Fast recurring algorithms of image thresholding based on two-dimensional Renyi's entropy
Yiquan Wu · Chinese Journal of Stereology and Image Analysis · 2007
Thresholding is a basic problem in image processing.Threshold selection is crucial to thresholding.With the continuing development of thresholding technique,the requirement of thresholoding algorithms is increasingly accurate and fast.In this paper,two fast recurring two-dimensional Renyi's entropic thresholding algorithms are proposed,which may reduce repeat computation and increase the computational velocity.Some experiments were made on all kinds of images.The results show that two-dimensional Renyi's entropic thresholding algorithm achieves a good segmentation,and the computational time of its two fast recurring algorithms is less than 0.1% of that of the original algorithm.That makes two-dimensional Renyi's entropic thresholding algorithm much more practical.