A New Image Thresholding Algorithm Based on Fuzzy sets Theory

Many classical measures partition image according to a single property. Moreover, many schemes suffer from the lack of evaluation of image quality at the global level. This paper proposes a novel two phases image thresholding measure that uses both global and local image properties for grayscale images. In the local phase, we present a novel thresholding technology which proposes threshold as multi-properties (ultra-fuzzy entropy and ultra-fuzzy similarity) based on type II fuzzy (ultra-fuzzy) sets. In the global phase, a nonlinear contrast intensification function is used to further enhance the image. In experiments conducted on various classic images, this algorithm showed notable visual improvement in comparison with common measures.