Robot visual posture guide calibration method and system based on contour features

Abstract

The invention relates to the technical field of robot vision, and discloses a robot vision posture guide calibration method and system based on contour features, and the method comprises the steps: extracting all contour data collected by a camera within a preset time length before a current moment, dividing the preset time length into a plurality of time intervals, and enabling the preset time length to comprise a plurality of preset moments; calculating a contour observation index of the robot palm based on the dispersion degree of the contour data in each time interval on all pixel points on the image plane; posture data of the tail end of the palm of the robot are collected, the joint change metric value of the posture of the palm of the robot is calculated, and the palm is guided to move towards the better observation posture. The contour features of the robot are associated with the motion state, the method is not limited to static hand-eye calibration, the hand-eye calibration precision is improved by dynamically obtaining calibration data with better quality, and manual intervention is reduced.

Description

Technical Field

[0001] This invention relates to the field of robot vision technology, and more specifically, to a robot vision posture guidance and calibration method and system based on contour features. Background Technology

[0002] With the rapid development of industrial automation and intelligent manufacturing, robots are increasingly widely used in precision assembly, workpiece gripping, and 3D inspection. As a key component in perceiving the environment, the calibration accuracy of robot vision systems directly affects the robot's positioning and operational performance. Hand-eye calibration is a core issue in robot vision systems, aiming to determine the transformation relationship between the camera coordinate system and the robot coordinate system, which is a prerequisite for vision-guided robot movement.

[0003] Traditional hand-eye calibration methods typically rely on high-precision calibration boards or specific markers, collecting multiple sets of pose data through manual intervention or fixed paths for calculation. While these methods are effective in static or structured environments, in dynamic, unstructured environments, the collected data often contains noise and anomalies due to factors such as robot motion jitter, changes in ambient lighting, and occlusion, leading to decreased calibration accuracy or even failure. Furthermore, traditional methods fail to fully integrate the robot's own motion state, resulting in the following problems: first, they fail to effectively fuse the spatiotemporal distribution characteristics of contour features in the image plane; second, they lack refined analysis of abrupt changes and stable phases of joint motion; and third, they fail to establish a quantitative correlation mechanism between visual contour stability and the robot's motion state. Summary of the Invention

[0004] This invention provides a robot visual posture guidance calibration method and system based on contour features, which associates robot contour features with motion state, no longer limited to static hand-eye calibration, improves hand-eye calibration accuracy, reduces human intervention, and enhances robot motion stability and operational accuracy.

[0005] To achieve the above objectives, this invention provides a robot visual pose guidance calibration method based on contour features, comprising: Extract all contour data collected by the camera within a preset time period before the current moment, and divide the preset time period into multiple time intervals, wherein the preset time period includes multiple preset moments; Based on the degree of dispersion of the contour data of all pixels on the image plane within each time interval, the contour observation index of the robot's hand at the current moment is calculated. Collect angle data of the joints at the end of the robot's hand at each preset time, sieve all the angle data to obtain standard angle data, and calculate the joint change metric value of the robot's hand posture based on the standard angle data; The robot's posture deviation coefficient is obtained by weighted summation of the contour observation index and the joint change metric, and the robot's posture guidance hand-eye calibration strategy is set based on the posture deviation coefficient.

[0006] Furthermore, when calculating the contour observation index of the robot's hand at the current moment based on the dispersion of all pixels of the contour data on the image plane within each time interval, the following are included: Determine the pixel points of the contour data in the u and v directions of the image plane within each time interval; Calculate the uv direction variance of the pixel corresponding to each time interval, wherein the uv direction variance is calculated by calculating the first variance of all u directions, calculating the second variance of all v directions, and taking the sum of the first variance and the second variance as the uv direction variance; A straight line is fitted to all the variances in the UV direction to obtain the fitted straight line for the UV direction variance. The mean slope of the fitted line corresponding to the variance of the uv direction is determined and used as the contour observation index of the robot hand.

[0007] Furthermore, when filtering all the angle data to obtain standard angle data, the following steps are taken: Determine an angle data point, and extract the left and right angle data corresponding to the angle data according to a preset time. Determine the absolute value of the first difference between the left-side angle data and the right-side angle data, and use it as the angle span data; Determine the mean value of the angle data corresponding to all angle data, and calculate the absolute value of the second difference between the angle data and the mean value of the angle data as the angle skew data; Determine the left preset time interval corresponding to the angle data, and determine the right preset time interval corresponding to the angle data; Calculate the standard angle data coefficient corresponding to the angle data based on the angle span data, angle deviation data, left preset time interval and right preset time interval; When the standard angle data coefficient is less than the preset standard angle data coefficient, the corresponding angle data is taken as non-standard angle data. When the standard angle data coefficient is greater than or equal to the preset standard angle data coefficient, the corresponding angle data is used as the standard angle data.

[0008] Furthermore, when calculating the standard angle data coefficient corresponding to the angle data based on the angle span data, angle deviation data, preset time interval on the left, and preset time interval on the right, the calculation includes: The standard angle data coefficients are calculated using the following formula: ; Where m is the standard angle data coefficient, n1 is the angle span data, n2 is the angle deviation data, b1 is the preset time interval on the left, b2 is the preset time interval on the right, d1 is the first coefficient, d2 is the second coefficient, d3 is the third coefficient, d1>d2>d3, d1+d2+d3=1.

[0009] Furthermore, when calculating the joint change metric of the robot hand posture based on the standard angle data, the following steps are included: The standard angle data is analyzed in stages to obtain the steady angle variation coefficient and the fluctuating angle variation coefficient; The weighted sum of the steady angle change coefficient and the fluctuating angle change coefficient is determined as the joint change metric of the robot hand posture.

[0010] Furthermore, when performing phased analysis on the standard angle data to obtain the steady-state angle variation coefficient, the process includes: Extract the same standard angle data from the standard angle data to obtain multiple standard angle data sets; Count the number of the first standard angle data set in the aforementioned standard angle data set; Extract one standard angle from each of the standard angle datasets and calculate the first standard angle data and value; Determine the mean of all standard angle data, remove all standard angle data sets that are less than the mean of the standard angle data, and count the number of second standard angle data sets in the remaining standard angle data sets. Extract one standard angle data point from the remaining standard angle data set, and calculate the second standard angle data and value; The steady angle variation coefficient is calculated based on the number of data sets in the first standard angle set, the number of data sets in the second standard angle set, the sum of the first standard angle data and the sum of the second standard angle data.

[0011] Further, when calculating the steady-state angle variation coefficient based on the number of data sets in the first standard angle dataset, the number of data sets in the second standard angle dataset, the sum of the first standard angle data, and the sum of the second standard angle data, the calculation includes: The coefficient of variation of the steady angle is calculated according to the following formula: ; Where v is the stationary angle variation coefficient, c1 is the number of data sets of the first standard angle, c2 is the number of data sets of the second standard angle, k1 is the sum of the data of the first standard angle, and k2 is the sum of the data of the second standard angle.

[0012] Furthermore, when performing phased analysis on the standard angle data to obtain the fluctuation angle variation coefficient, the process includes: Extract the remaining distinct standard angle data and sort the distinct standard angle data in ascending order; The fluctuation angle variation coefficient is calculated using the following formula: ; Where j is the fluctuation angle variation coefficient, h is the number of different standard angle data, and f g f is the number of distinct standard angle data for the g-th time. g+1 This represents the number of distinct standard angle data points for the (g+1)th time.

[0013] Furthermore, when setting the robot's posture guidance hand-eye calibration strategy based on the posture deviation coefficient, the following steps are included: Multiple preset attitude deviation coefficients are set in advance; Multiple preset rotation angles can be set in advance; Based on the relationship between the attitude deviation coefficient and multiple preset attitude deviation coefficients, a corresponding preset rotation angle is selected, and the end effector of the robot is controlled to rotate in the target direction based on the selected preset rotation angle.

[0014] To achieve the above objectives, the present invention also provides a robot visual pose guidance calibration system based on contour features, comprising: The data extraction module is used to extract all contour data collected by the camera within a preset time period before the current moment, and divide the preset time period into multiple time intervals, wherein the preset time period includes multiple preset moments; The first calculation module is used to calculate the contour observation index of the robot's hand at the current moment based on the degree of dispersion of the contour data of all pixels on the image plane within each time interval. The second calculation module is used to collect the angle data of the joints at the end of the robot's hand at each preset time, sieve all the angle data to obtain standard angle data, and calculate the joint change metric value of the robot's hand posture based on the standard angle data. The hand-eye calibration module is used to perform a weighted summation of the contour observation index and the joint change metric to obtain the robot's posture deviation coefficient, and to set the robot's posture guidance hand-eye calibration strategy based on the posture deviation coefficient.

[0015] Compared with the prior art, the beneficial effects of the present invention are as follows: This invention discloses a robot visual posture guidance calibration method and system based on contour features. The method includes: extracting all contour data collected by the camera within a preset time period before the current moment, dividing the preset time period into multiple time intervals, each including multiple preset moments; calculating the contour observation index of the robot hand based on the dispersion of all pixels of the contour data on the image plane within each time interval; collecting the angle data of the robot joints to obtain standard angle data, and calculating the joint change metric of the robot; weighted summing the contour observation index and the joint change metric to obtain the robot's posture deviation coefficient, and setting a posture guidance hand-eye calibration strategy to associate the robot's contour features with its motion state, no longer limited to static hand-eye calibration, improving hand-eye calibration accuracy, reducing human intervention, and enhancing the robot's motion stability and operational accuracy. Attached Figure Description

[0016] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings: Figure 1 A flowchart illustrating the robot visual pose guidance calibration method based on contour features in an embodiment of the present invention is shown. Figure 2 A schematic diagram of the structure of a robot vision posture guidance and calibration system based on contour features in an embodiment of the present invention is shown. Detailed Implementation

[0017] The specific embodiments of the present invention will be described in further detail below with reference to the accompanying drawings and examples. The following examples are for illustrative purposes only and are not intended to limit the scope of the invention.

[0018] In the description of this application, it should be understood that the terms "center", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this application.

[0019] The terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this application, unless otherwise stated, "a plurality of" means two or more.

[0020] In the description of this application, it should be noted that, unless otherwise expressly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection between two components. Those skilled in the art can understand the specific meaning of the above terms in this application based on the specific circumstances.

[0021] The following is a description of preferred embodiments of the present invention in conjunction with the accompanying drawings.

[0022] like Figure 1 As shown, embodiments of the present invention disclose a robot visual pose guidance calibration method based on contour features, including: S110: Extract all contour data collected by the camera within a preset time period before the current moment, and divide the preset time period into multiple time intervals, wherein the preset time period includes multiple preset moments; In this embodiment, the preset duration is preferably the past two minutes, and the preset time is preferably 120, including the 1st second, the 2nd second, the 3rd second, ..., the 120th second.

[0023] In this embodiment, every 10 preset moments are defined as a time interval.

[0024] In this embodiment, the contour data is pixel coordinate data.

[0025] S120: Calculate the contour observation index of the robot's hand at the current moment based on the dispersion of all pixels of the contour data on the image plane within each time interval; In some embodiments of this application, when calculating the contour observation index of the robot hand at the current moment based on the dispersion of contour data across all pixels on the image plane within each time interval, the following steps are included: Determine the pixel points of the contour data in the u and v directions of the image plane within each time interval; Calculate the uv direction variance of the pixel corresponding to each time interval, wherein the uv direction variance is calculated by calculating the first variance of all u directions, calculating the second variance of all v directions, and taking the sum of the first variance and the second variance as the uv direction variance; A straight line is fitted to all the variances in the UV direction to obtain the fitted straight line for the UV direction variance. The mean slope of the fitted line corresponding to the variance of the uv direction is determined and used as the contour observation index of the robot hand.

[0026] In this embodiment, the u direction is the horizontal direction of the image, the v direction is the vertical direction of the image, and the origin is at the upper left corner of the two-dimensional image.

[0027] In this embodiment, the numerical variance in the u direction of each pixel is calculated, and the numerical variance in the v direction of each pixel is calculated.

[0028] The beneficial effects of the above technical solution are: by calculating the variance in the u and v directions, the dispersion of contour data on the image plane can be accurately quantified, thereby obtaining the contour observation index of the robot hand, providing a reliable basis for the subsequent calculation of the posture deviation coefficient, and effectively integrating the spatiotemporal distribution characteristics of contour features in the image plane.

[0029] S130: Collect angle data of the joints at the end of the robot's hand at each preset time, sieve all the angle data to obtain standard angle data, and calculate the joint change metric value of the robot's hand posture based on the standard angle data. In some embodiments of this application, the process of filtering all angle data to obtain standard angle data includes: Determine an angle data point, and extract the left and right angle data corresponding to the angle data according to a preset time. Determine the absolute value of the first difference between the left-side angle data and the right-side angle data, and use it as the angle span data; Determine the mean value of the angle data corresponding to all angle data, and calculate the absolute value of the second difference between the angle data and the mean value of the angle data as the angle skew data; Determine the left preset time interval corresponding to the angle data, and determine the right preset time interval corresponding to the angle data; Calculate the standard angle data coefficient corresponding to the angle data based on the angle span data, angle deviation data, left preset time interval and right preset time interval; When the standard angle data coefficient is less than the preset standard angle data coefficient, the corresponding angle data is taken as non-standard angle data. When the standard angle data coefficient is greater than or equal to the preset standard angle data coefficient, the corresponding angle data is used as the standard angle data.

[0030] In this embodiment, the left angle data and right angle data are determined based on a preset time. As mentioned above, if a preset time is selected as the 5th second, then the left angle data is the angle data corresponding to the 4th second, and the right angle data is the angle data corresponding to the 6th second. It should be noted that, in order to avoid errors, the standard angle data coefficients corresponding to the 1st second and the 120th second are not calculated here.

[0031] In this embodiment, the preset time interval on the left refers to the time interval between the preset time corresponding to the angle data and the 1st second, and the preset time interval on the right refers to the time interval between the preset time corresponding to the angle data and the 120th second.

[0032] In this embodiment, the preset standard angle data coefficient is preferably 12, but it can be adjusted adaptively according to actual needs.

[0033] In this embodiment, by repeating the above steps, the standard angle data coefficient corresponding to each angle data can be calculated, except for the angle data corresponding to the 1st second and the 120th second.

[0034] The beneficial effects of the above technical solution are: by screening the angle data, abnormal data can be effectively removed to obtain standard angle data, thereby improving the accuracy of joint change measurement calculation and providing more reliable data support for robot posture guidance.

[0035] In some embodiments of this application, when calculating the standard angle data coefficient corresponding to the angle data based on the angle span data, angle deviation data, left preset time interval, and right preset time interval, the calculation includes: The standard angle data coefficients are calculated using the following formula: ; Where m is the standard angle data coefficient, n1 is the angle span data, n2 is the angle deviation data, b1 is the preset time interval on the left, b2 is the preset time interval on the right, d1 is the first coefficient, d2 is the second coefficient, d3 is the third coefficient, d1>d2>d3, d1+d2+d3=1.

[0036] In some embodiments of this application, calculating the joint change metric of the robot hand posture based on the standard angle data includes: The standard angle data is analyzed in stages to obtain the steady angle variation coefficient and the fluctuating angle variation coefficient; The weighted sum of the steady angle change coefficient and the fluctuating angle change coefficient is determined as the joint change metric of the robot hand posture.

[0037] In this embodiment, the steady angle change coefficient and the fluctuating angle change coefficient are weighted based on subjective weighting method and objective weighting method. Here, the weight of the steady angle change coefficient is preferably 0.65, and the weight of the fluctuating angle change coefficient is preferably 0.35. The specific weights can be adjusted adaptively according to the actual situation.

[0038] The beneficial effects of the above technical solution are: the present invention determines the weighted sum of the steady angle change coefficient and the fluctuating angle change coefficient as the joint change metric of the robot, realizes the refined analysis of the sudden change and steady phase of joint motion, and lays the foundation for the posture guidance of the robot.

[0039] In some embodiments of this application, when performing phased analysis on the standard angle data to obtain the steady-state angle variation coefficient, the following steps are included: Extract the same standard angle data from the standard angle data to obtain multiple standard angle data sets; Count the number of the first standard angle data set in the aforementioned standard angle data set; Extract one standard angle from each of the standard angle datasets and calculate the first standard angle data and value; Determine the mean of all standard angle data, remove all standard angle data sets that are less than the mean of the standard angle data, and count the number of second standard angle data sets in the remaining standard angle data sets. Extract one standard angle data point from the remaining standard angle data set, and calculate the second standard angle data and value; The steady angle variation coefficient is calculated based on the number of data sets in the first standard angle set, the number of data sets in the second standard angle set, the sum of the first standard angle data and the sum of the second standard angle data.

[0040] In this embodiment, the standard angle data in each standard angle data set is the same, but the standard angle data between each standard angle data set is different. That is, the same standard angle data is combined to obtain a standard angle data set. It should be noted that the number of standard angle data in a standard angle data set is at least two.

[0041] The beneficial effects of the above technical solution are: by calculating the steady angle change coefficient based on the number of data sets of the first standard angle, the number of data sets of the second standard angle, the sum of the first standard angle data and the sum of the second standard angle data, it is possible to accurately measure the characteristics of the robot joint in the steady motion phase, provide key support for the accurate calculation of joint change measurement values, and help to guide the robot's posture adjustment more accurately.

[0042] In some embodiments of this application, the calculation of the steady-state angle variation coefficient based on the number of the first standard angle data set, the number of the second standard angle data set, the sum of the first standard angle data and the sum of the second standard angle data includes: The coefficient of variation of the steady angle is calculated according to the following formula: ; Where v is the stationary angle variation coefficient, c1 is the number of data sets of the first standard angle, c2 is the number of data sets of the second standard angle, k1 is the sum of the data of the first standard angle, and k2 is the sum of the data of the second standard angle.

[0043] In some embodiments of this application, when performing phased analysis on the standard angle data to obtain the fluctuation angle variation coefficient, the process includes: Extract the remaining distinct standard angle data and sort the distinct standard angle data in ascending order; The fluctuation angle variation coefficient is calculated using the following formula: ; Where j is the fluctuation angle variation coefficient, h is the number of different standard angle data, and f g f is the number of distinct standard angle data for the g-th time. g+1 This represents the number of distinct standard angle data points for the (g+1)th time.

[0044] The beneficial effects of the above technical solution are as follows: The fluctuation angle change coefficient obtained through the above calculation method can accurately capture the characteristic changes of robot joints during the fluctuation motion phase. This is crucial for a comprehensive and accurate analysis of the robot joint motion state, further ensuring the accuracy of joint change metric calculations, thus providing a more precise basis for robot posture guidance. Simultaneously, this calculation method has strong adaptability and stability, effectively addressing the complexity and diversity of robot joint movements in different scenarios.

[0045] S140: The contour observation index and the joint change metric are weighted and summed to obtain the robot's posture deviation coefficient, and the robot's posture guidance hand-eye calibration strategy is set according to the posture deviation coefficient.

[0046] In this embodiment, the contour observation index and joint change measurement value are weighted based on subjective weighting method and objective weighting method. Here, the weight of the contour observation index is preferably 0.6 and the weight of the joint change measurement value is preferably 0.4. The specific weights can be adjusted adaptively according to actual needs.

[0047] The beneficial effects of the above technical solution are: the present invention can establish a quantitative correlation mechanism between visual contour stability and robot motion state, further ensuring the accuracy of robot hand-eye calibration.

[0048] In some embodiments of this application, setting the robot's posture-guided hand-eye calibration strategy based on the posture deviation coefficient includes: Multiple preset attitude deviation coefficients are set in advance; Multiple preset rotation angles can be set in advance; Based on the relationship between the attitude deviation coefficient and multiple preset attitude deviation coefficients, a corresponding preset rotation angle is selected, and the end effector of the robot is controlled to rotate in the target direction based on the selected preset rotation angle.

[0049] In this embodiment, the number of preset attitude deviation coefficients is preferably 3, including a first preset attitude deviation coefficient, preferably 3, a second preset attitude deviation coefficient, preferably 5, and a third preset attitude deviation coefficient, preferably 7. The specific values ​​can be adjusted according to actual needs.

[0050] In this embodiment, the number of preset turning angles is preferably 4, including a first preset turning angle, preferably 2°, a second preset turning angle, preferably 4°, a third preset turning angle, preferably 6°, and a fourth preset turning angle, preferably 8°. The specific values ​​can be adjusted according to actual needs.

[0051] In this embodiment, when the attitude deviation coefficient is less than the first preset attitude deviation coefficient, the first preset rotation angle is selected; when the attitude deviation coefficient is greater than or equal to the first preset attitude deviation coefficient and less than the second preset attitude deviation coefficient, the second preset rotation angle is selected; when the attitude deviation coefficient is greater than or equal to the second preset attitude deviation coefficient and less than the third preset attitude deviation coefficient, the third preset rotation angle is selected; and when the attitude deviation coefficient is greater than or equal to the third preset attitude deviation coefficient, the fourth preset rotation angle is selected.

[0052] The beneficial effects of the above technical solution are: the present invention selects the corresponding preset rotation angle based on the relationship between the posture deviation coefficient and the preset posture deviation coefficient, thereby realizing the dynamic hand-eye calibration of the robot, no longer limited to static hand-eye calibration, improving the accuracy of hand-eye calibration, reducing manual intervention, and enhancing the robot's motion stability and operational accuracy.

[0053] To further illustrate the technical concept of this invention, the technical solution of this invention will now be described in conjunction with specific application scenarios.

[0054] Correspondingly, such as Figure 2 As shown, this application also provides a robot visual pose guidance calibration system based on contour features, including: The data extraction module is used to extract all contour data collected by the camera within a preset time period before the current moment, and divide the preset time period into multiple time intervals, wherein the preset time period includes multiple preset moments; The first calculation module is used to calculate the contour observation index of the robot's hand at the current moment based on the degree of dispersion of the contour data of all pixels on the image plane within each time interval. The second calculation module is used to collect the angle data of the joints at the end of the robot's hand at each preset time, sieve all the angle data to obtain standard angle data, and calculate the joint change metric value of the robot's hand posture based on the standard angle data. The hand-eye calibration module is used to perform a weighted summation of the contour observation index and the joint change metric to obtain the robot's posture deviation coefficient, and to set the robot's posture guidance hand-eye calibration strategy based on the posture deviation coefficient.

[0055] In the description of the above embodiments, specific features, structures, materials, or characteristics may be combined in any suitable manner in one or more embodiments or examples.

[0056] Although the invention has been described above with reference to embodiments, various modifications can be made and components can be replaced with equivalents without departing from the scope of the invention. In particular, as long as there is no structural conflict, the features in the embodiments disclosed in this invention can be combined with each other in any way. The fact that not all of these combinations are described in this specification is merely for the sake of brevity and resource conservation.

[0057] It will be understood by those skilled in the art that the above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A robot visual pose guidance calibration method based on contour features, characterized in that, include: Extract all contour data collected by the camera within a preset time period before the current moment, and divide the preset time period into multiple time intervals, wherein the preset time period includes multiple preset moments; Based on the degree of dispersion of the contour data of all pixels on the image plane within each time interval, the contour observation index of the robot's hand at the current moment is calculated. Collect angle data of the joints at the end of the robot's hand at each preset time, sieve all the angle data to obtain standard angle data, and calculate the joint change metric value of the robot's hand posture based on the standard angle data; The robot's posture deviation coefficient is obtained by weighted summation of the contour observation index and the joint change metric, and the robot's posture guidance hand-eye calibration strategy is set based on the posture deviation coefficient.

2. The robot visual pose guidance calibration method based on contour features according to claim 1, characterized in that, When calculating the contour observation index of the robot's hand at the current moment based on the dispersion of all pixels of the contour data on the image plane within each time interval, the following are included: Determine the pixel points of the contour data in the u and v directions of the image plane within each time interval; Calculate the uv direction variance of the pixel corresponding to each time interval, wherein the uv direction variance is calculated by calculating the first variance of all u directions, calculating the second variance of all v directions, and taking the sum of the first variance and the second variance as the uv direction variance; A straight line is fitted to all the variances in the UV direction to obtain the fitted straight line for the UV direction variance. The mean slope of the fitted line corresponding to the variance of the uv direction is determined and used as the contour observation index of the robot hand.

3. The robot visual pose guidance calibration method based on contour features according to claim 1, characterized in that, When filtering all angle data to obtain standard angle data, the following steps are taken: Determine an angle data point, and extract the left and right angle data corresponding to the angle data according to a preset time. Determine the absolute value of the first difference between the left-side angle data and the right-side angle data, and use it as the angle span data; Determine the mean value of the angle data corresponding to all angle data, and calculate the absolute value of the second difference between the angle data and the mean value of the angle data as the angle skew data; Determine the left preset time interval corresponding to the angle data, and determine the right preset time interval corresponding to the angle data; Calculate the standard angle data coefficient corresponding to the angle data based on the angle span data, angle deviation data, left preset time interval and right preset time interval; When the standard angle data coefficient is less than the preset standard angle data coefficient, the corresponding angle data is taken as non-standard angle data. When the standard angle data coefficient is greater than or equal to the preset standard angle data coefficient, the corresponding angle data is used as the standard angle data.

4. The robot visual pose guidance calibration method based on contour features according to claim 3, characterized in that, When calculating the standard angle data coefficient corresponding to the angle data based on the angle span data, angle deviation data, preset time interval on the left, and preset time interval on the right, the calculation includes: The standard angle data coefficients are calculated using the following formula: ； Where m is the standard angle data coefficient, n1 is the angle span data, n2 is the angle deviation data, b1 is the preset time interval on the left, b2 is the preset time interval on the right, d1 is the first coefficient, d2 is the second coefficient, d3 is the third coefficient, d1>d2>d3, d1+d2+d3=1.

5. The robot visual pose guidance calibration method based on contour features according to claim 1, characterized in that, When calculating the joint change metric of the robot hand posture based on the standard angle data, the following is included: The standard angle data is analyzed in stages to obtain the steady angle variation coefficient and the fluctuating angle variation coefficient; The weighted sum of the steady angle change coefficient and the fluctuating angle change coefficient is determined as the joint change metric of the robot hand posture.

6. The robot visual pose guidance calibration method based on contour features according to claim 5, characterized in that, When performing phased analysis on the standard angle data to obtain the steady-state angle variation coefficient, the following steps are included: Extract the same standard angle data from the standard angle data to obtain multiple standard angle data sets; Count the number of the first standard angle data set in the aforementioned standard angle data set; Extract one standard angle from each of the standard angle datasets and calculate the first standard angle data and value; Determine the mean of all standard angle data, remove all standard angle data sets that are less than the mean of the standard angle data, and count the number of second standard angle data sets in the remaining standard angle data sets. Extract one standard angle data point from the remaining standard angle data set, and calculate the second standard angle data and value; The steady angle variation coefficient is calculated based on the number of data sets in the first standard angle set, the number of data sets in the second standard angle set, the sum of the first standard angle data and the sum of the second standard angle data.

7. The robot visual pose guidance calibration method based on contour features according to claim 6, characterized in that, When calculating the steady-state angle variation coefficient based on the number of data sets in the first standard angle dataset, the number of data sets in the second standard angle dataset, the sum of the first standard angle data and the sum of the second standard angle data, the following steps are included: The coefficient of variation of the steady angle is calculated according to the following formula: ； Where v is the stationary angle variation coefficient, c1 is the number of data sets of the first standard angle, c2 is the number of data sets of the second standard angle, k1 is the sum of the data of the first standard angle, and k2 is the sum of the data of the second standard angle.

8. The robot visual pose guidance calibration method based on contour features according to claim 5, characterized in that, When performing phased analysis on the standard angle data to obtain the fluctuation angle variation coefficient, the following steps are included: Extract the remaining distinct standard angle data and sort the distinct standard angle data in ascending order; The fluctuation angle variation coefficient is calculated using the following formula: ； Where j is the fluctuation angle variation coefficient, h is the number of different standard angle data, and f g f is the number of distinct standard angle data for the g-th time. g+1 This represents the number of distinct standard angle data points for the (g+1)th time.

9. The robot visual pose guidance calibration method based on contour features according to claim 1, characterized in that, When setting the robot's posture guidance hand-eye calibration strategy based on the posture deviation coefficient, the following are included: Multiple preset attitude deviation coefficients are set in advance; Multiple preset rotation angles can be set in advance; Based on the relationship between the attitude deviation coefficient and multiple preset attitude deviation coefficients, a corresponding preset rotation angle is selected, and the end effector of the robot is controlled to rotate in the target direction based on the selected preset rotation angle.

10. A robot visual posture guidance and calibration system based on contour features, applied to the robot visual posture guidance and calibration method based on contour features as described in any one of claims 1-9, characterized in that, include: The data extraction module is used to extract all contour data collected by the camera within a preset time period before the current moment, and divide the preset time period into multiple time intervals, wherein the preset time period includes multiple preset moments; The first calculation module is used to calculate the contour observation index of the robot's hand at the current moment based on the degree of dispersion of the contour data of all pixels on the image plane within each time interval. The second calculation module is used to collect the angle data of the joints at the end of the robot's hand at each preset time, sieve all the angle data to obtain standard angle data, and calculate the joint change metric value of the robot's hand posture based on the standard angle data. The hand-eye calibration module is used to perform a weighted summation of the contour observation index and the joint change metric to obtain the robot's posture deviation coefficient, and to set the robot's posture guidance hand-eye calibration strategy based on the posture deviation coefficient.

Images