Spectral band selection for color machine vision used for plant identification

摘要

Abstract: hrome video camera and a set of narrow bandpass optical filters were used to capture images in 13 spectral bands from 400 nm to 1000 nm. Images of 4 plant species were captured. Pixel data was extracted from the images in user defined areas on the plant leaves. Attempts were made to classify the plants based on the spectral information in the extracted data. !7