MinorII-Rice-Leaf-Disease

Identification of diseases from the images of a plant is one of the interesting research areas in the agriculture field, for which machine learning concepts of the computer field can be applied. Rice is a widely cultivated economical crop in the world. During cultivation the earliest and accurate diagnosis of the rice plant diseases is able to reduce the damage, resulting in environmental protection and better return.

In traditional practices, identification is performed either by visual observation or by testing in a laboratory. The visual observation requires expertise and it may vary subject to an individual which may lead to an error while the laboratory test is time consuming and may not be able to provide the results in time.

To overcome these issues, an image based machine learning approach to detect and classify plant diseases has been presented.

Steps done to do so -

1. Image Acquisition
   
2. Image Pre-Processing
   i) Background Removal
   ii) Sharpening/Blur Removal (optional)
   
3. Image Segmentation
   i) Disease Portion Segmentation
   ii) Leftover Green Region Removal
   
4. Feature Extraction
   i) Color Feature Extraction
   ii) Texture Feature Extraction
   
5. Classification
   i) SVM
   ii) Decision Tree