Glaucoma Detection using HOG and Feed-forward Neural Network

The second most common cause of blindness in the world is glaucoma. It is an eye condition that doesn't show any early symptoms until considerable vision loss has already taken place. To reduce the risk of vision loss, glaucoma diagnosis is therefore crucial. This study provides an assessment of detecting retinopathy, glaucoma, and healthy images using a feature extraction method called Histogram of Oriented Gradients (HOG), and feed-forward neural network used as a classifier to classify the refunds images. The Contrast Limited Adaptive Histogram Equalization (CLAHE) method, morphology-based homomorphic filter, wavelet-based homomorphic filter, andmulti-scale top-hat transformation can all be used to enhance retinal images and increase their quality and dynamic range. Retinal fundus images are enhanced to increase contrast and bring out the retinal vessels. From the HOG we extracted the different mean, variance, Skewness, kurtosis, entropy, and energyfeatures set. Further, the extracted features set are given to the feedforward neural networks to classify the images.

I. Introduction

The glaucoma word was derived from the Greek word, which means ‘clouded’, that describes a human with a dilated cornea. Glaucoma damages the optic nerve which results in permanent blindness [1]. To detect glaucoma image processingtechnique can be employed. Extracting the features like Cup-Disc ratio (CDR) and Neuro retinal rim (NRR) from retinal images leads to the categorization of retinopathy eye and healthy eye. Glaucoma is the initial stage of permanent vision loss. Frequently, patients do not experience signs of vision loss until the disease has progressed. The extreme rise in intraocular pressure inside the eye is what hascaused this injury. This intraocular pressure (IOP) increasesto abnormal levels, occasionally sharply reaching 60–70 mmHg. When sustained for a long duration, pressures higher than25 to 30 mm Hg can result in visual loss. Glaucoma patients experience abnormally high pressure as a result of a drainage canal system that is partially or entirely blocked. In generally healthy eyes, there is a balance between the fluids that are produced inside the eye and those that depart it. Inter-Ocular Pressure (IOP) is maintained by this equilibrium, however, inglaucoma, this equilibrium is lost, which causes IOP to rise[2]. It is shown that a family history of genetic conditions can contribute to the development of a disease. It is anticipatedthat at least half of the sufferers in wealthy nations experienceno symptoms of the condition, with the situation being morelikely worse in developing nations. A patient with suspected glaucoma can undergo a number of tests, including tonometry to assess the eye's pressure, laser therapy, or surgery; ifidentified early, the condition can be delayed or even reversed. These procedures aim to lower intraocular pressure [3]. For the purpose of detecting Glaucoma earlier, numerous research projects have been conducted. The disease can be detected using feature learning-based approaches and segmentation [4]. Depending on the imaging modalities, different features of the retinal anatomy, such as the retinal nerve fiber layer,peripapillary atrophy, optic nerve head (ONH), and cup, must be investigated in order to diagnose glaucoma. The optic nerve head is a beaming, rounded region in fundus pictures, while a cup is a smaller rounded region inside the ONH. Previous research related to glaucoma detection is based on the increase in cup size which results in the optic disc. The features like