Detection and classification technique of Yellow Vein Mosaic Virus disease in okra leaf images using leaf vein extraction and Naive Bayesian classifier

Okra (Abelmoschus esculentus (L) Moench), is widely grown all over tropical, subtropical and warm temperature regions of the world. It is a popular crop in India due to its ease of cultivation and adaptability to varying moisture conditions. But the crop is prone to damage by various diseases caused by various insects, fungi, nematodes and viruses. The most common disease of okra is Yellow Vein Mosaic Virus (YVMV), spread by white fly (Bemisiatabaci). This paper presents an efficient technique to detect and classify the presence of YVMV disease in okra leaf with the joint use of image processing, K-means and Naive Bayesian classifier. The proposed technique is experimented on 79 standard diseased and non-diseased okra leaf images. The input leaf images are of four classes, namely Highly Susceptible (HS), Moderately Susceptible (MS), Tolerable (T) and Resistive (R), depending upon the severity of the YVMV infection. The proposed technique achieves 87% success rate using 10 features only.