Comparisons of planar detection for service robot with RANSAC and region growing algorithm

Abstract

Considering the life independence of people with disabilities or a limited range of movement, we intend to develop an intelligent mobile service robot equipped with robot manipulator and computer vision. In order to implement an automatic moving, grasping and delivering task, the abilities of service robot to deal with the images they see and collect useful information are really essential. In this paper, two different algorithms (i.e., the RANSAC algorithm and the region growing algorithm) of the planar detection for the service robot are introduced with detailed principles and implementation processes. For the foundation of our mobile service robot's development, three specific experiments are designed to compare the algorithms on accuracy, robustness and computational complexity. Finally, after the comparison between two algorithms, the suitability of two algorithms for the service robot using in the environment detection will be presented and analyzed.