Abstract: In the past, detecting fruit diseases relied on human visual inspection, which was often unreliable due to subjective judgment and limitations in detecting microorganisms. This approach was time-consuming, costly, and less accurate. However, using MATLAB-based approaches for quick and accurate diagnosis is a better choice compared to outdated methods. Symptoms of infection or disease can manifest on fruits, leaves, and lesions of plants, and this project aims to accurately diagnose the condition based on submitted images through image segmentation, preprocessing, feature extraction, and labeling. Various factors such as insect transmission, weather, and environmental conditions can cause infectious diseases in fruits, caused by viruses, fungi, or bacteria. The project will focus on identifying the cause of contamination in fruits to determine the type of infection, by extracting major and minor axes of fruit characteristics from images for effective classification.
Keywords: K-Means Clustering, Local Binary Pattern, Multi-class Support Vector Machine, Texture Classification
|
DOI:
10.17148/IJARCCE.2023.12490
