Bi-Histogram Equalization with Adaptive Multi-Plateau Limits for Enhancing Magnetic Resonance Images

In this study, a highly efficient algorithm for enhancing the visual quality of brain Magnetic Resonance images is proposed. This method sequentially processes the gray-level histogram of the brain MR image by defining critical plateau limits. Since the lower and upper halves of the histogram are separately processed, the proposed algorithm produces significant results that facilitates subsequent morphological analysis effectively. Adaptive multiple plateau limits are computed on the upper and lower histogram. Thus the proposed algorithm not only rectifies the visual imperfections in the image but also avoids any unwanted noise to be introduced into the input image. The experimental results show that the proposed method is superior to many existing histogram-based algorithms, in terms of various standard quantitative metrics.