基于信息熵的隐私保护图像检索方案

PRIVACY PROTECTION IMAGE RETRIEVAL SCHEME BASED ON INFORMATION ENTROPY

  • 摘要: 针对现有图像检索方案无法对用户隐私进行有效保护且在实际应用中效率低、检索精确度差的问题,设计一种基于信息熵和局部敏感哈希的两层检索方案。利用图像信息熵的特性筛选出更具代表性的图像特征;使用高效的同态加密算法加密特征;采用基于信息熵的混沌加密算法保护图像安全。安全分析和实验表明其在实现有效保护用户隐私的同时,性能也达到了较好的实用水准。

     

    Abstract: Aimed at the problems that the existing image retrieval schemes cannot effectively protect users' privacy, low efficiency and poor retrieval accuracy in practical application, a two-layer retrieval scheme based on information entropy and local sensitive Hash is designed. More representative image features were selected by using the characteristics of image information entropy. Efficient homomorphic encryption algorithm was used to encrypt features. Chaotic encryption algorithm based on information entropy was used to protect image security. Security analysis and experiments show that its performance has reached a good practical level while effectively protecting users' privacy.

     

/

返回文章
返回