Contrast modification for pre-enhancement process in multi-contrast rubeosis iridis images

Abstract

Existing researchers for rubeosis iridis disease focused on image enhancement as a collective group without considering the multi-contrast of the images. In this paper, the pre-enhancement process was proposed to improve the quality of iris images for rubeosis iridis disease by separating the image into three groups; low, medium and high contrast. Increment, decrement and maintenance of the images' original contrast were further operated by noise reduction and multi-contrast manipulation to attain the best contrast value in each category for increased compatibility prior subsequent enhancement. As a result, this study proved that there have three rules for the contrast modification method. Firstly, the histogram equalization (HE) filter and increasing the image contrast by 50% will achieve the optimum value for the low contrast category. Experimental revealed that HE filters successfully increase the luminance value before undergoing the contrast modification method. Secondly, reducing the 50% of the image contrast to achieve the optimum value for the high contrast category. Finally, the image contrast was maintained for the middle contrast category to optimise contrast. The mean square error (MSE) and peak signal-to-noise ratio (PSNR) of the outputs were then calculated, yielding an average of 18.25 and 28.87, respectively.

Future Iris Imaging with Advanced Fuzzified Histogram Equalization

“…Using PSNR as a measure evaluates the accuracy of the image enhancement in terms of pixellevel fidelity. The proposed approaches achieved higher PSNR values than existing state-of-the-art methods [1], [18], [23], as shown in Table II, indicating that AFHE, CLAHE, and FCE preserve image quality and reduce distortion throughout the enhancement process. PSNR is commonly employed to evaluate methods such as HE, CLAHE, and FCE that modify image brightness; however, it may not adequately measure perceptual quality or task performance.…”

“…This technique reduces potential noises in the image while enhancing contrast in grayscale images. A study in [18] presented an image enhancement method using HE to increase the quality of iris images for rubeosis iridis disease. The processes were divided into three image groups: low, medium, and high.…”

Image noise reduction by deep learning methods

“…When shooting outdoors, these noises can be especially noticeable on dark backgrounds and at night. To remove such noise in images, various machine learning methods can be used [8]- [10] such as filters, CNNs, and other algorithms. CNNs can be trained on a large number of images with different types of noise to automatically remove noise on new images.…”