基于灰度-模糊熵直方图的阈值分割方法

Gray level-gradient based thresholding method utilizes the edges to characterize the spatial relationship between pixels, and its performance is dramatically improved. In this method, the gradient operator is adopted to obtain the edge information, however, the gradient operator is sensitive to noise and can not efficiently detect the weak or fuzzy edge. To overcome this drawback, in this paper, a new thresholding method based on gray level -fuzzy entropy (GLFE) histogram is proposed. GLFE histogram consists of pixels' gray level and local fuzzy entropy of its neighboring pixels. The local fuzzy entropy can characterize the spatial relationship between pixels. Furthermore, a nonlinear fuzzy membership function is used to map the original image to fuzzy domain in the procedure of calculating fuzzy entropy, making it not only suppress noise but also enhance the weak or fuzzy edge, and thus the local fuzzy entropy can character the spatial relationship between pixels more efficiently. The experimental results demonstrate the effectiveness of the proposed method.