Adaptive fuzzy logic based Bi — Histogram equalization for contrast enhancement of mammograms

Mammography is the primary and most reliable technique for breast cancer detection. Mammography being low cost low dose X-ray procedure, it can produce low contrast noisy images of poor visual quality. This can result in subtle malignancies remaining unnoticed. Histogram Equalization and its variations are some of the commonly used methods for contrast enhancement. However, these methods when applied to mammograms produce images with washed out appearance and sometimes, with uncontrollable change in luminance and brightness. To improve the quality of mammograms for better perception, adaptive fuzzy logic based bi-histogram equalization (AFBHE) is proposed in this paper which combines fuzzy logic with Brightness Preserving Bi-Histogram Equalization. The merit of the proposed method is that it is adaptive in nature where all the parameters are computed based on the characteristics of the mammographic images. In the first stage, grey level intensities are transformed to an adaptive fuzzy plane with values ranging from 0 to 1. In the second stage, bi-histogram equalization is applied to the fuzzy domain. In the final stage, the fuzzy plane is mapped back to grey level image. The proposed method has been tested on several mammograms. Both subjective quality assessment and objective quality assessment techniques like PSNR, SSIM, NCC, UIQI and DE has been applied to evaluate the performance of the proposed method. Experimental results show that the proposed method is found to have an edge over other contemporary methods in terms of visual quality, local information preservation and controlled contrast enhancement.