Citrus Diseases Recognition Using Deep Improved Genetic Algorithm

Agriculture is the backbone of each country, and almost 50% population directly involved in farming. In Pakistan, several kinds fruits are produced exported other countries. Citrus an important fruit, its production Pakistan higher than fruits. However, diseases citrus such as canker, scab, blight, a few more impact quality quantity this Fruit. The manual diagnosis these required expert person who always time-consuming costly procedure. agriculture sector, deep learning showing significant success last five years. This research work proposes automated framework using best feature selection for classification. proposed framework, augmentation technique applied initially by creating training data from existing samples. They were then modifying two pre-trained models named Resnet18 Inception V3. modified trained augmented dataset through transfer learning. Features extracted model, which further selected Improved Genetic Algorithm (ImGA). features both fused array-based approach that finally classified supervised classifiers Support Vector Machine (SVM) name more. experimental process conducted on three different datasets-Citrus Hybrid, Leaf, Fruits. On datasets, best-achieved accuracy 99.5%, 94%, 97.7%, respectively. evaluated step compared with some recent techniques, method shows improved performance.