Enhanced Rice Leaf Disease Detection using Densenet201 for Improved Classification Accuracy and Reduced Misclassification Error

Rice is a vital staple grain grown in numerous countries. Rice leaf diseases can result in substantial harm to crop cultivation, resulting in diminished yields and financial losses. Conventional methods for detecting diseases are typically lengthy, entail a lot of effort, and necessitate specialized knowledge. Automated leaf disease detection methods assist farmers in identifying illnesses with minimal or no human intervention. Prior research about this detection of diseases in rice leaves mostly relied on the utilization of image processing and ML methodologies. The agricultural sector extensively necessitates automated identification and examination of rice ailments to prevent the squandering of financial and other resources, minimize crop yield reduction, enhance processing effectiveness, and achieve bountiful and disease-free harvests. The advancements achieved in deep learning methodologies have had a substantial influence on the field of agricultural disease diagnoses. The occurrence of leaf diseases in Rice Plants can significantly reduce crop productivity, making it crucial to accurately diagnose these diseases to prevent their negative impact. The majority of the current approaches are insufficient in detecting leaf images and accurately identifying their condition. This study introduces the DenseNet201 technique for automatic leaf detection, which offers improved efficiency compared to existing methods. The new approach guarantees a 98 % accuracy rate and a misclassification error value of 0.0167, highlighting the significance of the suggested method compared to the existing one.