Stereo Matching Technique And Its Application In Mobile Robots,

Stereo matching has been a bottleneck problem for more than thirty years since stereo vision came forth. To meet the demand of lunar vehicle, this paper focuses on the stereo matching and its use in the mobile robots navigation.First, an area based matching method is introduced. To improve the adaptability of the method, a texture enhanced technique based on an improved LoG filter is presented. To improve the efficient of the method, box filtering, average sampling and pyramid analysis are used so that the matching time is cut down to 1/10. .Second, a feature based matching method is presented, which is based on the area based matching method. This feature based matching method can be used to track feature points so that a new method of computing the optical flow field is presented. The experiments show that our method can obtain a much better result in precision and efficiency than the traditional methods. Last, the tracking method for feature points and circular consistency checking method are combined to design a visual odometer. The experiments show that the designed visual odometer can improve the precision of localization of the vehicle efficiently.