An Image Watermarking Hiding Algorithm Based On The Maximum Entropy And Characteristics In Shearlet Transform Domain

Shearlet transform is a transform which considers all characteristics of the optimal image representation method. It can make full use of the image's geometrical characteristics to track the directional information of image. Compared with wavelet, shearlet transform can be more efficiently at analyzing edge, contour, texture and other significant details of image. These details are very critical for processing image. Therefore, shearlet transform has been widely applied in image processing field. The development of shearlet transform provides a new approach for the development of image watermarking. Considered superior characteristics of shearlet transform, the purpose of this paper is mainly to apply Fast Finite Shearlet Transform and Extended-Discrete Shearlet Transform in image watermark. The major content of this paper include:1. Study about the characteristics in Extended-Discrete Shearlet Transform domain and Fast Finite Shearlet Transform domain. Based on the theory of shearlet transform, the paper focused on analyzing characteristics in Extended-Discrete Shearlet Transform domain and Fast Finite Shearlet Transform domain, which provided a theoretical basis for applying Fast Finite Shearlet Transform and Extended-Discrete Shearlet Transform to image watermark.2. Based on the characteristics in Fast Finite Shearlet Transform domain, an image watermark algorithm based on the maximum entropy and the characteristics in Fast Finite Shearlet Transform was proposed. The proposed algorithm chosen watermark's hiding location by generalized Arnold Transform, and decided hiding strength by Noise Visibility Function model, and hided watermark information by Spread Spectrum algorithm. The experimental results showed the proposed algorithm obtained a good performance.3. Based on the distribution characters of shearlet coefficients in Extended-Discrete Shearlet Transform domain, an image watermark algorithm based on the maximum entropy and characteristics in Extended Discrete Shearlet Transform domain was proposed. The algorithm combined Mean Shift algorithm and the description of image texture based on Gray-Level Co-occurrence Matrix to choose watermark's hiding location and strength of. The experimental results showed that the proposed algorithm had good performance.