An Improvement of Hough Transform for Building Feature map

Map building is very important to accomplish autonomous navigation and other complex intelligent tasks. A novel method of feature map building was proposed to process the uncertainty data and noise data of sonar. This method has introduced sonar arc to describe detectable region for sonar, and extracted line by a novel Hough transform based on randomized Hough transform and multi-resolution Hough transform to accomplish autonomous map building of mobile robot. So uncertainty in sonar sensor responses can be successfully reduced. The experiment results indicate that the method is more effective and better real-time performance.