Abstract

Since India's economy is predominately agricultural, timely diagnosis of plant diseases is essential to reducing financial losses. Millions of rupees are annually spent to protect crops from a variety of diseases, but this is only possible because of the antiquated techniques for identifying plant diseases. Plant disease detection by humans is, at best, unreliable. Even with the assistance of specialists in plant diseases, the time, effort, and expertise needed to detect the particular sickness, there is no assurance that the results will be accurate. Image processing and machine learning are two such technologies that have demonstrated efficacy in this area. We provide a use case of machine learning for a specific issue in plant disease diagnosis using camera-collected plant photos.

Description

“Open CV,” “SVM,” “feature extraction,” “segmentation.”