In this paper, an efficient watermarking algorithm for color image is proposed based on HSI color space and Discrete Wavelet Transform (DWT). The color image is firstly converted from RGB to HSI format to obtain the intensity component. Then sub-blocks of intensity are divided and sorted according to the human visual system (HVS) characteristic. Finally, the scrambled binary watermark, which is achieved by Arnold transformation, is embedded by modifying the low frequency DWT coefficients of the selected sub-blocks. A series of experiments are carried out to test the effectiveness of the proposed algorithm, and experimental results show that the scheme is feasible and robust against JPEG compression and different image processing operation.