3D modeling and codec of real objects are hot issues in the field of virtual reality. In this paper, we propose an automatic registration two range images method and a cycle based automatic global registration algorithm for rapidly and automatically registering all range images and constructing a realistic 3D model. Besides, to meet the requirement of huge data transmission over Internet, we present a 3D mesh encoding/decoding method for encoding geometry, topology and attribute data with high compression ratio and supporting progressive transmission. The research results have already been applied successfully in digital museum.
3D modeling, range images registration, 3D model compression and codecWith the development of 3D data acquisition technique, people can conventionally acquire 3D geometry data of real objects using 3D scanning devices. Due to the complexity of the geometry shape, real objects should be scanned by 3D scanning devices from multiple viewpoints. Point cloud data acquired with range information are called range images. A complete 3D model can be constructed using the range images registration technique to align these range images. However, the registration work is sometimes onerous, since the current methods for registering range images need the users intervention, and, for large complicated real objects, the 3D scanned data are pretty big. Therefore, it is meaningful to explore some rapid and automatic registration methods. Nowadays 3D models have found more and more applications over Internet; therefore we are in urgent need of a good 3D model codec method that can fast transfer and render these 3D models. Considering the abovementioned issues and taking 3D digital modeling and exhibition of precious relics as a background, we have done some work on 3D model construction, compression and codec technique research. This paper proposes a rapid and automatic range images registration method for constructing realistic 3D models, as well as a three-dimensional mesh encoding/decoding method for coding geometry, topology and attribute data with a high compression ratio and supporting progressive transmission. Our methods have already been applied successfully in 3D modeling of precious relics and 3D exhibition