A MRA-based Color Transformation Scheme between Objects

Object´s colors are frequently changed for enhancing some potential features of the object for wide applications in different fields. In this paper, multiple regression analysis (MRA) is respectively conducted on the R, G, and B planes of the target and source objects to find the corresponding best fitting function of each plane. Where, target and source objects are extracted from target and source images with employing color objects extraction algorithm using multiple thresholds (COEMT), respectively. The new values of each pixel of target objects are then evaluated by using the best fitting functions. These new values are then combined into the result RGB recolored objects. The experimental results show that the presented approach has three major advantages: (i) the proposed algorithm is simple, effective and accurate in transferring color between objects without any change in the object details, (ii) the proposed algorithm is time saving, (iii) there are no restrictions in the dynamic ranges of colors of objects.