PLANT DISEASE RECOGNITION AND CLUSTERING USING FUZZY ALGORITHM ON DATA MINING
| Autor: | R Sabitha, G Kiruthiga |
|---|---|
| Jazyk: | angličtina |
| Rok vydání: | 2021 |
| Předmět: | |
| Zdroj: | ICTACT Journal on Soft Computing, Vol 11, Iss 4, Pp 2429-2432 (2021) |
| Druh dokumentu: | article |
| ISSN: | 0976-6561 2229-6956 |
| DOI: | 10.21917/ijsc.2021.0346 |
| Popis: | Deep learning models are capable of performing sophisticated calculations, but they''''re not suitable for mobile and handheld devices due to their vast size and demanding computational requirements. Using an automated process, we intend to identify the diseases of plants so that we can build a process which begins with pre-processing, separates diseased leaf area, calculates features based on the Gray-Level Co-occurrence Matrix (GLCM), chooses and classifies features, and ends with decision making. Through threshold segmentation, we were able to isolate the diseased leaf areas in the maize plants, and then use this information to create fuzzy decision rules for the assignment of images of Common Rust to its severity class. These results were obtained with six colour and texture features. Plant disease clustering is performed with the Fuzzy Algorithm. The measurements show higher performance to the conventional methodologies and are ranked highest in terms of feature extraction method. This suggests that leaf-based plant disease diagnosis is the most appropriate method. These capabilities can be considered by the addition of new disease classifications or specific crop or disease classifications. |
| Databáze: | Directory of Open Access Journals |
| Externí odkaz: |
|