Cabbage disease detection system using k-NN algorithm
Identification of plant diseases is key to avoiding losses in agricultural yields and product quantities. Plant disease study means the study of disease patterns that can be visually seen on plants. The main objective of this research is to develop a prototype system with the help of machine learnin...
Saved in:
| Main Author: | |
|---|---|
| Format: | Academic Exercise |
| Language: | en en |
| Published: |
2022
|
| Subjects: | |
| Online Access: | https://eprints.ums.edu.my/id/eprint/33210/1/Cabbage%20Disease%20Detection%20System%20Using%20K-nn%20Algorithm.24pages.pdf https://eprints.ums.edu.my/id/eprint/33210/2/Cabbage%20Disease%20Detection%20System%20Using%20K-nn%20Algorithm.pdf https://eprints.ums.edu.my/id/eprint/33210/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Identification of plant diseases is key to avoiding losses in agricultural yields and product quantities. Plant disease study means the study of disease patterns that can be visually seen on plants. The main objective of this research is to develop a prototype system with the help of machine learning to detect cabbage diseases which are Alternaria Leaf Spot disease, Mosaic Virus disease, Downy Fungus disease, Bacterial Soft Rot disease, and Black Rot disease . It is very difficult to monitor plant diseases manually because it requires a large amount of work, deep expertise in plant diseases, and also requires excessive processing time. The image sample pixel will need to convert first using an otsu method and histogram method in the image processing and segmentation technique. Then, the segmented cabbage sample will use the GLCM method for feature extraction. It is a method of extracting second-order statistical texture features to detect diseases more efficiently. Finally, the KNN algorithm will be used to classify the disease based on sample nature and a cabbage disease data set. Consequently, by employing the KNN technique, the cabbage diseases are recognized at average 90% percent accuracy rates. This prototype has a very great potential to be further improved in the future. |
|---|