Groundnut Crop Disease Detection Using Learning Rate Annealing Methods

Agricultural fields are vulnerable to diseases due to atmospheric changes, making early detection crucial for crop health and yield improvement. However, visual identification of crop diseases is challenging and there is a shortage of plant disease experts. To address this, advanced deep learning models are used for accurate disease detection. Despite various efforts using different models, results have been inconsistent, indicating room for improvement. This project focuses on groundnut crops, a significant cash crop in India. Groundnuts are prone to several diseases, impacting the agricultural economy. Here, a groundnut leaves dataset is used to develop and evaluate the disease detection model. After reviewing multiple studies, it is analyzed that InceptionResnetV2 outperforms the current models for crop disease detection, but there is still a scope for improvement. In order to address the issue of plant disease detection, this project aims to build a model making use of Learning Rate Annealing Techniques along with InceptionResnetV2 model to provide a best model that can produce greater results in disease detection of plants that can help the field of agriculture.