星球漫游车超广角实时立体视觉系统

A Real Time Stereo Vision System for Planetary Rover with a Very Large Field of View

给出了一种用于星球漫游车障碍检测和定位的超大视场立体视觉系统的实现方法．该系统采用具有超广角镜头(对角视场角约160度)的双目或三目摄像机获取场景立体图像对，利用摄像机标定参数对大变形图像进行修正等预处理，然后在外极线、连续性等约束条件下，基于查找表和 Intel MMX 指令集，使用 SAD 算法快速进行对应点匹配计算．实验表明，该系统在图像分辨率为320×120像素、视差为64级时，利用普通工控机恢复稠密深度图的速度为10帧/秒，并能使机器人以1米/秒的速度行走．

Implementation of a real time stereo vision system for obstacle detection andlocalization is discussed for planetary rover.The system employs binocular or trinocularstereo with fish-eye lenses to capture stereo pairs with a very large field of view(FOV～160°).Camera calibration parameters are used to perform distorted image correction,epipolarrectification and LoG filtering,and the SAD similarity is computed to estimate dense depthmaps under the epipolar and smooth constraints.Using the Intel MMX instruction set andLUT,the system can complete dense depth mapping and obstacle detection in real time.Thetest result shows that the system can produce dense depth maps at 10Hz with a resolutionof 320X120 pixels and a disparity of 64 levels using standard PCs,are allow the robot tosafely travel at 1m/s.