A Multi-scale Enhancement Method to Medical Images Based on Fuzzy Logic

A main challenge of medical image enhancement is to solve conflicts between noise resistance and detail sharpening. A multi-scale enhancement method based on fuzzy logic (MEMBFL) is presented in this paper. The proposed method can prevent noise increasing during the sharpening of the image details. With this method, the original image is filtered by the hybrid low pass filter which is designed according to local context of the image in order to remove out the impulse noise and no-impulse noise. Then the filtered image is decomposed into several band pass images. Each pixel in band pass image is adaptively assigned a different enhancement factor by evaluating the local feature. The processed band pass images can be composed into an enhanced image by the process opposite to the decomposition process. Experimental results show that the proposed technique achieves good performances while can be easily used in medical image processing systems