A dual vision-guided mobile robot control approach for multi-target path planning and intelligent pickup

Wenyong Zhang, Dawen Xia, Guoyan Chang, Yang Hu, Yujia Huo, Ziqiang Wang, Fujian Feng, Zhiquan Zheng, Yantao Li, Huaqing Li

Abstract: With the rapid development of the Internet of Robotic Things (IoRT), the autonomous sensing and decision-making capabilities of IoRT in multi-target pickup tasks have become a research hotspot. However, existing methods still face significant challenges in optimizing multi-target path planning and intelligent pickup operations. To this end, we propose a dual vision-guided mobile robot control approach based on the IoRT that combines multi-target path planning and intelligent pickup, called MTPP, in which robots employ dual cameras as sensors to pick up multiple targets. Specifically, the MTPP approach consists of an ant colony optimization algorithm which considers the rotation angle and the target distance (AD-ACO), and the dual-camera intelligent pickup method (DIP). In the AD-ACO algorithm, we present a new method to calculate the actual distance between targets based on the pixel coordinates of two points (PDC). Then, we convert the robot’s rotating angle when pickup each target into the corresponding arc length of rotation. Next, the actual distance and the arc length of rotation are summed to obtain the effective distance. The effective distance is employed as a parameter of the ant colony optimization algorithm for path planning to generate the optimal path. Moreover, the DIP method is composed of a pickup robot, a robot positioning method, and a dual-camera navigation method, where the robot is utilized for collecting targets, the positioning method is used for real-time robot localization, and the dual-camera navigation method is adopted to guide the robot to the targets. Finally, our MTPP approach is applied to a smart tennis collection system enabling efficient and accurate operations.