A New Method for Low Polygonal Modeling of Colored Object
Kazuki Kitaguchi, Masaki Saito · 2006
This paper proposes a new method for low polygonal modeling that approximates object shape and color simultaneously. The basic idea of this method is online extraction of vertices for creating colored polygons. A composite image sensor, a 3D color scanner which consists of an infrared laser projector and two cameras, is employed to digitize 3D points and their colors along the projected light on object surface. Vertices are extracted from every scan data so as to satisfy the approximation criteria for object shape and color. Temporary polygons are produced by connecting the vertices of two successive scan data. Low polygon model is built by merging smooth and same colored temporary polygons. We have carried out some experiments with real objects. The modeling accuracy and processing speed of our method are discussed through the results.