Mobile application using DCDM and cloud-based automatic plant disease detection

Area of Science:

• Agricultural Science
• Computer Science
• Data Science

Background:

• Plant diseases cause significant annual losses to global agricultural produce, estimated at over 18%.
• Traditional methods for plant disease detection are often inefficient, uncertain, and labor-intensive.
• There is a critical need for advanced, automated techniques to monitor plant health and identify diseases early.

Purpose of the Study:

• To develop and evaluate an automated farm-based module for accurate plant disease diagnosis.
• To leverage cloud computing and deep learning for real-time monitoring and management of farm data.
• To improve agricultural productivity and reduce crop losses through early disease identification.

Main Methods:

• Implementation of a farm-based module with cloud data centers and data conversion devices.
• Utilizing mobile cameras and bots for image acquisition of plant health status.
• Development of a DCDM deep learning model trained on 40,000 images, with analysis of 10,000 images for disease classification.

Main Results:

• Real-time diagnosis of plant leaf diseases achieved 98.78% accuracy in a laboratory setting.
• A cloud-based image diagnostic and classification service provided results in an average of 0.349 seconds.
• The system successfully recorded plant name, disease type, and images in a cloud database.

Conclusions:

• The proposed automated system significantly enhances the accuracy and speed of plant disease detection.
• Cloud-based image analysis and deep learning models offer a scalable solution for agricultural monitoring.
• This technology has the potential to boost productivity in the agricultural and irrigation sectors by minimizing disease-related losses.