Machine Learning Based Automatic Leaf Diseases Detection

Abstract

The method for applying machine learning to automatically detect leaf diseases is presented in this paper. A convolutional neural network was used to extract pertinent features from leaf image datasets that included healthy and diseased leaves. The dataset was compiled and pre-processed. Accuracy, precision, and recall measures were used to assess the machine learning algorithm after it had been trained on the labeled dataset. According to the findings, the algorithm was very precise and recallable in its ability to detect leaf illnesses, making it a potential method for practical use. This strategy may help with early leaf disease identification and prevention, increasing crop productivity and lowering the demand for toxic pesticides. Here we are identifying the Bacterial spot, Early blight.