Sonar Features Extraction Algorithm for a Mobile Robot

Sonar sensor is an attractive range sensor for mobile robots, but the lack of good spatial resolution result it difficult to apply in localization or mapping. To overcome the bad angular resolution and the frequent number of outliers in sonar data, different sonar data fusion techniques are used. In this paper a method to extract point and line features from sonar data returns is presented. The method is based on the TBF algorithm, but line feature extraction is added. To demonstrate the effectiveness of the method, an experiment is carried out along a corridor and results are given.