RETRACTED ARTICLE: An Intelligent Optimized Deep Network for Retinopathy Diabetes Segmentation

Retinopathy diabetes is the most interesting field in the optical world for improving vision-related problems. However, it is difficult to forecast the affected part and its types in the retinal image because of improper analyzing features. So, the current work intended to develop a novel Hybrid Horse-herd based Convolutional ResNet Segmentation Framework (HHbCRSF) for predicting the extraction of diabetes cells in the retinal images. Initially, the noise filtering function was activated in the hidden layer of the HHbCRSF to remove the noise parameters. Then the refined data is taken as the input for the classification phase for analyzing and tracking the affected features. Finally, the affected region was segmented, and the types were classified. This planned system is implemented in the Python framework. The successive score was measured as Accuracy, error rate, sensitivity, and specificity. In that, the newly developed design scored the maximum Accuracy and less misclassification score for every disease.