• Dibyendu Mandal
• Souvaghya Hore
• Dibyendu Ghosh
• Sujay Kumar Roy

Diabetic retinopathy originated from a root optical disorder is invariably predominant cause to ratify the impairments caused in vision and other cases leading to blindness in patients suffering from diabetes mellitus. It is very obvious that prior identification and detection sometimes turns up to be crucial and vital at the same other end to prevent the ailment and related complications. To forecast a preventive outlook a comprehensive and clarified real time predictive model has been proposed in this work designed with MATLAB. Starting from preprocessing, enhancing the image, segmenting optical regions independently and most importantly the critical examination in segmenting blood vessels of optical cells has been undergone. Through methodologies of thresholding and independent morphological segregation the actual exudates are detected in each stage. The extremities of diabetic retinopathy have been distinguishably classified into 3 major stages like moderate, severe and mild. Extracted features mainly highlight the necessity of features extracted from various models making the system most reliable, accessible and more accurate. Leading this tool more forward this methodology can be implemented to reflect the prediction as detection in classification of severity of images in consideration.

Retinopathy, MATLAB, diabetes, thresholding, features, optical cells, detection, segregation

"Deep Learning-Enhanced System for Detection and Grading of Diabetic Retinopathy", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.12, Issue 5, page no.c791-c794, May-2025, Available :http://www.jetir.org/papers/JETIR2505308.pdf

"Deep Learning-Enhanced System for Detection and Grading of Diabetic Retinopathy", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.12, Issue 5, page no. ppc791-c794, May-2025, Available at : http://www.jetir.org/papers/JETIR2505308.pdf