Parallel Color Image Segmentation Based on Fuzzy Membership
Yanli Hou · Computer Technology and Development · 2011
Indoor mobile robot has a wide range of applications which is one of the hotspots in robot research.To improve the real-time performance for visual navigation of the mobile robot,a parallel color image segmentation algorithm using peer group filter(PGF) and fuzzy membership is studied.Firstly,the mobile robot navigation task is decomposed into different subtasks corresponding to the different objects,and then a general parallel structure based on multiple processors is proposed,used to extract the different objects simultaneously.Secondly,a peer group is determined corresponding to the interested object in a subtask using fisher discrimination,at the same time,defining the fuzzy membership function to improve the segmentation effect.At last,simulations are performed on the presented parallel algorithm,and the simulation result shows that the presented algorithm is not only simple but also effective.