Gaussian Distributive Czekanowskis Region-based Deming Regression for Plant Leaf Disease Identification

  • Mohammed Zabeeulla A N, Dr. Chandrasekar Shastry

Abstract

Automatic identification of plant leaf disease is asignificant part to predict the infected region. Early disease detection is essential to achieve higher crop yield and production in the agriculture field. Most of the developed processing scheme performs the leaf disease identification but it consumes more processing time. In order to improve the accuracy of disease identification, a novel method called, Gaussian Distributive Czekanowskis Region-based Deming Regression (GDCR-DM) is presented. The GDCR-DM uses the deep Multilayer feedforward Perceptive neural learning concept comprises many layers to learn the given input plant images and provides accurate disease identification accuracy. The deep neural learning uses four different processing steps in the different layers. Initially, the image acquisition process is carried out to collect the multiple plant leaf images from the dataset. Then the collected image is sent to the input layer of deep neural network. The input image is preprocessed in the hidden layer to smoothen the image using the Gaussian distributive trilateral filtering technique to enhance the image quality for accurate plant leaf disease identification. Then the Czekanowski's dice indexive region-based segmentation technique is applied to find the region of interest part fromthe plantleaf image in the second hidden layer. In the third hidden layer, the feature extraction process is performed usingthe Deming regression function to extract the shape, texture, and color features from the input segmented images. Finally, the extracted features are analyzed to classify the input leaf image into the normal or disease with the help of the activation function. The proposed GDCR-DM classifier accurately classifies the given input image and minimizes the error rate.Experimental evaluation is performed using a plant leaf image dataset with different parameters such as peak signal to noise ratio, Disease identification accuracy on an average by 88%, false-positive rate, and computation time with respect to the number of plant leaf images. The observed qualitative and quantitative results illustrate our proposed GDCR-DM attains superior performance than the state-of-the-art methods.

Published
2021-12-02
How to Cite
Mohammed Zabeeulla A N, Dr. Chandrasekar Shastry. (2021). Gaussian Distributive Czekanowskis Region-based Deming Regression for Plant Leaf Disease Identification. Design Engineering, 1581- 1601. Retrieved from http://www.thedesignengineering.com/index.php/DE/article/view/7113
Section
Articles