Main Article Content
Purpose Numerous clinicians utilise digital histopathology images to diagnose diseases such as cancer and to get information about the aetiology of the illness. Typically, histopathology images exhibit poor image quality artefacts such as low contrast across various areas of the image, blurring, and inadequate lighting. The primary objective of this article is to offer a better and successful hybrid strategy for delivering overall contrast enhancement, accurately enhancing fine details, and producing a natural and distortion-free histopathological image.
Methods The novel hybrid method leverages the advantages of many established enhancing approaches to generate a natural and distortion-free histopathology picture from low-quality histopathology photos. After obtaining the brightness channel using a luminance measurement technique, two inputs are created utilising local histogram equalisation and retinex theory. The brightness component is changed by fusing the derived inputs with their neighbouring weights on a multi-level basis. Through a careful selection of inputs, their neighbouring weights, and a multi-level fusion technique. An extensive quantitative analysis demonstrates that the proposed method outperforms existing image enhancement techniques. The proposed method produced a maximum average peak signal-to-noise ratio of 29, a maximum average UIQI, SSIM, and FSIM of 0.99, 0.93, and 0.96, respectively, and a minimum average AMBE of 2.22.
Results The proposed approach was able to give superior overall improvement while preserving naturalness and superior performance in all performance measures such as AMBE, SSIM, FSIM, UIQI, and PSNR.
Conclusion The proposed approach produces images of higher quality, which is extremely beneficial for disease inspection and diagnosis.