Abstract: In this paper, an image content based perceptual watermarking model using ridgelet transform is presented. It provides a new visual model, which can exactly evaluate the Just Noticeable Distortion (JND) tolerance of human visual system. The watermarking algorithm is designed to address the issue of how to improve the robustness and the security of the embedded watermark according to the image content. In order to improve the robustness and transparency, the watermark bits are inserted in the visual significant ridgelet coefficients without exceeding the maximum strength of JND. With the spread spectrum technique and the secret key protection, the security of the embedded watermark is also guaranteed very well. Through analyzing the statistical property of the watermarked image, a threshold selection method based on Neymann-Pearson principle is presented and then the embedded watermark can be blindly detected by correlation coefficients between the original image and the watermark sequence. Experiments show that the proposed method can achieve a better tradeoff between the robustness and transparency.