An image segmentation method based on maximizing fuzzy correlation and its fast recursive algorithm

• An image segmentation method is proposed by maximizing fuzzy c-partition of the whole image.
• A fast recursive algorithm is proposed to reduce the computation burden of the proposed method.
• Experimental results on synthetic, brain MR and casting images show that the proposed method has good performance.

In this paper, an image segmentation method is proposed that integrates fuzzy 2-partition into Yen’s maximum correlation thresholding method. A fuzzy 2-partition of the image is obtained by transforming the image into fuzzy domain by means of two parameterized membership functions. Fuzzy correlation is defined to measure the appropriateness of the fuzzy 2-partition. An ideal threshold is calculated from the optimal membership functions’ parameters, which make the corresponding fuzzy 2-partition have maximum fuzzy correlation. In the process of searching the optimal parameters of membership functions, a fast recursive algorithm is presented in order to reduce the computation complexity. Experimental results on synthetic image, brain magnetic resonance (MR) images and casting images show that the proposed method has a satisfactory performance.

Figure optionsDownload as PowerPoint slide