Biomedical Signal Processing and Control 56 (2020) 101677 Contents lists available at ScienceDirect Biomedical Signal Processing and Control journal homepage: www.elsevier.com/locate/bspc A novel reformed histogram equalization based medical image contrast enhancement using krill herd optimization Pankaj Kandhway a , Ashish Kumar Bhandari a,∗ , Anurag Singh b a Department of Electronics and Communication Engineering, National Institute of Technology Patna, 800005, India b Department of Electronics and Communication Engineering, International Institute of Information Technology, Naya Raipur, India a r t i c l e i n f o Article history: Received 29 March 2019 Received in revised form 6 August 2019 Accepted 25 September 2019 Keywords: Reformed histogram equalization Plateau limit Krill herd optimization Salp swarm algorithm Medical imaging a b s t r a c t In this paper, a novel krill herd (KH) based optimized contrast and sharp edge enhancement framework is introduced for medical images. Plateau limit and fitness function are proposed in this paper to achieve the best-enhanced image. A new plateau limit is applied to clip the histogram using minimum, maximum, mean, and median of the histogram with a tunable parameter. The residue pixels are reallocated to the relative vacancy available on histogram bins. This method explores KH meta-heuristic algorithm to auto- matically adjust the tunable parameter based on a novel fitness function. Fitness function contains two different objective functions, which use edge, entropy, gray level co-occurrence matrix (GLCM) contrast, and GLCM energy of image for best visual, contrast enhancement and improved different characteristic information of the anatomical images. This method is compared with a different state of the art methods to check the viability and vigorous of the scheme and salp swarm algorithm (SSA) optimization is also used for the fair comparison of the proposed approach. The results show that the proposed framework is having superior performance compared to all the existing methods, both qualitatively and quantitatively, in terms of contrast, information content, edge details, and structure similarity. © 2019 Elsevier Ltd. All rights reserved. 1. Introduction Medical imaging plays a vital role to examine the health con- dition of a patient and provides an effective treatment. Medical images are used for the treatment and diagnosis of a large num- ber of diseases [1], but these images are complex in nature due to the presence of several overlapped objects in the image, which makes it difficult for the diagnostic process. Medical images, such as magnetic resonance imaging (MRI), X-ray, mammographic, and computerized tomography (CT) images, do not carry enough fea- tures for an accurate diagnosis due to low lighting conditions, environmental noises, technical restrictions of imaging devices, etc. Therefore, medical images have low quality and contrast. Con- trast collectively deals with the pixel intensity differences between structures and distinct objects in the image. Region of interest (ROI) [2] or an object can be easily observed in a good contrast image. Numerous image enhancement methods have been proposed so far in the literature for contrast and quality enhancement of the medical images. Some of them are histogram equalization, gamma ∗ Corresponding author. E-mail addresses: pankaj.kandhway@gmail.com (P. Kandhway), bhandari.iiitj@gmail.com (A.K. Bhandari), anurag2685@gmail.com (A. Singh). correction, and transform based approaches which have been widely used for improving the features, contrast and visual percep- tion of both medical and natural images. Histogram equalization (HE) based algorithms are extensively used for contrast enhance- ment because of their effectiveness and simpler implementation. Global histogram equalization (GHE) technique generates seri- ous constraints such are visual artifacts, noise amplification, level saturation effect, under and over-enhancement, which are not acceptable in medical imaging. Several HE-based frameworks have been emerged based on different mechanisms to overcome these limitations of the GHE. Sub-histogram, histogram clipping, and dynamic histogram equalization are the basic mechanisms that have been utilized by authors [3–7] to enhance the image’s qualities and contrast. Joseph et al. [8] introduced a fully customized enhancement framework for medical images where an arbitrary clip-limit is exploited for the clipping process, but it is tedious to set a threshold value. In anatomical images, local details can be more signifi- cant than global contrast, but local enhancement (LE) frameworks generate undesired artifacts, block effects, and also increase com- putational complexity. To enhance the fidelity features and contrast of the images, adaptive histogram equalization (AHE) technique [9] divides the image into small blocks and improve the pixel values of each block based on the several operations. All enhanced blocks are https://doi.org/10.1016/j.bspc.2019.101677 1746-8094/© 2019 Elsevier Ltd. All rights reserved.