Hybrid Vision Transformer and CNN for Detection of Overripe Coffee Berry Disease (OCBD) in Coffee Plantation

The article proposes an innovative Hybrid Vision Transformer and CNN architecture for the precise identification of Overripe Coffee Berry Disease (OCBD) in coffee farms. The suggested model adeptly integrates the advantages of Vision Transformers and CNNs, allowing it to capture both the global and local characteristics of coffee berries. Experimental findings indicate that the Hybrid model outperforms conventional CNN-based methods, with an impressive accuracy of 96.7% on a varied dataset. The model's resilience to fluctuating illumination conditions and obstructions makes it a viable alternative for practical applications in coffee plantation management.