Transforming Crop Health Monitoring: Fine-Tuned InceptionV3 for Groundnut Disease Classification

The militant and interactive role of diseases in groundnut (Arachis hypogaea) consists of crop yield, quality, and food safety. Conventional processes of disease identification, mainly relying on eye exams, are often time-consuming and can induce misdiagnosis. This study provides an innovative method of groundnut disease classification using a fine-tuned InceptionV3 deep neural network. Through this deep learning (DL) technique, we purposed to build an automated system capable of differentiating between healthy, collar rot, afla rot, and wilt of those groundnut varieties using digital images. Our methodology covers the gathering and preprocessing of a large dataset of plant images of groundnuts, and then the further training of the InceptionV3 model to make it suitable for our classification problem. The model was rated on its accuracy, recall, precision, and F1 score across the disease categories. The model did a great job, as shown by the results where there was an average accuracy of 95.28%. This means the model is accurate in distinguishing between the different health states of groundnut plants. The results of this study reveal the power of appropriately designed DL models in agricultural disease detection, as well as the exact contribution of the fine-tuned InceptionV3 model to visual disease diagnosis. This study extends the rapidly developing application area of precision agriculture technologies and provides an innovational instrument for agronomists and farmers for early disease detection, and such a facility helps to develop timeous and better control strategies. More research is suggested to increase the data size test the scalability of the model and confirm its performance in different agricultural projects.