Classification of Fundus Diseases Using Hybrid Techniques Based on the Fused MobileNet and Handmade Traits

Early detection of fundus diseases is climacteric for timely ministration and prevention of darkness. These diseases in early stage have similar symptoms, making it difficult to distinguish between different types of diseases. Computer-assisted automated diagnostic techniques are needed to overcome the challenges in distinguishing between fundus disease types. This study focuses on classifying four fundus diseases like CTRT (Cataract), DR (Diabetic retinopathy), GLA (Glaucoma), and AMD (Age-related Macular Degeneration) using hybrid techniques pursuant to the merged MobileNet and handmade traits. Methodology involved in this study classifying the OIHA collection using an ANN according to merged traits from the MobileNet and DenseNet121 models along with handmade traits. The ANN achieved an accuracy of 98.6% according to the merged MobileNet and handmade traits.