DeepLeafNet: Multiclass Classification of Soybean Plant Leaves with ResNet50V2 for Enhanced Crop Monitoring and Disease Detection

The soybean, which is a fundamental component of worldwide agriculture, plays a crucial role in meeting the global need for protein and oil. Nevertheless, the spread of many diseases poses a significant risk to both the quantity and quality of soybean harvests. The timely and accurate categorization of diseases is crucial for the successful implementation of crop management strategies. This study aims to enhance the categorization of soybean diseases by addressing ten different disease classes using the ResNet50V2 deep learning model. In order to accomplish this objective, the Soybean Disease Leaf Image Classification Dataset has been employed for both the purposes of training as well as evaluating the model. The ResNet50V2 model's architectural depth and inclusion of skip connections augment its capacity to grasp complicated patterns present in leaf images, hence bolstering its efficacy in performing robust categorization. The experimental findings provide evidence for the effectiveness of our methodology since the model attains a notable training accuracy of 97% and validation accuracy rate of 96% when diagnosing illnesses in soybean plant leaves with a minimum loss of 0.05 and 0.38. Moreover, this research endeavor makes a valuable contribution to the expanding domain of computer vision in the agricultural sector, facilitating the automation and enhancement of plant disease diagnosis processes.