Range data acquisition and reconstruction of three-dimensional object

Abstract

Three-Dimensional object recognition is one of the fundamental problems in computer vision. In three-dimensional object recognition, the depth information is indispensable for further processing. This thesis deals with range data acquisition of three-dimensional objects and reconstruction of them. To acquire range data, following two procedure is necessary, i.e., Camera calibration procedure and range data acquisition procedure. First, Camera calibration stage, simple calibration technique is implemented to find unknown parameters of vision system. Second, Acquisition stage, we developed a range finder one of the active ranging techniques, so called light striping technique. This technique is mainly composed of laser projector, turntable and vision system. In light striping a single plane of light is projected onto a target object, which causes a stripe of light to appear on the image. Thinning operation extracts line information of stripe image. And range data is calculated by geometric triangulation. Several objects are tested and reconstructed and this range finder performs well with accuracy and flexibility.