Research and implementation of perceptual hash algorithm for image retrieval
Wenlie Zhu, Jing Chang · AIP conference proceedings · 2019
With the extensive development of the Internet, digital image resources are more and more abundant. How to effectively retrieve and utilize a large amount of image information has become a hot issue in the current research. In this paper, a perceptual hash algorithm based on SIFT features is proposed by combining the perceptual hash algorithm with the SIFT feature of the image, which helps to overcome the influence of the perceptual hash algorithm on geometric attacks, image scale changes and other abnormal conditions. Improve the accuracy of the perceptual hash algorithm. By studying the application scene of perceptual hash algorithm, using the image-aware hash algorithm to perceive the target object in the image scene, matching the relevant features of the image, the image retrieval technology based on the hash algorithm is realized.