Adaptive Line Extraction for Slam Line Based Applications in Unknown Environments

This paper presents an approach for line extraction from 2D laser range finder which has excellent precision to determine of line boundary, while having efficient speed and less complexity. This noble algorithm is more robust in measurement noise and cluttered data, adaptive in different kind of surface and fit a line to a set of uncertain points via Hough transform and recursive split formalism. In this algorithm first the data from laser is segmented to separate regions by determining start and end points throw Hough transform and noisy or sparse points are removed as well, and then extracts lines with a noble fast recursive split algorithm. Furthermore, it has good real-time capability to integrate the information of laser scanner into the navigation algorithm of the mobile robot and it is tested in NAJI II Autonomous rescue robot.