6533b826fe1ef96bd1285094

RESEARCH PRODUCT

Colour segmentation based on a light reflection model to locate citrus fruits for robotic harvesting

subject

description

Abstract Colour segmentation with a vision system is a good procedure to identify and locate fruits in robotic harvesting. Natural illumination conditions present in these environments produce a very variable illumination of the scene, in addition, fruits are usually partially occluded, and complete visual information about them is not available. The colour segmentation used for these purposes must take into account the appearance of highlights and shadows that natural illumination conditions produce. A method based on the Dichromatic Reflection Model for the light reflected from the surface object is reported here. Through the assumption of this model the light rays reflected from points of the object surface form a planar cluster in the colour space. A directional space of the vectors in the colour space is used as colour features to perform image segmentation regardless of the relative luminosity of lights with proportional power spectral distributions. By means of a training algorithm using colour sample points, the parameters used in colour segmentation are calculated. To locate the fruits, the regions from the segmented image labeled as fruit colour are taken as the fruit position in the image. The results of the application of the method reported here are given in terms of their performance for fruit detection, reaching 95% of visible fruits detected and 6% failure rate.