| Abstrak/Abstract |
There are several techniques of pre-processing such as contrast enhancement and image restoration. This research will discuss about the comparison between the methods of contrast enhancement such as contrast stretching, adaptive histogram, and Contrast Limited Adaptive Histogram Equalization (CLAHE) with the aim to detect glaucoma according to texture features on retinal fundus images. To assess the quality of the image used in this study as glaucoma images of DRISTHI-GS and normal image of the RIM-ONE after applying contrast enhancement technique. The results obtained are CLAHE get the lowest MSE value, and the highest PSNR value. Multi-Layer Perceptron will be used as the classifier, and also accuracy, sensitivity, and specificity as assessment parameters. The results obtained are CLAHE get 100% of each of the valuation parameters, histogram equalization getting the accuracy of 97, 97%, and contrast stretching only get the accuracy of 95, 95%. So CLAHE is the best contrast enhancement method in this study to detect glaucoma on retinal fundus image using texture features. |
