DISCRIMINATION OF CULTURAL RELICS SIMILARITY BASED ON PHASH ALGORITHM AND SIFT OPERATOR

Abstract. Among the immovable cultural relics such as grottoes, stone carvings of the same period usually show similar artistic styles and repetitive statues. Through the study of the similarity between cultural relics in the same period, we can have reference value for the historical significance, artistic characteristics and cultural relics restoration of cultural relics. The traditional discrimination of cultural relics similarity often depends on the subjective experience judgment of experts, which is time-consuming and inefficient. Often, different experts may issue different conclusions, the lack of convincing objective evaluation basis, can not provide effective scientific support for the protection and restoration of cultural relics. In order to provide objective criteria for evaluating the similarity of cultural relics and avoid the interference of human factors, this paper proposes the concept of similarity index of cultural relics based on perceptual Hash (pHsah) algorithm. For orthophoto images of cultural relics with similarity index within the threshold, the SIFT operator is further used to find the corresponding feature points. This paper takes the Buddha statues in Yungang Grottoes as the research object, objectively calculates the similarity index of cultural relics, combines the pHash algorithm and Sift operator, realizes the objective quantitative discrimination of cultural relics similarity, effectively avoids the subjective influence of experts on the judgment of cultural relics similarity.