Fast two-dimensional entropic thresholding algorithm
Wen-Tsuen Chen, Chia-Hsien Wen, Chin-Wen Yang · Proceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIE · 1993
Two-dimensional entropic thresholding is one of the important thresholding techniques for image segmentation. Usually, the global threshold vector is selected from L2 (gray level, local average) pairs through a `maximum' optimization procedure with O(L4) computation complexity. This paper proposes a fast two-phase 2D entropic thresholding algorithm. In order to reduce the computation time, we estimate 9L2/3 candidate threshold vectors from a quantized image of the original in advance. The global threshold vector is then obtained by checking candidates only. The optimal computation complexity is O(L8/3) by quantizing the gray level in L2/3 levels. Experimental results show that the processing time of each image is reduced from more than two hours to about two minutes. The required memory space is also greatly reduced.