Modified Barnacles Mating Optimization with Deep Learning based Tomato Disease Detection and Classification

Tomatoes are the second largest crop grown in India. They are more commonly known vegetables and be possible to find in all kitchens. It is the most popularly produced crop after potato and sweet potato. On the other hand, various diseases might damage the quantity and quality of tomato crops. Leaf disease detection and classification employs different DL algorithms which is a cutting-edge image processing technique yet produces comparatively new and proficient outcomes. This study develops a new Modified Barnacles Mating optimization with Deep Learning based Tomato Disease Detection and Classification (MBMODL-TDDC) method. The MBMODL-TDDC method accurately determines the effectual identification of different kinds of tomato diseases. To accomplish this, the presented MBMODL-TDDC method uses a Wiener Filtering (WF) based noise removal process at the preprocessing step. In addition, the MBMODL-TDDC technique employs a RetinaNet object detector for tomato disease detection. Moreover, Hybrid Deep Learning (HDL) algorithm was applied for classifying tomato diseases into different types. Eventually, the MBMO algorithm is used to adjust the hyperparameter values of the HDL model. The simulation results of the MBMODL-TDDC technique are tested on the tomato disease dataset. The result analysis portrayed the improvement of the MBMODL-TDDC algorithm over other current methodologies.