Infrared images enhancement using fuzzy dissimilarity histogram equalization

Abstract

In this article, a novel double-plateau limit histogram equalization (HE) method using the modification of the fuzzy dissimilarity histogram (FDH) is presented for the contrast improvement of Infrared (IR) images. Initially, FDH is formulated using fuzzy contextual information from the input IR image. Next, the average of the multi-peaks of the FDH is considered as the threshold limit of histogram sub-division. After that, sub-histograms are clipped based on their individual plateau limit values. Plateau-limit values of the two sub-histograms are based on the mean of individual histogram, and the remaining portion of each clipped sub-histogram is equally redistributed with the non-zero bins. Here, clipping of the histogram and redistribution process together can control over enhancement of image. Finally, the modified sub-histograms are independently equalized to form the desired output images. Simulation outcomes reveal that the proposed HE method effectively enhances the visual quality of the image. Compared with the other traditional state-of-the-art contrast enhancement methods, visual assessment and quantitative measurements like discrete entropy (DE), feature similarity index metric (FSIM), spectral residual similarity index measure (SR-SIM) and linear index of fuzziness, efficiently confirm the advancement of the proposed method. Moreover, a comparative result of the methods is demonstrated with a dataset of 100 IR images to show the superiority of the proposed method.