Image perceptual hashing method based on two-sided random projection

Abstract

The invention discloses an image perceptual hashing method based on two-sided random projection, which mainly solves the problem of quick searching of massive image data. The method comprises the following steps of (1) preprocessing original image data; (2) obtaining a two-sided random projection matrix; (3) carrying out low-rank approximating; (4) updating a low-rank matrix; (5) judging if the number of iteration times of the updated low-rank matrix is maximum or not; (6) grouping projection vectors; (7) orthogonalizing the projection vectors; (8) obtaining hashing codes; (9) obtaining a hamming distance; and (10) outputting test results. The hashing method has the advantages that the better projection vector can be obtained, the effective hashing codes can be obtained, the memory consumption is reduced, the searching time is saved, the comprehensive performance of precision ratio and recall ratio of image searching can be improved, and the hashing method is applied to the image searching services of electronic businesses and mobile terminal equipment.

Description

technical field [0001] The invention belongs to the field of image processing, and further relates to an image perception hashing method based on bilateral random projection in the field of fast retrieval of large-scale image data. The invention can effectively carry out binary coding on the image, improves the performance of image retrieval, and has practical application value. Background technique [0002] In recent years, with the development of the Internet and information technology, big data has attracted more and more attention. According to the research results of the International Data Corporation, as of 2012, the amount of data generated globally has jumped to the ZB level. Image data is an important member of big data and an important way for people to communicate. Especially in today's era of rapid development of the Internet, e-commerce and mobile terminals, images are an indispensable part of people's production and life. In order to quickly and efficiently ...

AI Technical Summary

Problems solved by technology

The disadvantages of the method proposed in this patent application are: the locality-sensitive hashing method is a non-data-driven method based on a probability model. On the one hand, although the accuracy of the algorithm is theoretically guaranteed, the accuracy is relatively low; Sexual influence, the stability of the method is not satisfactory

However, the disadvantages of the method proposed in this patent application are: on the one hand, the direction of the principal component is obtained by linear dimensionality reduction, while most of the data have nonlinear characteristics, and the subsequent iterative optimization of the direction of the principal component will Increase the time complexity of the algorithm; on the other hand, for big data, the construction of multi-hash tables will undoubtedly greatly increase the occupancy rate of storage space

However, the disadvantage of the method proposed in this patent application is that this method still does not avoid the premise that the training data in the spectral hash model is forced to obey a uniform distribution, which limits its application value

Images