EigenHistograms: Using Low Dimensional Models of Color Distribution for Real Time Object Recognition

Abstract

Distribution of object colors has been used in computer vision for recognition and indexing. Most of the recent approaches to this problem have been focused on defining optimal spaces for representing pixel values that are related to physical models and that present some invariance. We propose a new approach to identify individual object color distributions by using statistical learning techniques and to allow their compact representation in low dimensional spaces. This approach outperforms generic “optimal” spaces when color illumination is constant, allowing changes in object pose and illumination direction. This approach has been tested for real time industrial inspection of multicolored objects.