2002 Volume 64 Issue 5 Pages 123-133
The first major task of a fruit harvesting robot is the recognition of the fruit. This paper presents a color model suitable for the recognition of Fuji apples during harvest using a machine vision system under variable lighting conditions. Three color models; RGB model, rg-chromaticity method, and LRCD (Luminance and Red Color Difference) method, were tested to determine the thresholds for the segmentation of the apple fruit from the images. The decision-theoretic approach was applied to the three models to determine the thresholds. Results showed that the rg-chromaticity method was hardly influenced by the different lighting conditions and had the highest recognition rate and the lowest noise rate, specifically under the back lighting condition. Therefore, it was concluded that the rg-chromaticity method was suitable as one of the recognition methods of Fuji apples in a robotic harvesting operation.
