中文摘要：

在所构建的摄像机与激光雷达组合感知系统上,对基于特征点的配准问题进行了研究。结合3D激光雷达与摄像机工作特性,针对激光雷达混合像素及边界点问题,设计了基于相关扫描序列滤波器（Associated Scanning Sequence Filter,ASSF）,很大程度上减少了混合像素对配准特征点提取的影响,提高了提取准确度,使3D激光雷达与摄像机配准结果更为精确;另外,提出了基于结构化特征的激光雷达坐标原点位置标定方法,减少了激光雷达的量测误差。基于以上研究结果,使组合感知系统能够准确地对移动机器人行驶环境进行感知。

英文摘要：

With the integrated perception system of camera and laser ranger,the problems based on feature point matching are researched.Combined with 3D laser ranger and camera working features,referring to the problem of laser ranger mixpixels and edging points,associated scanning sequence filter（ASSF） is designed.The influence of registration feature point extraction from mixpixels is largely decreased,extraction precision is increased,and the registration results of 3D laser ranger and camera are more accurate.Besides,laser ranger coordinate origin position calibration method based on structure feature is proposed,which reduces the measuring error of laser ranger.The research results show that the integrated perception system can accurately perceive mobile robot running environment.