An Efficient Deep Learning-based VGG19 Approach for Rice Leaf Disease Classification

As the staple crop and commodity par excellence, rice deserves close scrutiny throughout its lifetime from production through consumption. This is especially true in countries like India. A major threat to the health of rice plants in general as well as their yield is that they are susceptible to a variety of diseases. This is a big step forward in rice leaf disease classification. The use of the an efficient and fine-tuned VGG19 model helps to achieve an extremely high accuracy rate (over 97.5%). This highlights the capabilities of modern deep learning methods, an important breakthrough for the agriculture field. There was a very complex equation in the relationship between epochs and model performance, as probed by careful exploration. Despite this, an epoch of 28 showed the highest efficiency and a set of 7 displayed suboptimal results. At epoch 28 the proposed approach reached a testing loss of 0.0253, verifying its adeptness in distinguishing healthy from diseased rice leaves. This research is based on an integrated theory of image analysis that includes the different conditions and stages in the development of rice plants. Taking a holistic view brings rice leaf health conditions into focus. In conclusion, this study is an important step forward in the field of rice leaf disease classification with state-of-the-art deep learning techniques that greatly improve accuracy.