Automatic Prediction and Classification of Diseases in Melons using Stacked RNN based Deep Learning Model

Abstract

Several studies have investigated in deep learning classification techniques to predict and classify the diseases by processing and segmenting the leaf images. This paper exhibits the correlations between Watermelon yield and proposed disease prediction indices in the Indian largest Watermelon-exporting form. Only Disease pixels with 100 <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">%</sup> Watermelon plant were included in the analysis process. The prediction and identification stages includes image acquisition, image processing, image segmentation are used. Here Image processing technique utilized to distinguish the illness in melon plants. After the segmentation is completed, K-means clustering is used in which one of the clusters contains the diseased spots being extracted. The Stacked RNN is applied for classification of diseases in melons. This prediction and classification of leaf diseases gives the top performance with high accuracy and computational efficiency compared with other existing models. This prediction and classification of disease is implemented in the TensorFlow and Keras framework. The performance is estimated regarding the accompanying measurements: exactness, review and f-measure. The optimization algorithm used in the classification of disease is Adam and the activation used is ReLu and SoftMax.