Application of image processing techniques in the characterization of plant leafs

Abstract

The number of applications using machine vision and digital image processing techniques in the agricultural sector is increasing rapidly. These applications include land/aerial remote sensing of crops, detection and recognition of pathological stress conditions, shape and color characterization of fruits, among many other topics. In fact, quantification of the visual properties of horticultural products and plants can play an important role to improve and automate agricultural management tasks. In this paper, is described a plant leaf characterization system based on a personal computer. This system uses a digital scanner to acquire leaf images with a resolution of 150 dpi. These images are afterwards processed in order to compute some leaf characteristic parameters, such as: leaf area and perimeter, existence of holes, width and length. With the implemented algorithms the errors between the measurements and the real values were typically less than /spl mnplus/3% and /spl mnplus/2.5% for the area and linear measurements, respectively. These tests and results were realized using sets of known size images and leaf images that were measured with the proposed system and with a commercial calibrated leaf area system LiCor from Delta-T devices.