Cropsaver: Next Generation Pest Management Using Improvised Squeeze Net Model And Support Vector Machine

Crop pests are a major problem for farmers in agriculture, as they can significantly reduce both yield and quality, thereby affecting global food security and economic stability. In this study, we proposed a solution in the form of a deep neural network for pest detection and classification, utilizing the improved Squeeze Net architecture using transfer learning and abstracting high level features by image processing. Pesticide recommendation is done by SVM algorithm. By leveraging a pre-trained Squeeze Net model on a large dataset and fine-tuning it on our pest-specific dataset, we can significantly reduce the computational resources and time required for training, while still maintaining the model's exceptional feature recognition capabilities. The model's performance has been evaluated using accuracy, precision, recall, and F1-score to determine its proficiency in distinguishing the class of pests. This proposed model shows an improved SqueezeNet model of better accuracy for pest detection and classification.