Detection of The Erosion Area of Lotus Leaf Using Hyperspectral Imaging

Hyperspectral imaging (HSI) has become one of the representative imaging technologies in the 21st century. It is widely used in many fields such as geography, ecology, chemistry, medical imaging, and military defense, and so on. In this paper, we utilize HSI technology to detect the insect erosion areas of lotus leaf which are not easy to be inspected by human eyes. We collect the HSI photos of lotus leaf by with a 470-900nm hyperspectral camera. Given limited amounts of spectral samples of erosion, we use a detection algorithm to find out all areas of insect erosion in the leaf. The experiments show that the spectra of erosion area are significantly different from those of normal mesophyll tissue in NIR spectrum, and they could be detected and recognized by using the algorithm. This discovery will help develop prevention strategies and classification technology of insect hazards in the future.