Intelligent Tennis Ball Pick-Up Trolley Based On Visual Recognition
Yi En Chen · 2025
In order to more accurately identify tennis balls in complex tennis court environments, a cross shaped window self attention mechanism was introduced. Choose a separable core attention module to handle the recognition problem of tennis balls at different distances and scales. Based on the structure and kinematics of the robotic arm, it ensures that the robotic arm can accurately and quickly complete the ball picking action, and combined with path planning strategies, enables the car to autonomously navigate to the tennis ball position and safely return, while avoiding collisions with other obstacles in the field. The results show that the method proposed in this paper has excellent detection performance, with an accuracy of 98.3%. The intelligent cart displays a high degree of consistency between the actual movement trajectory and the preset expected trajectory, with an accuracy rate of up to 98.2%. The results show that the car can work stably and reliably in complex environments, significantly improving the efficiency and accuracy of tennis ball picking.