Hybrid CNN & Random Forest Model for Effective Onion Leaf Disease Diagnosis

The classification and detection of several onion leaf diseases including Sour Skin, Onion Smut, Downy Mildew, Botrytis Leaf Blight, Purple Blotch, Smudge, White Rot, along with Rust are the topic of this study paper. The study uses accuracy, recollection, F1-score, and other metrics to assess a machine learning model's effectiveness in diagnosing various disorders. The model has an overall accuracy range of 94% to 97%, with an average weighted F1-score of 81.50%, demonstrating its resilience in disease identification. The model's consistency in properly recognizing disease cases is highlighted through a micro-average F1-score of 81.37%. This study helps to improve the detection and control of onion leaf illnesses, which is critical for maintaining onion plant health and output. This study makes a substantial contribution to the discipline of onion leaf disease management by providing an accurate and effective method of identifying illnesses and categorization.