Detection of Age and Defect of Grapevine Leaves Using Hyper Spectral Imaging

Abstract

This paper demonstrates the potential of using hyperspectral imaging for detecting age and defects of grapevine leaves. For age detection studies a number of grapevine healthy leaves and for defect detection analysis 9 different defective leaves have been selected. Hyperspectral images of these leaves covered spectral wavelengths from 380 nm to 1000 nm. A number of features from the brightness in ultra violet (UV), visible (VIS) and near infrared (NIR) regions were derived to obtain the correlation of age and defect. From the experimental studies it has been observed that the mean brightness in terms of original reflection in visible range has correlation with different ages of grapevine leaves. Moreover, the position of mean 1st derivative brightness peak in VIS region, variation index of brightness and rate of change of brightness i.e., mean 1st derivative brightness at NIR provide a good indication about the age of the leaves. For defect detection whole area and selective areas containing the defects on the leaves have been experimentally analysed to determine which option provides better defect detection. Variation index of brightness was also employed as a guide to obtain information to distinguish healthy and unhealthy leaves using hyperspectral imaging. The experimental results demonstrated that hyperspectral imaging has excellent potential as a non-destructive as well as a non-contact method to detect age and defects of grapevine leaves.