This project classifies fruit quality into five levels (Fresh to Rotten) using image preprocessing, data augmentation, and clustering (K-Means, Agglomerative, DBSCAN). Deep learning models are trained on clustered data, with AlexNet + Agglomerative Clustering achieving top results (Acc: 0.84, F1: 0.83).