| Structured light techniques are one of the active research areas in 3D sensing. Traditional structured light methods require several different patterns to recover the depth, without ambiguity or albedo sensitivity, and are corrupted by object movement during the projection/capture process. A novel projection pattern, called the composite pattern, is introduced in this dissertation. It efficiently combines multiple patterns into a single composite pattern projection and the 3D reconstruction can be achieved from a single camera captured image. Therefore, it allows for video based range acquisition for dynamic scenes. Experiments are performed to evaluate the composite pattern technique for single frame 3D acquisition. Because our structured light system requires only standard image capture and projection devices, unlike time of arrival techniques, they are relatively low cost. Attaining low cost video 3D data acquisition extends potential applications that are presently limited to 2D video imaging. In particular, the composite pattern is studied to estimate the 3D hand rotation for the potential applications in human-computer interaction. A set of video clips are also made to demonstrate the video-based 3D acquisition and possible applications in human computer interfacing. However, the 3D results from the composite pattern technique are not precise comparing with the high-accuracy multiple-pattern approaches. A novel post-processing algorithm for the composite pattern is also introduced and analyzed in this dissertation. It is adapted from correspondence matching techniques used in stereo vision to improve the 3D accuracy. Simulations and experiments illustrate the improvement with the post-processing algorithms. |
