Research On Image Data Classification Of Cultural Relics Assisted By Image Search

In recent years,cultural,museum and archaeological work has received extensive attention.Following the completion of the 13 th five year plan for the development of national cultural relics,the State Council put forward the 14 th five year plan for cultural relics protection and scientific and technological innovation.It is the first time to discuss the combination of scientific and technological innovation and the field of culture and Museum,and put forward the specific requirements for the construction of a large database of national cultural relics resources.Recommend the high-definition data acquisition,display and utilization of cultural relics image information,and establish a standard system for the digitization of cultural relics.The cultural relic image data classification technology based on image search is an important part of the planning.Although it was occasionally mentioned in the previous research on the classification of cultural relics image data,the applicability of the actual cultural relics image classification is slightly poor due to the lack of professional background and lack of understanding of the classification standards of the cultural relics industry.Based on the cultural relics image classification standard of the cultural and museum industry,this thesis makes a comparative analysis of the cultural relics image data classification by using interdisciplinary technical means,and discusses the classification effect.From the perspective of cultural relics industry,the effectiveness of image classification is realized.The main research work is as follows:1.Sort out the relevant theories and practices of digitization in the cultural and Expo industry,the development process of image search technology,mature cases of image search in other fields,and the online classification standards of cultural and Expo image data in China.2.Select the largest number of pottery,stone,porcelain,jade and bronze artifacts from the cultural relics image data collected through field investigation and online sorting,establish a cultural relics image database,carry out image preprocessing,compare the segmentation effects of cultural relics images such as threshold,segmentation and regional segmentation,and complete image noise reduction for cultural relics images.3.Using image search technology to compare and verify the image classification of five kinds of artifacts,and verify the success and efficiency of image search technology to classify the same texture and shape.Through sift,histogram,main tone and perceptual hash,this thesis analyzes the characteristics of cultural relics,and compares their classification success and efficiency.Sift and perceptual hash can effectively help the classification of cultural relics images.Finally,it discusses the necessity of using this kind of technology in the field of culture and Museum from the aspect of talent capital.According to the experimental results,the results can effectively promote the digital management of the cultural and Expo industry and facilitate the research of cultural relics.After the technical scheme is mature,it looks forward to helping the construction of grass-roots museums,promoting the standardization of cultural relics data,and improving the interconnection between museums and the masses.