Apple and Tomato Leaves Disease Detection using Emperor Penguins Optimizer based CNN

• Main Authors: Suguna, R, C N S, Vinoth Kumar, Deepa, S, M. S., Arunkumar
• Format: Conference Proceeding
• Language: English
• Subjects: Agriculture; Apple Leaves; Computational modeling; Convolutional Neural Network; Crops; Disease Detection; Emperor Penguin Colony Optimizer; Feature extraction; Image segmentation; Plants (biology); Productivity; Tomato Leaves
• ISSN: 2575-7288
• DOI: 10.1109/ICACCS57279.2023.10112941

The economy of the country is heavily dependent on agricultural productivity. The largest number of the crop yield trust on growth of the healthy plants. Nearly all the times the crop yields are heavily affected by leaves diseases. The vital role of the agriculture field is to discover the plant's leaves diseases in a timely premature point. An automatic disease detection technique is mandatory in agriculture field to identify and determine the kind of plant leave diseases in an earlier stage and give the right pesticides. The healthy plants are the major effect on quality and productivity of the agriculture products. An image segmentation and classification system is presented in this study to recognize the plant leaf diseases automatically. In this work, YOLO network of version 4 is used for object spotting (detection) and segmentation for detecting the affected leaves such as Apple and Tomato Leaves. Here, Convolutional Neural Network (CNN) acts as both feature extraction and also classifier, i.e. the color features and statistical attributes are bringing out using CNN. Finally, the learning rate of CNN is optimized by using Emperor Penguin Colony Optimizer (EPC). The prominent aspects of EPO is producing and computing the huddle boundary, distance, temperature, and effective mover all at the single point time. The experiments are done on publicly available datasets to analyze the performance of proposed algorithm with existing techniques.