履带式挖掘机器人CCD道路识别

为实现液压挖掘机器人的无人操纵、自主作业,对小松液压挖掘机进行机器人化改造,在对挖掘机行走机构、回转机构、工作装置实现全液压电液比例控制技术的基础上,通过电荷耦合器件(Charge Coupled Devices,CCD)摄像头采集室外规定路径图像,图像经平滑、去噪、二值化、边缘提取等处理后,对路径进行识别,将路径识别特征量作为履带式液压挖掘机左右行走液压马达的驱动控制信号,实现挖掘机对路径轨迹的自主跟踪.在二值图像基础上,采用直接边缘检测和跟踪边缘检测相结合的识别方法,保证路径识别可靠性.改造后的挖掘机器人集机械、液压、自动控制、机器视觉等技术于一体,采用摄像头获取路径图像,构成视觉伺服控制系统,控制挖掘机的行走机构,经实验验证是可行的.

CCD-enabled road recognition for caterpillar excavator robots

To realize unmanned and autonomous operations on hydraulic excavator robots,the robot of a Komatsu hydraulic excavator is refined.Based on the full electro-hydraulic proportional control technology,which is applied for the excavator walking and rotary mechanisms and working devices,the predetermined-path images are collected by an outdoor charge coupled devices（CCD）camera.On the basis of such image processing techniques as smoothing,noise-elimination,binary and edge extraction,the road recognition is conducted.By treating path recognition features as driving control signals for right and left hydraulic excavator walking motors,the autonomous path tracking is carried out.By combining the direct and tracking edge detections on binary images,the reliability of path recognition can be guaranteed.Due to the integrated technology containing mechanical,hydraulic,automatic control and machine vision techniques,this approach is proven feasible by the means of CCD-enabled path image acquisition,vision-oriented servo control system and excavator-related walking mechanism.