Computer Science & Electrical

Computer Science & Electrical

Archive
Join as an Editor/Reviewer

Enhancing Rice Disease Binary Classification. An Analysis of Deep Neural Network Models

Volume: 147  ,  Issue: 1 , April    Published Date: 16 April 2024
Publisher Name: IJRP
Views: 324  ,  Download: 102 , Pages: 10 - 18    
DOI: 10.47119/IJRP1001471420246284

Authors

# Author Name
1 Yubraj Bhattarai
2 Kumar Prasun
3 Gajendra Sharma
4 Prajwal Rai

Abstract

The study examines deep neural network models for the binary classification of rice plants, with a specific emphasis on distinguishing between healthy and ill states. The utilization of convolutional neural network designs such as VGG16, Inception V3, ResNet18, and MobileNet is commonly observed. The study gathered a sample of 500 photographs depicting both good and unhealthy conditions in Kathmandu and Sindhupalchowk. The objective was to evaluate the efficacy of the models in generalizing the health states of rice. The results indicated that model Resnet exhibited superior performance with a high level of accuracy, whereas model VGG and Inception had lower accuracy. The model exhibits promising potential in influencing the diagnosis and early detection of rice diseases.

Keywords

  • Rice Disease Classification
  • Deep Neural Network
  • Image-based Disease Detection