基于PCA的哈希图像检索算法  

Image retrieval based on PCA hash

作　　者：马绍覃 张鸿[1,2] MA Shao-qin;ZHANG Hong(School of Computer Science and Technology,Wuhan University of Science and Technology,Wuhan 430065,China;Hubei Province Key Laboratory of Intelligent Information Processing and Real-Time Industrial Systems,Wuhan University of Science and Technology,Wuhan 430065,China)

机构地区：[1]武汉科技大学计算机科学与技术学院,湖北武汉430065 [2]武汉科技大学智能信息处理与实时工业系统湖北省重点实验室,湖北武汉430065

出　　处：《计算机工程与设计》2020年第2期483-487,共5页Computer Engineering and Design

基　　金：国家自然科学基金项目（61373109）

摘　　要：图像哈希算法的步骤大致分为投影和量化两个阶段,为提高哈希编码的性能,分别对这两个阶段进行研究。在投影阶段,通过主成分分析算法将数据投影到新的特征子空间中,以降低原始特征之间的冗余性;在量化阶段,为减少量化所带来的损失,提出一种单双比特结合的量化方法;利用得到的哈希编码进行图像检索。在两个常用的图像数据集上的实验结果表明,提出的算法较现有的主流图像哈希算法在多个评价指标下均有所提高。The steps of image hashing algorithm can be roughly divided into two stages,namely projection and quantization.To improve the performance of hash coding,the two phases were analyzed.In first phase,to reduce the redundancy among some features,the data were projected into the new feature subspace using the principal component analysis algorithm.In the quantization phase,considering the characteristics of the principal component analysis algorithm,to minimize the loss caused by quantization,a single-bit and double-bit combination quantization method was proposed.Image retrieval was performed using the binary hash codes.The experimental results on two commonly used image datasets show that the proposed algorithm is improved compared with the existing mainstream image hashing algorithms under multiple evaluation indicators.

关 键 词：投影 量化 哈希 图像检索 主成分分析 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]