A Fast Method of Automatic Multilevel Thresholding

摘要

Previous optimal thresholding methods are very popular and efficient in the case of bi-level thresholding. But they are very computationally expensive when extended to multilevel thresholding to optimize the objective functions. This paper proposed a new method for automatic multilevel threshold selecting based on max'mum entropy theorem. New method has some advantages, such as more simply, batter stabilization and easier performance comparing with some traditional entropy methods. It can calculate thresholds quickly. We employed this approach to perform image enhancement, segmentation and thresholding, and obtained satisfactory results.