ABSTRACT: The presence of several types of fuzzy inference system allows us to control the expected results of a problem that to be resolved. Image processing is the most interesting science that concern with fuzzy theory and linear algebra. In this paper, the Sugeno Fuzzy Inference System is adopted with the QR matrix decomposition method to propose a watermarking algorithm using the parameter generated by FIS to control the choice of the location to embed the information of the watermark into the R matrix obtained from the QR matrix decomposition method. Some attacks are applied to test the robustness represented by NC and the imperceptibility represented by PSNR of the proposed scheme.

Keywords –Sugeno fuzzy Inference System, QR matrix decomposition,watermarked image, the Peak Signal to Noise Ratio(PSNR), normalized correlation(NC).