordinate system associated with the rigid body in a known coordinate system. The invention is suitable for calibrating the position and attitude of the rigid body.

Description

A Converged Binocular System for Calibration of Rigid Body Coordinate System Space Transformation

technical field

The invention belongs to the technical field of rigid body coordinate system space transformation and positioning, and particularly relates to a convergent binocular system used for rigid body coordinate system space transformation calibration.

Background technique

The space transformation of rigid body coordinate system is the basis of rigid body kinematics. In the field of robotics, the accuracy of the space transformation of rigid body coordinate system directly affects the working performance of the robot. Due to the error of mechanical manufacturing and assembly, the design size does not match the actual size. Kinematics calibration must be performed before being put into use. Usually, the manufacturer will use the laser tracker to calculate the actual key dimensions through standardized steps and robust algorithms to ensure the performance of the robot. However, the end tools used by users vary widely, how to be economical and fast There has been no general solution for the coordinate transformation relationship between the key points of the end tool and the center of the robot end flange. In the field of surgical robot navigation, infrared optical navigation is the most common technology, but the infrared optical navigation system can generally only accurately track the standard tools provided by the manufacturer, so the standard tools are usually fixed on the actual tools during use. For tools with simple shapes and good rigidity, some manufacturers have proposed a convenient and quick principal point rotation calibration method to achieve accurate conversion between the standard tool coordinate system and the actual tool key point coordinate system, but for tools with complex shapes and poor rigidity There are still problems of slow calibration speed and poor accuracy.

SUMMARY OF THE INVENTION

The present invention is for the problems of slow calibration speed and poor accuracy when the existing tool with poor rigidity is performing the calibration space transformation and calibration. A convergent binocular system for calibration of rigid body coordinate system space transformation is proposed.

A convergent binocular system for spatial transformation and calibration of a rigid coordinate system according to the present invention includes a camera a1, two plane light sources 2, two light source fixing plates 3, a cross adjustment table a, and a cross adjustment table b , camera b5, base 8 and host computer 9;

The cross adjustment table a includes a longitudinal sliding table a10 and a transverse sliding table a11; the transverse sliding table a11 is arranged on the upper side of the longitudinal sliding table a10, and the sliding direction of the transverse sliding table a11 is perpendicular to the sliding direction of the longitudinal sliding table a10;

The cross adjustment table b includes a transverse sliding table b6 and a longitudinal sliding table b7; the transverse sliding table b6 is arranged on the longitudinal sliding table b7, and the sliding direction of the transverse sliding table b6 is perpendicular to the sliding direction of the longitudinal sliding table b7;

The cross adjustment table a, the cross adjustment table b and the two light source fixing plates 3 are all fixed on the upper surface of the base 8 to form a visual measurement space; the visual measurement space is used to place the rigid body to be calibrated;

The sliding direction of the longitudinal sliding table a10 is perpendicular to the sliding direction of the longitudinal sliding table b7;

The sliding direction of the lateral slide a11 is perpendicular to the sliding direction of the lateral slide b6;

Two light source fixing plates 3 are arranged adjacently, and the two light source fixing plates 3 are both fixed with the plane light source 2;

The sliding directions of the longitudinal sliding table a10 and the longitudinal sliding table b7 are respectively perpendicular to the two light source fixing plates 3;

A camera a1 is fixed on the lateral slide a11, and the shooting surface of the camera a1 is opposite to a plane light source;

A camera b5 is fixed on the lateral slide b6, and the shooting surface of the camera b5 is opposite to another plane light source; the camera b5 and the camera a1 are used for image acquisition of the rigid body to be calibrated in the visual measurement space;

The camera b5 and the camera a1 upload the collected images of the rigid body to be calibrated to the host computer 9 respectively;

Four motors are used to control the lateral slide b6, the longitudinal slide b7, the longitudinal slide a10 and the lateral slide a11 in one-to-one correspondence;

The host computer 9 is used to quickly match the three-dimensional shape of the image of the rigid body 4 to be calibrated, calculate the coordinates of the key points to be calibrated in the visual measurement space, and use the pictures of different poses of the rigid body to obtain the coordinates of multiple sets of key points in the visual measurement space. The coordinates of a certain point of the rigid body in the known external coordinate system with multiple groups of rigid bodies, and then obtain the corresponding relationship between the visual measurement space and the external known coordinate system, so as to calculate the key points associated with the rigid body to be calibrated in the external known coordinate system. s position.

Further, the host computer 9 is also used to judge the ratio of the pixel area of the rigid body image to be calibrated to the entire width, and when the ratio is not within the ratio threshold range, send control signals to one or more of the four motors to control the longitudinal direction. One or more of the slide a10, the vertical slide b7, the horizontal slide b6, or the vertical slide b7 are moved until the ratio of the pixel area of the rigid body image to be calibrated to the entire frame is within the proportional threshold range.

Further, the threshold range of the ratio of the pixel area of the rigid body image to be calibrated to the entire width is 30%-50%.

Further, the control signal input ends of the four motors are respectively connected with the host computer 9 through data lines;

The host computer 9 is also used to judge whether the rigid body 4 to be calibrated is complete in the image, and when the rigid body 4 to be calibrated is incomplete in the image, it sends control signals to one or more of the four motors to control the longitudinal slides a10, One or more of the longitudinal slide b7, the lateral slide b6 or the longitudinal slide b7 are moved until a complete image of the rigid body to be calibrated is obtained.

The convergent binocular system used for the spatial transformation and calibration of the rigid body coordinate system of the present invention can quickly and accurately calibrate the pose of the key point coordinate system associated with the rigid body in the known coordinate system. The present invention mainly adopts a convergent binocular system, two cameras are placed in the form of intersecting optical axes, and adjusting the relative poses of the two cameras can flexibly change the size of the intersecting visual space to adapt to the size and calibration resolution requirements of different rigid bodies, Through the relevant camera calibration technology, the camera space coordinate system is established. Through certain image processing algorithms, the camera space coordinates of any point in the intersecting visual measurement space can be accurately represented, and the rigid body can be controlled to perform certain steps of translation and rotation. The pose (position pose) of the keypoint coordinate system associated with the rigid body in a known coordinate system. The invention has the advantages of high sensitivity, adjustable measurement space, strong usability, wide application range, etc., effectively reduces the cost of the calibration process, and improves the calibration efficiency and calibration accuracy.

Description of drawings

FIG. 1 is a schematic structural diagram of a convergent binocular system for calibration of rigid body coordinate system space transformation according to the present invention;

2 to 4 are schematic diagrams of adjusting the position of the camera a or the position of the camera b to visually measure the spatial position and size change; wherein,

Figure 2 is a schematic diagram of the spatial position and size of visual measurement when the distances between camera a and camera b and the rigid body to be calibrated are equal;

3 is a schematic diagram of the spatial position and size of visual measurement when the distance between the camera a and the rigid body to be calibrated is greater than the distance between the camera b and the rigid body to be calibrated;

4 is a schematic diagram of the spatial position and size of visual measurement when the distance between the camera a and the rigid body to be calibrated is smaller than the distance between the camera b and the rigid body to be calibrated.

Detailed ways

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative work fall within the protection scope of the present invention.

It should be noted that the embodiments of the present invention and the features of the embodiments may be combined with each other under the condition of no conflict.

The present invention will be further described below with reference to the accompanying drawings and specific embodiments, but it is not intended to limit the present invention.

Embodiment 1: The present embodiment is described below with reference to FIG. 1 and FIG. 2 . The present embodiment describes a convergent binocular system for spatial transformation and calibration of a rigid coordinate system. The system includes a camera a1 and two plane light sources 2 , two light source fixing plates 3, a cross adjustment table a, a cross adjustment table b, a camera b5, a base 8 and a host computer 9;

The cross adjustment table a includes a longitudinal sliding table a10 and a transverse sliding table a11; the transverse sliding table a11 is arranged on the upper side of the longitudinal sliding table a10, and the sliding direction of the transverse sliding table a11 is perpendicular to the sliding direction of the longitudinal sliding table a10;

The cross adjustment table b includes a transverse sliding table b6 and a longitudinal sliding table b7; the transverse sliding table b6 is arranged on the longitudinal sliding table b7, and the sliding direction of the transverse sliding table b6 is perpendicular to the sliding direction of the longitudinal sliding table b7;

The cross adjustment table a, the cross adjustment table b and the two light source fixing plates 3 are all fixed on the upper surface of the base 8 to form a visual measurement space; the visual measurement space is used to place the rigid body to be calibrated;

The sliding direction of the longitudinal sliding table a10 is perpendicular to the sliding direction of the longitudinal sliding table b7;

The sliding direction of the lateral slide a11 is perpendicular to the sliding direction of the lateral slide b6;

Two light source fixing plates 3 are arranged adjacently, and the two light source fixing plates 3 are both fixed with the plane light source 2;

The sliding directions of the longitudinal sliding table a10 and the longitudinal sliding table b7 are respectively perpendicular to the two light source fixing plates 3;

A camera a1 is fixed on the lateral slide a11, and the shooting surface of the camera a1 is opposite to a plane light source;

A camera b5 is fixed on the lateral slide b6, and the shooting surface of the camera b5 is opposite to another plane light source; the camera b5 and the camera a1 are used for image acquisition of the rigid body to be calibrated in the visual measurement space;

The camera b5 and the camera a1 upload the collected images of the rigid body to be calibrated to the host computer 9 respectively;

Four motors are used to control the lateral slide b6, the longitudinal slide b7, the longitudinal slide a10 and the lateral slide a11 in one-to-one correspondence;

The host computer 9 is used to quickly match the three-dimensional shape of the image of the rigid body 4 to be calibrated, calculate the coordinates of the key points to be calibrated in the visual measurement space, and use the pictures of different poses of the rigid body to obtain the coordinates of multiple sets of key points in the visual measurement space. The coordinates of a certain point of the rigid body in the known external coordinate system with multiple groups of rigid bodies, and then obtain the corresponding relationship between the visual measurement space and the external known coordinate system, so as to calculate the key points associated with the rigid body to be calibrated in the external known coordinate system. s position.

Further, the host computer 9 is also used to judge the ratio of the pixel area of the rigid body image to be calibrated to the entire width, and when the ratio is not within the ratio threshold range, send control signals to one or more of the four motors to control the longitudinal direction. One or more of the sliding table a10, the vertical sliding table b7, the horizontal sliding table b6 or the vertical sliding table b7 are moved until the ratio of the pixel area of the rigid body image to be calibrated to the entire frame is at the proportional threshold interval position.

Further, the threshold range of the ratio of the pixel area of the rigid body image to be calibrated to the entire width is 30%-50%.

Further, the control signal input ends of the four motors are respectively connected with the host computer 9 through data lines;

The host computer 9 is also used to judge whether the rigid body 4 to be calibrated is complete in the image, and when the rigid body 4 to be calibrated is incomplete in the image, it sends control signals to one or more of the four motors to control the longitudinal slides a10, One or more of the longitudinal slide b7, the lateral slide b6 or the longitudinal slide b7 are moved until a complete image of the rigid body to be calibrated is obtained.

The host computer of the present invention includes an image processing module, the image processing module performs image processing on the received image of the rigid body to be calibrated, and firstly adjusts the positions of the two cameras flexibly according to the size of the captured image and the rigid body to be calibrated to adapt to different shapes When judging whether the collected rigid body to be calibrated is complete, it is only necessary to compare the images. The principle is shown in Figure 2 to Figure 4. To simplify the analysis, it is assumed that the two cameras are placed horizontally and only pixels are considered. The localization problem of spatial point at the level of 1. The size of the spatial point corresponding to a single pixel is composed of the intersection of rays at the boundary of a single pixel emitted at the optical center of the two cameras. Only a single row of pixels (similar to multiple rows of pixels) is analyzed. The intersecting part is a quadrilateral, and the size of the quadrilateral is the resolution. Since the resolution of each point in the measurement space is different, the unit is related to the resolution of the pixel at the optical center. When the resolution is adjusted, the camera a moves to the right, and the measurement The overall space shrinks and moves toward the direction of camera b, the resolution is improved, and the measurement space becomes narrower and longer, which is suitable for rigid bodies with narrow and long shapes; the two cameras are approached along the optical center of the other camera at the same time, and the overall measurement space is greatly reduced. To further improve the resolution, it is suitable for occasions where the rigid body to be measured is small but the precision is high.

Obviously, the larger rigid body to be calibrated has a lower resolution. By analyzing the image, calculate the pixel area ratio of the camera shooting format occupied by the rigid body to be calibrated in each posture, and ensure that it is below 50% and more than 30% to ensure that The rigid body to be calibrated has enough movement space and a certain measurement accuracy. When it is higher than 50%, the system controls the motor to stay away from the rigid body to be calibrated. When it is lower than 30%, the system controls the motor to approach the rigid body to be calibrated, and then conducts the internal and external parameters of the system. Calibration, and finally use the stereo vision system composed of two cameras to quickly match the three-dimensional shape, and calculate the coordinates of the key points to be calibrated in the stereo vision system, and perform a certain movement through the rigid body to obtain multiple sets of key points in the stereo. The corresponding relationship between the coordinates of the vision system and the pose of a certain point on the rigid body in the external known coordinate system, so as to calculate the pose of the key point coordinate system associated with the rigid body to be calibrated in the external known coordinate system.

During use, first put the rigid body 4 to be calibrated into the visual measurement space (that is, the space where the shooting spaces of the two cameras meet), and adjust the horizontal slide table b6, the vertical slide table b7, the vertical slide table a10, and the horizontal slide table a11 respectively. The calibration rigid body 4 is located at a suitable position in the visual measurement space, and the rigid body 4 to be calibrated is externally controlled to move to all poses required for calibration in the visual measurement space, and the camera a1 and camera b5 are used to take pictures and transmit the data to the host computer 9 , the images of all the poses are processed through the image processing algorithm, and the poses of the key point coordinate system associated with the rigid body 4 to be calibrated in the external known coordinate system are calculated.

Although the invention has been described herein with reference to specific embodiments, it should be understood that these embodiments are merely illustrative of the principles and applications of the invention. It should therefore be understood that many modifications may be made to the exemplary embodiments and other arrangements can be devised without departing from the spirit and scope of the invention as defined by the appended claims. It should be understood that the features described in the various dependent claims and herein may be combined in different ways than are described in the original claims. It will also be appreciated that features described in connection with a single embodiment may be used in other described embodiments.

Claims (4)

1. A convergent binocular system for spatial transformation and calibration of a rigid body coordinate system, characterized in that the system comprises a camera a (1), two plane light sources (2), two light source fixing plates (3), a cross Adjustment table a, cross adjustment table b, camera b (5), base (8) and upper computer (9); The cross adjustment table a includes a longitudinal sliding table a (10) and a transverse sliding table a (11); the transverse sliding table a (11) is arranged on the upper side of the longitudinal sliding table a (10), and the transverse sliding table a (11) The sliding direction of is perpendicular to the sliding direction of the longitudinal sliding table a (10); The cross adjustment table b includes a transverse sliding table b(6) and a longitudinal sliding table b(7); the transverse sliding table b(6) is arranged on the longitudinal sliding table b(7), and the sliding table b(6) slides The direction is perpendicular to the sliding direction of the longitudinal sliding table b(7); The cross adjustment table a, the cross adjustment table b and the two light source fixing plates (3) are all fixed on the upper surface of the base (8) to form a visual measurement space; the visual measurement space is used for placing the rigid body to be calibrated; The sliding direction of the longitudinal sliding table a (10) is perpendicular to the sliding direction of the longitudinal sliding table b (7); The sliding direction of the lateral slide a (11) is perpendicular to the sliding direction of the lateral slide b (6); Two light source fixing plates (3) are arranged adjacently, and the two light source fixing plates (3) are both fixed with a plane light source (2); The sliding directions of the longitudinal sliding table a (10) and the longitudinal sliding table b (7) are respectively perpendicular to the two light source fixing plates (3); A camera a (1) is fixed on the lateral slide a (11), and the shooting surface of the camera a (1) is opposite to a plane light source; A camera b(5) is fixed on the lateral sliding table b(6), and the shooting surface of the camera b(5) is opposite to another plane light source; the camera b(5) and the camera a(1) are used for visual measurement of the space The rigid body to be calibrated is used for image acquisition; The camera b (5) and the camera a (1) respectively upload the collected images of the rigid body to be calibrated to the upper computer (9); Four motors are used to control the transverse slide table b (6), the longitudinal slide table b (7), the longitudinal slide table a (10) and the transverse slide table a (11) in one-to-one correspondence; The upper computer (9) is used to quickly match the three-dimensional shape of the image of the rigid body to be calibrated (4), calculate the coordinates of the key points to be calibrated in the visual measurement space, and use the pictures of different poses of the rigid body to obtain multiple sets of key points in the visual field. The coordinates of the measurement space and the coordinates of a certain point of multiple sets of rigid bodies in the external known coordinate system, and then the corresponding relationship between the visual measurement space and the external known coordinate system is obtained, and the key points associated with the rigid body to be calibrated are calculated. position in the coordinate system. 2. a kind of convergent binocular system for rigid body coordinate system space transformation calibration according to claim 1, is characterized in that, host computer (9) is also used for judging the ratio that the pixel area of rigid body image to be calibrated accounts for the whole width , when the ratio is not within the ratio threshold range, send control signals to one or more of the four motors to control the longitudinal slide a (10), the longitudinal slide b (7), and the lateral slide b (6) Or one or more of the vertical slides b(7) are moved until the ratio of the pixel area of the rigid body image to be calibrated to the entire frame is within the proportional threshold range. 3 . The convergent binocular system for spatial transformation and calibration of rigid body coordinate system according to claim 2 , wherein the pixel area of the rigid body image to be calibrated accounts for the proportion of the entire width of the image, and the threshold range is 30%-50%. 4 . 4. a kind of convergent binocular system for rigid body coordinate system space transformation calibration according to claim 1, is characterized in that, the control signal input end of four motors is respectively connected with host computer (9) by data line; The host computer (9) is also used for judging whether the rigid body (4) to be calibrated is complete in the image, and when the rigid body (4) to be calibrated is incomplete in the image, it sends a control signal to one or more of the four motors, respectively, One or more of the longitudinal slide a (10), the longitudinal slide b (7), the lateral slide b (6) or the longitudinal slide b (7) are controlled to move until a complete image of the rigid body to be calibrated is obtained.

Images