Tennis ball picking robot and positioning recognition ball picking method thereof

Abstract

The invention relates to the technical field of sports equipment, and discloses a tennis ball picking robot and a positioning and recognizing ball picking method.On one hand, the tennis ball picking robot comprises a machine base, a moving mechanism, a recognizing and positioning mechanism and a ball picking mechanism, the moving mechanism is connected to the bottom of the machine base, and the recognizing and positioning mechanism is connected to the machine base; the ball picking mechanism is connected to the front side of the machine base, the ball picking mechanism comprises a mounting base, two sets of ball inlet guide frames arranged on the outer side of the mounting base, a ball picking module arranged on the mounting base and a ball storage box arranged on the rear side of the ball picking module, and a ball inlet is formed in the middle of the lower end of the mounting base; a ball guide chute communicated with the ball inlet is formed in the surface of the mounting seat; and the two groups of ball inlet guide frames are closed up from the front side of the ball inlet to the center to form a ball guide channel.

Description

Technical Field

[0001] The invention relates to the technical field of sports equipment, and in particular to a tennis ball picking robot and a positioning, identifying and picking ball method thereof. Background Art

[0002] In the past decade, tennis has shown a growing trend worldwide, and more and more people are participating in tennis. In sports, the problem of collecting balls is inevitably involved. Existing tennis ball picking equipment is divided into two types: one, pure manual ball picking equipment, and two, intelligent ball picking equipment. The current status of the industry and field is that various types of manual ball picking equipment emerge in an endless stream, and intelligent ball picking equipment needs to be developed. Both have certain limitations. After consulting the information, it was found that the friction wheel technology used in the tennis ball picking robot can well achieve the task of picking up the ball, and with the recent rise of mechanical vision technology, combined with the existing tennis ball picking technology, the tennis ball picking work is realized. The main advantage of the first device is lightweight, but it only plays a role in assisting the ball picking and does not improve the current situation of manpower picking up the ball. The second intelligent ball picking device improves and meets the need to use machinery to replace manual labor, but the manufacturing cost is relatively expensive, or simply uses electricity to provide energy, replacing manpower but the actual use experience is very poor. In view of this, the technicians in this field designed a tennis ball picking robot and its positioning and identification method for picking up the ball. Summary of the invention

[0003] The technical problem to be solved by the present invention is to overcome the deficiencies of the prior art and provide a tennis ball picking robot and a positioning, identifying and picking ball method thereof.

[0004] The technical solution of the present invention is as follows: on the one hand, the present invention discloses a tennis ball picking robot, comprising a machine base, a moving mechanism, an identification and positioning mechanism and a ball picking mechanism, wherein the moving mechanism is connected to the bottom of the machine base, the identification and positioning mechanism is connected to the machine base, the ball picking mechanism is connected to the front side of the machine base, the ball picking mechanism comprises a mounting seat, two groups of goal guide frames arranged on the outside of the mounting seat, a ball picking module arranged on the mounting seat and a ball storage box arranged on the rear side of the ball picking module, a ball inlet is arranged in the middle of the lower end of the mounting seat, a ball guide inclined groove is arranged on the surface of the mounting seat connected to the ball inlet, and the two groups of goal guide frames close toward the center in front of the ball inlet to form a ball guiding channel.

[0005] It can be seen from the above scheme that the moving mechanism is used to drive the tennis ball picking robot to move, the goal guide frame is used to guide the tennis ball from the ball guide channel to the ball entrance, the identification and positioning mechanism is used to identify and locate the position of the tennis ball, and the ball picking module moves the tennis ball at the ball entrance along the ball guide chute to the ball storage box to pick up the tennis ball.

[0006] The ball picking module includes a driving motor symmetrically arranged on both sides of the ball inlet and a friction wheel arranged on the output end of the driving motor. It can be seen that the driving motor is used to drive the friction wheel to rotate, so as to realize friction extrusion and launching of the tennis ball at the ball inlet.

[0007] The identification and positioning mechanism includes a raised frame, a pan-tilt mounted on the raised frame, a servo, and a camera mounted on the pan-tilt, wherein the raised frame is mounted on the base. It can be seen that the pan-tilt is used to achieve tracking, and the camera is used to achieve identification and positioning.

[0008] The width of the ball guiding inclined groove is greater than the spacing between the two groups of friction wheels. A horizontal extension portion is provided at the end of the mounting seat, and the ball inlet is sunken at the bottom end of the horizontal extension portion.

[0009] The base is also provided with a control module, which includes a main control board and a battery. The control module is electrically connected to the moving mechanism, the identification and positioning mechanism and the ball picking mechanism respectively, and the control module signal is connected to a remote monitoring terminal.

[0010] The goal guide frame includes a connecting portion and a gathering portion integrally formed at the end of the connecting portion, the ball guiding channel is located between the two groups of the gathering portions, and the connecting portion is connected to the bottom of the raised frame. It can be seen that the goal guide frame is gathered to the ball entrance through the gathering portion.

[0011] The goal guide is arranged above the ground, and the distance between the goal guide and the ground is smaller than the diameter of the tennis ball. It can be seen that the goal guide is used to limit the position of the tennis ball and prevent dust and other debris from entering the ball guide channel, thereby improving the accuracy of tennis ball guidance and avoiding friction between the bottom of the goal guide and the field.

[0012] The moving mechanism includes four Mecanum wheels arranged at the bottom of the base. It can be seen that the Mecanum wheels are used to drive the tennis ball picking robot to move omnidirectionally to pick up the ball.

[0013] On the other hand, the present invention discloses a method for positioning, identifying and picking up a ball, the method comprising the following steps:

[0014] S1. Turn on the power, the Mecanum wheels drive the robot to rotate 360 ​​degrees, and the robot moves forward step by step. At this time, the camera performs a 360-degree scan with itself as the origin to search for the target to be picked up;

[0015] S2, determine whether there is a tennis ball within the camera recognition range, if not, return to S1; if yes, enter S3;

[0016] S3. Measure the distance of all targets in the camera's field of view and calculate the target distance using the following formula based on the triangular similarity principle:

[0017]

[0018] but

[0019]

[0020] Where Lm represents the distance between the robot and the target, Bpix represents the pixels occupied by the target in the camera, Apix represents the fixed image pixels, a represents half of the lens angle of view, and K represents a constant;

[0021] And calculate the distance factor based on the target distance:

[0022]

[0023] Where, distance = Lm, d represents the distance factor;

[0024] S4, assigning a category label to each of the tennis balls and obstacles such as sharp objects, and assigning a corresponding initial priority to each target based on the type of the target, wherein the initial priority of the tennis balls is set to a positive value, and the initial priority of objects such as sharp objects and obstacles is set to a negative value;

[0025] S5. Based on the potential threat posed by the target, the risk factor is calculated using the following formula:

[0026] r = risk_factor

[0027] Among them, r represents the risk factor, and risk_factor is a pre-set risk value;

[0028] S6. The robot calculates the final priority based on the initial priority, distance factor, and risk factor:

[0029] p=(initial_p+d)*|r|

[0030] Among them, p represents the final priority, initial_p represents the initial priority, d represents the distance factor, and r represents the risk factor;

[0031] S7, if the final priority value is positive, the robot will enter S8 to pick up the ball according to the maximum value of the final priority; if the final priority value is negative, the robot will avoid the maximum absolute value of the final priority pole;

[0032] S8, Mecanum wheel drives the camera to move so that the tennis ball to be picked up is accurately placed in the center of the camera's field of view;

[0033] S9, the goal guide frame collects the ball to the ball entrance, and the driving motors on both sides of the ball entrance drive the friction wheels to rotate and drive the tennis ball to be launched, and the tennis ball enters the ball storage box along the ball guide chute for collection;

[0034] S10. Determine whether the battery power is sufficient according to a preset battery threshold. If sufficient, continue to pick up the ball; if insufficient, the Mecanum wheel drives the robot to return to charge.

[0035] In S1, the single forward step distance of the Mecanum wheel is 5 to 10 meters. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 It is a schematic diagram of the structure of the present invention;

[0037] Figure 2 is a side view of the present invention;

[0038] Figure 3 It is a partial structural schematic diagram of the present invention;

[0039] Figure 4 It is a flow chart of picking up a ball of the present invention;

[0040] Figure 5 This is a ball picking and collection judgment flow chart of the present invention;

[0041] Figure 6 is a startup control flow chart of the present invention;

[0042] Figure 7 It is the design calculation diagram of the goal guide size. DETAILED DESCRIPTION

[0043] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the accompanying drawings in the embodiments of the present invention.

[0044] like Figures 1 to 7As shown, on the one hand, the present invention discloses a tennis ball picking robot, comprising a machine base 1, a moving mechanism 2, an identification and positioning mechanism and a ball picking mechanism, wherein the moving mechanism 2 is connected to the bottom of the machine base 1, the moving mechanism 2 comprises four Mecanum wheels arranged at the bottom of the machine base 1, the identification and positioning mechanism is connected to the machine base 1, the identification and positioning mechanism comprises a cushion frame 7, a pan / tilt 8 arranged on the cushion frame 7, a servo 12 and a camera 9 arranged on the pan / tilt 8, the cushion frame 7 is arranged on the machine base 1, the ball picking mechanism is connected to the front side of the machine base 1, the ball picking mechanism comprises a mounting seat 3, two sets of goal guide frames 4 arranged outside the mounting seat 3, a ball picking module 5 arranged on the mounting seat 3 and a ball storage box 6 arranged at the rear side of the ball picking module 5, a ball inlet 31 is arranged in the middle of the lower end of the mounting seat 3, and a surface of the mounting seat 3 is connected to the ball inlet 31 and is provided with The ball guide groove 32, two groups of the goal guide frames 4 are closed at the center in front of the ball entrance 31 to form a ball guide channel, the ball picking module 5 includes a driving motor 51 symmetrically arranged on both sides of the ball entrance 31 and a friction wheel 52 arranged on the output end of the driving motor 51, the ball guide groove 32 is connected to the ball entrance 31, the width of the ball guide groove 32 is greater than the distance between the two groups of friction wheels 52, a horizontal extension part 33 is arranged at the end of the mounting seat 3, the ball entrance 31 is sunken at the bottom of the horizontal extension part 33, the goal guide frame 3 is arranged above the ground, the distance between the goal guide frame 3 and the ground is less than the diameter of a tennis ball, the goal guide frame 4 includes a connecting part 41 and a gathering part 42 integrally formed at the end of the connecting part 41, the ball guide channel is located between the two groups of the gathering parts 42, and the connecting part 41 is connected to the bottom of the raised frame 7.

[0045] In one embodiment of the present invention, the camera 9 includes an OpenMV module, with an STM32F767CPU as the core, an integrated OV7725 camera chip, and a core machine vision algorithm is efficiently implemented in C language on a compact hardware module, providing a Python programming interface. The pan / tilt 8 and the servo 12 can achieve 180° rotation up, down, left, and right, and the base 1 is made of acrylic plate. The mounting seat 3 is an angled mounting seat, the drive motor 51 uses DJI's snail-2305 brushless servo motor, and the friction wheel 52 uses a 2305 friction wheel with polyurethane rubber as the edging material. In order to ensure that the friction wheel 52 is the first to contact the tennis ball when it contacts the car body, the horizontal length of the mounting seat 3 is lengthened to form a horizontal extension 33 to ensure that the friction wheel 52 is close to the ground. At the same time, in order to ensure sufficient strength of the mounting seat 3, a multi-angle bevel is provided through the mounting seat 3 for connection support. The mounting seat 3 is used to fix the snail-2305 brushless servo motor. In assembly, the top of the snail-2305 brushless servo motor is connected to the 2305 friction wheel through the M2 threaded hole and the bottom of the M3 threaded hole is connected to the mounting base 3. The tennis ball is guided and transmitted to the gap between the two friction wheels of the ball picking module, and the tennis ball is launched in the opposite direction by the principle of rotating and squeezing the tennis ball, so that it enters the ball storage box 6 for collection. Through the high-speed rotation of the friction wheel 52, the tennis ball is subjected to the reverse friction and squeezing effect of the friction wheel 52, and gains power to increase to the specified initial velocity, enters the predetermined route, and moves along the predetermined trajectory into the ball storage box 6, completing the ball picking and collection process.

[0046] In order to help improve the positioning effect of the trolley, the goal guide 4 is designed to help the tennis ball better contact the friction wheel 52 at the ball entry point, so that the trolley can collect tennis balls more efficiently. Therefore, after determining the blind spot distance L of the camera 9, the tangent length L1 between the goal guide 4 and the friction wheel 52 is measured, and the hypotenuse S of the goal guide 4 is calculated by the Pythagorean theorem. The size design of the goal guide 4 is completed. The size design calculation of the goal guide 4 is as follows Figure 5 , determine the minimum positioning distance of the camera 9, that is, after the camera 9 obtains the position information of the tennis ball, it drives to the distance from the camera 9 where the tennis ball information is gradually lost, and add some redundant length to this distance, then this length is the length L of the goal guide; next, the straight line L1 from one side of the goal guide 4 to the friction wheel 52 on this side and the vertical distance L2 from the friction wheel 52 to the farthest end of the goal guide 9 are regarded as the two right-angled sides of the right triangle, and the length S of the hypotenuse that sends the positioned tennis ball to the point where the friction wheel 52 enters the ball is calculated by the Pythagorean theorem, which is the length of the hypotenuse of the goal guide 4. Based on this right triangle calculation, the angle between the vertical length of the goal guide 4 and the hypotenuse length is 60°. The size of the goal guide 4 is calculated in this way.

[0047] The base 1 is also provided with a control module, which includes a main control board 10 and a battery 11. The control module is electrically connected to the mobile mechanism 2, the identification and positioning mechanism, and the ball picking mechanism, respectively, and the control module signal is connected to a remote monitoring terminal. In one embodiment of the present invention, the battery 11 is a dual battery design of one large and one small. The large battery is mainly used for power supply, and the small battery can be used as a backup power supply. The remote monitoring terminal is a mobile phone APP or a dedicated control platform. The user can view the robot's status information in real time, including key indicators such as power, position, and task progress. Once an abnormal situation or fault warning is detected, the system will immediately issue an alarm, and the user or technician can intervene remotely to make necessary adjustments or maintenance to ensure that the robot is always in the best working condition. This remote monitoring and maintenance mechanism not only simplifies daily management and maintenance work, but also provides a solid guarantee for the long-term stable operation of the robot in a complex environment.

[0048] On the other hand, the present invention discloses a method for positioning, identifying and picking up a ball, the method comprising the following steps:

[0049] S1, start the power supply, the Mecanum wheel drives the robot to rotate 360 ​​degrees, and the ball picking robot moves forward step by step. At this time, the camera 9 performs a 360-degree scan with itself as the origin to search for the target to be picked up;

[0050] S2, determine whether there is a tennis ball within the recognition range of the camera 9, if not, return to S1; if yes, enter S3;

[0051] S3, measure the distance of all targets in the field of view of the camera 9, and calculate the target distance using the following formula based on the principle of triangular similarity:

[0052]

[0053] but

[0054]

[0055] Wherein, Lm represents the distance between the robot and the target, Bpix represents the pixels occupied by the target in the camera 9, Apix represents a fixed image pixel, a represents half of the lens angle of view, and K represents a constant;

[0056] And calculate the distance factor based on the target distance:

[0057]

[0058] Where, distance = Lm, d represents the distance factor;

[0059] S4, assigning a category label to each of the tennis balls and obstacles such as sharp objects, and assigning a corresponding initial priority to each target based on the type of the target, wherein the initial priority of the tennis balls is set to a positive value, and the initial priority of objects such as sharp objects and obstacles is set to a negative value;

[0060] S5. Based on the potential threat posed by the target, the risk factor is calculated using the following formula:

[0061] r = risk_factor

[0062] Among them, r represents the risk factor, and risk_factor is a pre-set risk value;

[0063] S6. The robot calculates the final priority based on the initial priority, distance factor, and risk factor:

[0064] p=(initial_p+d)*|r|

[0065] Among them, p represents the final priority, initial_p represents the initial priority, d represents the distance factor, and r represents the risk factor;

[0066] S7, if the final priority value is positive, the robot will enter S8 to pick up the ball according to the maximum value of the final priority; if the final priority value is negative, the robot will avoid the maximum absolute value of the final priority pole;

[0067] S8, the Mecanum wheel drives the camera 9 to move so that the tennis ball to be picked up is accurately placed in the center of the field of view of the camera 9;

[0068] S9, the goal guide 4 collects the ball to the ball entrance 31, and the driving motors 51 on both sides of the ball entrance 31 drive the friction wheels 52 to rotate and drive the tennis balls to be launched, and the tennis balls enter the ball storage box 6 along the ball guide chute 32 for collection;

[0069] S10. Determine whether the battery power is sufficient according to a preset battery threshold. If sufficient, continue to pick up the ball; if insufficient, the Mecanum wheel drives the robot to return to charge.

[0070] In one embodiment of the present invention, through face recognition binding, after successful recognition, the robot starts to identify and track the tennis ball in the field of view. After successfully tracking the tennis ball, it can interact with the STM32 main control board to provide the coordinates and distance of the tennis ball, and the four Mecanum wheels start to rotate 360 ​​degrees in place. This action is not only to cover the maximum field of view, but more importantly, it starts the deep learning model carried by the robot and starts to analyze every frame from the high-definition camera in real time. The camera's field of view is carefully set to a square area of ​​400×400 pixels, ensuring sufficient detail capture and fast data processing speed. The deep learning model, after training with massive image data, can quickly identify and locate the target - whether it is a tennis ball or other types of balls, and can even identify potential obstacles and dangerous objects. The stepping distance of the Mecanum wheel is 5 to 10 meters for a single forward movement. It makes dynamic decisions based on the preset target priority. Tennis balls always come first, but when encountering potentially dangerous obstacles such as sharp objects, the robot will automatically adjust the path to avoid threats first and ensure the principle of safety first. This process is not just a simple recognition and sorting, but a deep integration of environmental perception and intelligent decision-making, which reflects the robot's adaptability in complex scenarios. In one embodiment of the present invention, the initial priority values ​​are tennis: 10, sharp objects: -5, obstacles: -10. The values ​​here are only examples. In actual applications, the priority values ​​may be adjusted according to specific needs.

[0071] In one embodiment of the present invention, when the tennis ball picking robot searches for a tennis ball through the camera 9, if the initial priority is 10, the distance is 5 meters, and there is no additional risk factor (i.e., sharp objects), then:

[0072]

[0073] In another embodiment of the present invention, if a sharp object is encountered at a distance of 10 meters and the risk factor is -5, then:

[0074]

[0075] In this way, even if the sharp object is far away, the robot will prioritize avoiding such potential dangers due to its high risk factor.

[0076] In the process of picking up the balls, the camera 9 is swung left and right to identify and stay away from dangerous objects on the left and right sides. After avoiding the dangerous objects, the tennis balls are collected. The ball storage box 6 is gradually filled, and it will return automatically after reaching a certain number. Generally, there are two ways to judge the return, one is to return when a fixed amount is reached, and the second is to return when the battery is low.

[0077] The SOC of a battery can be calculated in a variety of ways, one of the common methods is to estimate it through the integration method (also known as the ampere-hour method). This method is based on the measurement of the current during the battery's charge and discharge process, and calculates the energy consumed or stored by integration to deduce the remaining capacity.

[0078]

[0079] Where: SOC(t) is the battery SOC at time t (expressed as a percentage), SOC(t0) is the battery SOC at the initial time t0, Q nom is the rated capacity of the battery (in Ah or mAh), i(t) is the battery current at time t (positive numbers indicate charging, negative numbers indicate discharging)

[0080] In practical applications, since the calculation of continuous integrals is relatively complicated, a discrete accumulation method is usually used, namely:

[0081] This is derived

[0082]

[0083] In robotic applications, battery power management also includes consideration of factors such as battery health, temperature effects, and battery aging. In order to more accurately predict the remaining power, a machine learning algorithm can be used in combination with the battery model and historical data to make predictions. When the SOC drops to a preset threshold (such as 20% or 30%), the robot will trigger the command to return to charge and begin planning a path back to the charging station. The setting of this threshold needs to balance the robot's endurance and mission requirements to ensure that it can safely return to charge before the battery is exhausted.

[0084] Because the battery capacity and the battery consumption per minute can be determined, the remaining power can be obtained by subtraction. The walking time is calculated from the starting point and once it exceeds the set threshold time, it will automatically return.

[0085] Finally, it should be emphasized that the above is not intended to limit the present invention. For those skilled in the art, the present invention may have various changes and modifications. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A tennis ball picking robot, comprising a base (1), a moving mechanism (2), an identification and positioning mechanism and a ball picking mechanism, wherein the moving mechanism (2) is connected to the bottom of the base (1), the identification and positioning mechanism is connected to the base (1), and the ball picking mechanism is connected to the front side of the base (1), characterized in that: The ball picking mechanism comprises a mounting seat (3), two groups of goal guide frames (4) arranged outside the mounting seat (3), a ball picking module (5) arranged on the mounting seat (3), and a ball storage box (6) arranged at the rear side of the ball picking module (5); a ball inlet (31) is arranged in the middle of the lower end of the mounting seat (3); a ball guiding inclined groove (32) is arranged on the surface of the mounting seat (3) and is connected to the ball inlet (31); the two groups of goal guide frames (4) are closed toward the center in front of the ball inlet (31) to form a ball guiding channel.

2. A tennis ball picking robot according to claim 1, characterized in that: The ball picking module (5) comprises a driving motor (51) symmetrically arranged on both sides of the ball inlet (31) and a friction wheel (52) arranged on the output end of the driving motor (51).

3. The tennis ball picking robot according to claim 1, characterized in that: The identification and positioning mechanism comprises a raised frame (7), a pan / tilt platform (8) arranged on the raised frame (7), a steering gear (12), and a camera (9) arranged on the pan / tilt platform (8); the raised frame (7) is arranged on the machine base (1).

4. The tennis ball picking robot according to claim 2, characterized in that: The width of the ball guiding inclined groove (32) is greater than the spacing between the two groups of friction wheels (52); a horizontal extension portion (33) is provided at the end of the mounting seat (3); and the ball inlet (31) is sunken at the bottom end of the horizontal extension portion (33).

5. The tennis ball picking robot according to claim 1, characterized in that: The base (1) is also provided with a control module, the control module comprising a main control board (10) and a battery (11), the control module being electrically connected to the moving mechanism (2), the identification and positioning mechanism and the ball picking mechanism respectively, and the control module signal being connected to a remote monitoring terminal.

6. The tennis ball picking robot according to claim 3, characterized in that: The ball guide frame (4) comprises a connecting portion (41) and a gathering portion (42) integrally formed at the end of the connecting portion (41); the ball guide channel is located between the two groups of the gathering portions (42); and the connecting portion (41) is connected to the bottom of the elevated frame (7).

7. The tennis ball picking robot according to claim 1, characterized in that: The goal guide frame (3) is arranged above the ground, and the distance between the goal guide frame (3) and the ground is smaller than the diameter of a tennis ball.

8. The tennis ball picking robot according to claim 1, characterized in that: The moving mechanism (2) comprises four Mecanum wheels arranged at the bottom of the base (1).

9. A method for positioning, identifying and picking up a tennis ball applied to a tennis ball picking robot as claimed in any one of claims 1 to 8, characterized in that: The method comprises the following steps: S1, start the power supply, the Mecanum wheels drive the robot to rotate 360 ​​degrees, and the ball picking robot moves forward step by step. At this time, the camera (9) performs a 360-degree scan with itself as the origin to search for the target to be picked up; S2, determine whether there is a tennis ball within the recognition range of the camera (9), if not, return to S1; if yes, enter S3; S3. Measure the distance of all targets in the field of view of the camera (9) and calculate the target distance using the following formula based on the triangular similarity principle: but Where Lm represents the distance between the robot and the target, Bpix represents the pixels occupied by the target in the camera (9), Apix represents a fixed image pixel, a represents half of the lens angle of view, and K represents a constant; And calculate the distance factor based on the target distance: Where, distance = Lm, d represents the distance factor; S4, assigning a category label to each of the tennis balls and obstacles such as sharp objects, and assigning a corresponding initial priority to each target based on the type of the target, wherein the initial priority of the tennis balls is set to a positive value, and the initial priority of objects such as sharp objects and obstacles is set to a negative value; S5. Based on the potential threat posed by the target, the risk factor is calculated using the following formula: r = risk_factor Among them, r represents the risk factor, and risk_factor is a pre-set risk value; S6. The robot calculates the final priority based on the initial priority, distance factor, and risk factor: p=(initial_p+d)*|r| Among them, p represents the final priority, initial_p represents the initial priority, d represents the distance factor, and r represents the risk factor; S7, if the final priority value is positive, the robot will enter S8 to pick up the ball according to the maximum value of the final priority; if the final priority value is negative, the robot will avoid the maximum absolute value of the final priority pole; S8, the Mecanum wheel drives the camera (9) to move so that the tennis ball to be picked up is accurately placed in the center of the field of view of the camera (9); S9, the goal guide frame (4) collects the ball to the ball entrance (31), the driving motors (51) on both sides of the ball entrance (31) drive the friction wheels (52) to rotate and drive the tennis balls to be launched, and the tennis balls enter the ball storage box (6) along the ball guide chute (32) to be collected; S10. Determine whether the battery power is sufficient according to a preset battery threshold. If sufficient, continue to pick up the ball; if insufficient, the Mecanum wheel drives the robot to return to charge.

10. A method for positioning, identifying and picking up a ball according to claim 9, characterized in that: In S1, the single forward step distance of the Mecanum wheel is 5 to 10 meters.

Images