Improved fuzzy c-means clustering for grape leaf disease detection using mathematical morphology

Name: Aphichet Chaimanee

Organization : Mahasarakham University. Faculty of Engineering

Name: Fu, Xiaoli

Organization : Xiamen Institute of Technology. Faculty of Architecture

Name: Lyu, Guifen

Organization : Ningxia College of Construction

Name: Niwat Angkawisittpan

Organization : Mahasarakham University. Faculty of Engineering

Email : niwat.a@msu.ac.th

Name: Nuttapon Chaiduangsri

Organization : Mahasarakham University. Faculty of Engineering

Name: Sivarit Sultornsanee

Organization : Northeastern University. College of Engineering

keyword: Fuzzy C Means

ThaSH: Grapes -- Morphology

ThaSH: Anthracnose

ThaSH: Downy mildew diseases

Abstract: Grapevines may show various symptoms that negatively affect the vigor and production of the grapes. In this study, the classification of diseases was performed by identifying the distinctive features of grape leaf lesions on digital photographs to facilitate treatment and control of the symptoms of the disease from an early stage. A set of 5,700 images of grape leaves consisting of 1,500 images of downy mildew-infected grape leaves, 1,400 images of rust-infected grape leaves, 1,600 images of Anthracnose-infected leaves and 1,200 images of healthy leaves was used. Pre-processing procedures included 1) Image Sharpening using Laplacian Filters to enhance or improve the details in the images by reducing blur and enhancing the edges. 2) Gray Scale Image to convert the image into a white and black gradient (Halftone Image). 3) Noise Reduction using Smoothing Spatial Filters to reduce the noise based on blurring effect such as reducing fine details or connecting gaps in straight lines or curves. 4) Fuzzy C-means Clustering (FCMC), an algorithm that allows overlapping or redundancy of data in each cluster through the determination of membership degree of each cluster. This clustering method is widely used in fields such as medicine, science and engineering. 5) Mathematical Morphology, the comparison of characteristics and appearance of grape leaf lesions. The results showed that, from 5,700 images, the average accuracy of disease classification for downy mildew, rust, anthracnose and healthy leaves was 96.7742 percent, 97.9871 percent, 95.3877 percent and 96.7742 percent, respectively. The results obtained can be used for further development and applications in agriculture.

King Mongkut's University of Technology North Bangkok. Central Library

Address: BANGKOK

Email: library@kmutnb.ac.th

Created: 2023

Modified: 2024-12-12

Issued: 2024-12-12

บทความ/Article

application/pdf

BibliograpyCitation : In Electrical Engineering Academic Association (Thailand), Mahasarakham University. Faculty of Engineering and ASEFA. The 2023 International Electrical Engineering Congress (iEECON 2023) (pp.391-394). Mahasarakham : Mahasarakham University

eng

©copyrights King Mongkut's University of Technology North Bangkok

ลำดับที่.	ชื่อแฟ้มข้อมูล	ขนาดแฟ้มข้อมูล	จำนวนเข้าถึง	วัน-เวลาเข้าถึงล่าสุด
1	iEECON 2023pp.391-394.pdf	345.47 KB

ใช้เวลา

0.035843 วินาที