Image thresholding using measures of fuzziness

In image analysis, image thresholding which is used for separating the object from the background is one of the most common application. For the preprocessing purposes of an image, thresholding is a necessary tool. Here, a new method based on minimizing measures of fuzziness is presented. Every pixel in the image and the averaged image have a membership value. Using these membership values entropy measures of fuzziness of image set and averaged image set are calculated. Then, threshold is founded by optimally minimizing calculated measures. The experimental results indicated that the new method has good performance but takes a long time in averaging.