Enhance the growth of agriculture by predicting crop diseases using optimization and deep learning

Crop disease diagnostics is essential in addressing this problem, educating farmers about how to stop the spread of illnesses in their crops, and putting appropriate management in place. However, the advent of numerous crop-related illnesses impacts the agriculture sector’s production. Many techniques were developed to predict crop diseases early, but there are still issues of overfitting, less detection, and classification problems. To overcome these issues, design a novel Ant Lion-based Deep Belief Neural system for detecting and classifying crop diseases and enhancing agriculture’s growth. Initially, PlantVillage datasets were collected from the net source and trained in the system, and they were implemented in the MATLAB tool. Then, the noise and errors in the dataset were removed in the preprocessing phase, and the affected parts were segmented based on the pixels using the GrabCut algorithm. Additionally, feature extraction is employed using the Gray-Level Co-Occurrence Matrix, which extracts shape, texture, and color features. Finally, detect and classify the affected diseases in the crop using threshold values. The designed model can accurately predict crop diseases using ant lion fitness. The experimental results indicate the efficiency of the designed model by attaining better performance metrics, and the gained results are validated with other conventional models in terms of accuracy, precision, recall, F-score, AUC, and error rate.