Edge Detection in the medical MR brain image based on fuzzy logic technique

In this paper, a novel technique is proposed based on fuzzy logic reasoning for edge detection in medical Magnetic Resonance (MR) brain image. The MR image is segmented into regions using 2×2 binary matrices. In our proposed work, it is possible to detect the edge without determining the threshold values. The ranges of values are mapped by the edge pixels distinct from each other. The Fuzzy Inference System (FIS) designed has four inputs which correspond to four pixels of scanning binary matrix and the output finds whether the pixel comes under consideration is "black", "white" or "edge" pixel based on rule base. The target pixel is classified based on the rule base comprises of more than twenty rules. The first and second derivative of resultant MR brain image from FIS is used to trace the edge of the image. The proposed method results is compared in terms of robustness with the conventional edge detection operators such as “Sobel”, “Prewitt”, “Roberts” and “Canny”.