Contrast and Luminosity Enhancement of Retinal Images Using Weighted Threshold Histogram

Retinal images are used for detection of many eye diseases like diabetic retinopathy, glaucoma and age-related macular degeneration. Faulty quality retinal images are strenuous for disease diagnosis because of low contrast and uneven illumination. A novel technique is proposed to enhance the contrast and luminosity of retinal images. The contrast of the retinal image is improvised with the Weighted Threshold Histogram Equalization (WTHE) technique. The WTHE technique is applied on the L* channel of the L*a*b* color space. The luminosity of the retinal image is improved by using the gamma correction and luminance gain matrix. Gamma correction is done on the V channel of the HSV color space. This method uses two publicly available datasets: DRIVE and STARE datasets. The quality of enhancement is assessed using the Global Contrast Factor (GCF) metric. The average GCF value of the proposed model is 5.90548, surpassing the state-of-the-art method by nearly two points.