Abstract: Copyrights is considered the most important issue that the most of the jewelry designers and jewelry companies worries about and any other art designers, each designer have his own style and local features that is built in the jewels, each feature is unique and cannot be faked by global brands counterfeit. This study propose a jewels features recognition method, Speeded up Robust Feature (SURF) algorithm or Scale Invariant Feature Transform (SIFT) are used for allowing the proper detection of watermarks, even in the case where the features in image have been partially altered or rotated. For the recognition of the chosen object area after geometrical distortions, jewels features using SURF or SIFT are estimated and stored prior to the process for usage throughout the procedure of detection. In the step of detection, the SURF/SIFT features of the jewels image are obtained then they undergo matching against the stored images. From the results of matching, SURF/SIFT properties are utilized for computing the affine-transformation parameters and the object area is then recovered.