DWT域的图像哈希算法

AN IMAGE HASHING METHOD BASED ON DWT

图像哈希是根据人类视觉系统的特点,以图像内容为基础,从数字图像数据中提炼出剪短的二进制代码,映射成固定的数字串,可用于反映图像的重要视觉特征.通过对现有图像哈希算法的调查研究,提出一种新颖的哈希提取算法.该算法了结合图像离散小波变换(DWT,Discrete Wavelet Transform)系数和环形分割算法计算出图像特征向量,最后量化压缩编码成图像哈希序列.实验结果表明该算法具有良好的鲁棒性,且优于对比算法.

Image hashing is the process of encoding digital image data into compact binary codes. It reflects important vision characteristic of image. A novel image hashing method is presented in this paper through the investigation of the existing image hashing algorithms. The proposed algorithm extracted image feature vectors by combining DWT - coefficients and ring - based image division method, then sampled, compressed and encoded into an image hashing sequence. The experiment demonstrates the effectiveness of the proposed algorithm and shows that the method is robust to various types of attacks. At the same time the algorithm is superior to the contrast algorithms.