The Development Of Software For Cultural Relic Images Identifivation And Retrieval

Aimed at the identification and retrieval of cultural relic images, this paper established the image database of bronze mirrors by selecting their features and developed a software system for identification and retrieval of bronze mirrors. The research was done on the foundation of Content-Based Image Retrieval (CBIR) theoretically and the discussion with the related experts.The main points of the paper are as follows:(1) In order to benefit the extraction of image features, the pre-processing technologies for the bronze mirror images were used which include normalization, converting color images into gray-level images and de-noise etc.(2) The texture and shape features of the bronze mirror images were extracted respectively based on the gray-gradient co-occurrence matrix and its shape features. The similitude match algorithm for identifing the images was put forward too.(3) On the image retrieval, a new method was presented which includes two retrievals. The first one was based on the texture similarity of the images. The second one was carried out on the principle that images with similar texture have shape similarity of the gray-gradient co-occurrence matrix. As a result, more accurate retrieval was obtained.(4) A database for bronze mirror images was developed by integrating Visual Basic with Access and the system design was accomplished by integrating VB with Matlab seamlessly. A new method for data exchange between VB and Matlab was proposed. Furthermore, each module of the system was developed separately and relative program was written.(5)The functions and usage of each module were introduced in detail by an example. The feasibility of the system was proved.There are some new viewpoints on the image retrieval. The texture features and shape features were utilized simultaneously and the retrieval precision was improved by two retrievals. It was proved that the developed system could satisfy the basic demand of the cultural relic images identification and retrieval. It would provide a convenient tool for scientific management, query and preliminary identification of the cultural relic image information and could be helpful for relative cultural relics administrative department.