Related Experiment Video
Updated: Jul 11, 2025
09:47
Author Spotlight: Advancing Alzheimer's Research – Exploring Early Detection and Multi-Omics Approaches
Published on: December 15, 2023
1.1K
Plant disease identification using contextual mask auto-encoder optimized with dynamic differential annealed
1Research Scholar, Department of Computer Science and Engineering, University College of Engineering (BIT Campus), Anna University, Tiruchirappalli, Tamil Nadu, India.
Microscopy Research and Technique
|November 3, 2023
Summary
This study introduces a new AI method for early plant disease detection. The proposed PDI-CMAE-DDAOA model significantly improves accuracy and sensitivity in identifying diseased plants compared to existing approaches.
Area of Science:
- Agricultural Science
- Computer Vision
- Artificial Intelligence
Background:
- Plant diseases significantly reduce global food production, necessitating efficient monitoring.
- Manual plant disease identification is time-consuming and prone to errors.
- Early detection using AI and computer vision can mitigate disease impact and improve crop yields.
Purpose of the Study:
- To propose an advanced AI model, PDI-CMAE-DDAOA, for accurate plant disease identification.
- To enhance the classification performance of the contextual mask auto-encoder (CMAE) using dynamic differential annealed optimization algorithm (DDAOA).
- To achieve early and reliable detection of plant diseases for improved agricultural management.
Main Methods:
- Utilized the Plant Village dataset for image collection and preprocessing with an adaptive self-guided filter.
- Extracted statistical features (mean, variance, entropy, kurtosis) using the cosine similarity hidden Markov model (CSHMM).
- Employed a contextual mask auto-encoder (CMAE) optimized by DDAOA for classifying healthy and diseased plant regions.
Main Results:
- The PDI-CMAE-DDAOA model demonstrated superior performance metrics, including accuracy, precision, sensitivity, F1-score, and specificity, compared to existing methods (PDI-DENN, PDI-CAE-CNN, PDI-EN-CNN).
- Achieved higher accuracy (e.g., 23.34% improvement), sensitivity (e.g., 36.67% improvement), F1-score (e.g., 46.67% improvement), and specificity (e.g., 56.67% improvement) over baseline models.
- The DDAOA optimization enhanced CMAE convergence speed and improved classification precision, minimizing classification errors.
Conclusions:
- The proposed PDI-CMAE-DDAOA offers a robust and efficient solution for early plant disease detection.
- The integration of CSHMM for feature extraction and DDAOA for CMAE optimization significantly boosts classification accuracy.
- This AI-driven approach holds substantial potential for revolutionizing agricultural practices and ensuring food security.
