Early Identification of Crop Leaf Disease Detection System Through Artificial Intelligence (AI) Assisted Deep Learning Methodology

The productivity of agricultural systems is critical to sustaining global food security. Unfortunately, it is known that many important crops, such as potatoes, tomatoes and mango, suffer from leaf diseases that cause important yield and quality decreases. Routine diagnostic procedures of these diseases are based on visual observation, which may be laborious, time-consuming, and may be affected by subjectivity. To address these limitations, this work presents a powerful Crop Leaf Disease Detection System based on Deep Learning Methodology (Crop-DLM) for accurate detection and classification of the leaf disease in miscellaneous images of crop. The Crop-DLM is enriched with a dedicated probability score index for assisting the model to capture and differentiate probability score priors when scarce labelled data exist, demonstrating a large potential of this application. The system is highly effective, with a $99 \%$ overall accuracy, $99.12 \%$ sensitivity, and $99.05 \%$ precision, that dominate over those already available. By verifying such a deep learning-based architecture, the study goes beyond existing multi-crop disease detection approaches and provides applicable, efficient, and viable solutions to assist sustainable agricultural production systems.