Standard Intensity Deviation based Recursive Histogram Equalization Contrast Enhancement for Low-exposure Images

The low exposure images pose challenges in better visibility due to its low light conditions. The
visibility can be improved by contrast enhancement. To improve image contrast, the histogram
equalization (HE) is a famous method. The existing HE based algorithm for low exposure images
leads to an over enhancement problem and unnatural appearance. In this framework, the generalized
algorithm proposed for contrast improvement, it performs the separation of the histogram based on
respective standard intensity deviation value and then recursively equalizes all sub histograms
independently. The over-enhancement problem is minimized by this method. Added to this, the
presented methodology preserves image information and increases image brightness adaptively. A
total of 150 low exposure images are used to evaluate its performance and compared it with several
existing state-of-the-art algorithms. The experiment results are analyzed in terms of entropy, absolute
mean brightness error (AMBE), degree of entropy un-preservation (DEU) and output image
inspection. The proposed method results show significant improvement in enhancing low exposure
images.