基于单目视觉的智能寻迹小车设计与实现

Design and implementation of the self-tracing trolley on the basis of monocular vision

目的研究基于单目视觉的智能小车循迹运动控制技术。方法通过HQ7620数字摄像头采集路径信息,采用快速OTSU自适应阈值算法获取路径引导中心线,计算并获得小车移动方向和速度控制参数,用PID算法对小车的舵机和直流电机进行闭环控制。结果设计并实现了一种以MC9S12XS128单片机为核心控制模块的智能循迹小车系统。实验结果表明,该智能循迹小车能够沿预设的引导线快速而稳定地行驶。结论设计方案简单可靠,具有较好的动态性能和鲁棒性,可应用于小型智能自主移动机器人的视觉导航控制。

Objective--To study the tracing motion control technology of intelligent trolley based on monocular vision. Methods--The control system can control the steering gear and DC motor of the trolley in a closed loop with PID algorithm by obtaining the motion direction and speed control param- eters of the trolley through the calculation after collecting the path information with HQ7620 digital camera and getting the guidable path centerline with rapid OTSU adaptive threshold algorithm. Re- suits--An intelligent tracing control system of the trolley has been designed and implemented based on MCU （Micro-Control Unit） of MC9S12XS128. The experimental result shows that the intelligent tracing trolley with proposed control methods can move fast and stably along the predefined guidable line. Conclusion--The proposed control method is not only simple, safe and reliable but also has good dynamic performance and robustness. This control approach can be applied to the visual navigation control of small intelligent mobile robot.