Description

We propose a new color FR method. The newly introduced method takes advantage of “boosting” learning as a feature selection mechanism, aiming to find the optimal set of color-component features for the purpose of achieving the best FR result. We have taken the first attempt to incorporate feature selection scheme underpinning boosting learning into FR methods using color information. The proposed color FR method which largely consists of two parts: 1) color-component feature selection with boosting, and 2) color FR solution using selected color component features. To determine the best color component feature at each boosting round for recognizing the hard-to-classify sample subset of a training set, termed “learning set,” the effective selection criterion is proposed.