Survey over image thresholding techniques based on entropy

Image segmentation is an important problem of digital image processing and also a common difficult problem. The research conducts a survey of image thresholding methods based on entropy in recent years. First, the analysis of some representative thresholding methods has been made from several aspects, such as the ideas of algorithm, key technology. Then, the thresholding function of these typical methods is listed. At last, we evaluate the performance in the context of nondestructive testing applications based on the combined performance measures. The above work can give a valuable reference for image segmentation.