The Research And Implement Of The Classification And Retrieval Of Archaeological Fragments

A Cultural relic is one of important objects for research of archaeology because archaeologist can obtain information such as production, life, science & technology, art and so on from cultural relics. Because of nature and human-generated forces, there exist hundred and thousands of fragments at archaeological excavation sites. How to quickly reconstruct complete cultural relics from fractions is a question, which archaeologists are interested in. At present, this work is finished mainly by hand. It is not only time-consuming, but also damaged to cultural relics. So it is emergent to develop a computer-aided reconstruction system, where the classification and retrieval of archaeological fragments serve as the basis for the further reconstruction process.Combining with techniques in these fields of computer geometry, computer graphics and pattern classification, we finished the classification and retrieval system of archaeological fragments based on geometry features, texture features and color features. Because most of cultural relics are made on rotational plates by hand, we research fragments from rotational symmetric objects. During the features extraction process, we estimate the rotational axis from 3D partial data, which serve as the classification process. We calculate the profile using 2D projection points. We extract the geometry features according to the rotational axis and the profile. We extract texture features and color features from 2D texture images. In database, we allocate an unique code for every fragment and save geometry, texture and color features. We finished the classification and retrieval of archaeological fragments using geometry,texture and color features.Main contributions are as follows:1. We put forward the framework of the classification and retrieval system.2. We adopt a quick method of estimating the rotational axis. Firstly, we calculate the normals using quadratic fitting method. Secondly, we estimate the rotational axis based on the line geometry method. This method advances robust, making the rotational axis more accurate.3. We adopt the method of estimation of the profile using Square Distance Minimization (SDM). Firstly, we project 3D point clouds, thus get 2D point clouds. Secondly, we adopt SDM method to estimate the profile, where we put forward a simple method to initialize the B-Spline curve using the features of 2D point clouds after projection. We can obtain good fitting effect.4. We put forward the geometry features, texture features and color features description method. Using abundant and complete features can powerfully support the classification, retrieval and reconstruction process.The classification and retrieval system of archaeological fragments finished the management, classification and retrieval of fragments, which serve as the further reconstruction process. It can largely decrease the error because of measure and draw manually. Using the system not only enhances the accurate of data, but also saves thetime for archaeologist.