Identification of Diseases in Grapes Using Gray Level Co-Occurrence Matrix & Wavelet Statistical Features

Grapes are a crop that is susceptible to many diseases. However, the degree of susceptibility varies depending on the variety. When no pest management is carried out, damage can generally be severe. Downy mildew and powdery mildew are the major grape diseases in India. Evidently, these diseases can be easily predicted based on the climatic conditions determined by agricultural experts. Technological strategies using machine vision and artificial intelligence are being investigated to achieve intelligent farming forbetter yield. As a part of the prediction process in Grapes, this paper initially deals with the identification of type of disease that has occurred in a grape vine, with a special focus on its leaves. The first step in an effective pest management program is correct identification of the disease. This paper uses GLCM (Gray Level Co-occurrence Matrix) and Wavelet statistical Features to determine whether a given grape leaf is affected with Powdery Mildew or Downy Mildew by comparing the statistical features with that of an unaffected leaf. The developed algorithm's efficiency can successfully detect and classify the examined diseases with a precision of 94%.