An Efficient Approach of Color Image Matching by Combining Color Invariant and ORB Feature

Abstract

Through combine traditional prints with portable electronic devices by image matching algorithms, the traditional printing information can be expressed in a diverse way. During the process of image matching, color is an important component for distinction between images. Nevertheless, most of the existing approaches only use gray geometric-based feature extractors to realize image matching, such as ORB, SIFT, SURF, etc. For these approaches, neglecting color information may lead to poor illumination robustness and mismatching. In this paper, we present a new approach based on color invariant and ORB feature, called C-ORB. By combining color invariant and ORB feature, C-ORB could be more accurate, at the same time; C-ORB also retains the ORB’s efficiency and performance. To validate C-ORB, we perform experiments that test the properties of C-ORB relative to ORB. Experimental results show that C-ORB is more accurate and stable with respect to variations in the photometrical imaging conditions. Therefore, the research of this paper can improve the accuracy and reduce the error during the process of image matching in mobile electronic equipment.