Color recognition

Color provides a wealth of information about an image. Color reduction and color feature extraction can be used to compress this information into a manageable image characteristic. These simple image-processing techniques can be used to quickly describe images while providing enough information to allow image classification and identification. This paper investigates three possible applications of color recognition: flag identification, stamp identification, and landscape classification. The results from these three examples show that color reduction can provide a useful tool for image understanding. In addition, the low processor cost of this system allows it to be easily incorporated into larger systems.