학술논문

Coffee Leaf Diseases Classification: A CNN and Random Forest Approach for Precision Diagnosis
Document Type
Conference
Source
2024 International Conference on Automation and Computation (AUTOCOM) Automation and Computation (AUTOCOM), 2024 International Conference on. :210-214 Mar, 2024
Subject
Communication, Networking and Broadcast Technologies
Components, Circuits, Devices and Systems
Computing and Processing
Engineering Profession
Power, Energy and Industry Applications
Signal Processing and Analysis
Productivity
Deep learning
Damping
Training
Feature extraction
Magnetic heads
Random forests
Coffee leaf diseases
CNN model
Random Forest
Machine learning
Language
Abstract
The study explores the varieties of coffee leaf diseases which involves the collection of data about different kinds of coffee leaf diseases along with their traits and characteristics, and also their cure and prevention requirements. The different types of coffee leaf diseases include coffee leaf rust, brown eye spot, damping off, black rot, berry blotch, anthracnose, and root diseases. To avoid all these diseases, it is important to know about these diseases. To understand its causes, the paper includes different classes i.e. pooling layers, and other effective elements. The study includes a total of 5747 pictures. To estimate the model production, numerous matrices are used i.e. precision, recall, F1-score, support proportion, accuracy, and macro, weighted, and micro average. Based on these matrices, the productivity of this model is calculated as an overall accuracy of 77.00% and an average weighted F1-score of 77.00%. It also includes the highest precision of 84.63% and lowest of 68.11% and also the highest recall of 79.90% and the lowest is 74.66%. The model will be helpful for the early detection of diseases which will help us to act earlier on the disease before the disease will affect the plant itself as well as nearby plants present over there. It will better help in enhancement of the productivity.