An improved enhancement algorithm based on OTSU adaptive threshold and fuzzy set theory for robot digital roadmap
Yuan Luo, Hongmei Yang · Proceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIE · 2009
In robot path planning, the robot digital roadmap recognition is very important. The image preprocessing can affect the recognition result. Image enhancement is always used in image preprocessing to provide better input for the following automated image processing and fuzzy image enhancement is one of a effective approach that be used widely in image preprocessing. Through the research of the Pal fuzzy enhancement algorithm, there were two main bugs of the algorithm. The first one was the definition of the membership function, and the second was the selection of the transition point. In this paper, a novel improved algorithm is proposed by analyzing the deficiencies of Pal algorithm. First, use the 2D OTSU operator to automatically select the best threshold value, solve the disadvantage that the transition point should be set artificially, and eliminate the random selection. Then, in order to improve the image information loss problem, modify the core membership function of the fuzzy enhancement algorithm. Finally, the new algorithm was applied to process digital roadmap for Pioneer III robot. The experimental results indicate that the proposed image enhancement method can achieve better performance, and have higher processing speed than now-available fuzzy image enhancement algorithms. And it also possesses the property of high practicability and generality.