A Chaotic Butterfly Optimization Driven Deep Learning Approach for Tomato Leaf Disease Detection and Classification

Disease recognition in tomato crops is significant to minimize crop loss and safeguarding food safety. Automated models utilizing high-resolution imaging and computer vision (CV) models evaluate images to recognize various infectious agents. Machine and deep learning (ML, DL) models given training on labeled datasets, enable precise disease recognition and classification, empowering agriculturalists in taking timely and targeted actions to manage diseases. This study presents a Chaotic Butterfly Optimization Algorithm DL Approach for Tomato Leaf Disease Detection and Classification (CBOADL-TLDDC) technique. The CBOADL-TLDDC technique concentrates aims for detecting and categorizing discrete tomato leaf ailments. The CBOADL-TLDDC technique utilizes Gaussian filtering (GF) model, and also employs the ShuffleNet network for feature extraction. The CBOA and extreme gradient boosting (XGBoost) techniques are employed for hyperparameter tuning, and to detect and classify tomato leaf ailments. The investigational validation of the CBOADL- TLDDC technique exhibited a greater accuracy value of 93.75% over other models.