High-resolution Shape and Color Integration of Dynamic Rigid Body Using 3D Motion Sensing System
Summary
Recently, in the applications such as archiving and modeling, it is expected to scan 3D shape of a moving dynamic object with high resolution and high accuracy. For this purpose, 3D shape integration is one of the effective methods. This method integrates multiple measured range images into a single shape and can reconstruct high-resolution and high-precision shape whose performance exceeds a single depth image.
In this research, we propose a new type of solution for this application based on 3D motion sensing system using lasers. Instead of estimating the 3D motion from time-sequential depth images or color images, by using the results of the 3D motion sensing, the depth images obtained by any type of depth sensors can be fused into a single surface. Moreover, in this method, we can precisely track the any positions on the target surface. Therefore, we can know where the tracked point is located on the captured image, if the camera is calibrated. This allows color acquisition as additional functions. This work can be applied to capture of various moving objects in natural everyday scenes, including flying objects, walking humans, travel-ling vehicles, and so on.
Movie
If you want to use the original video, please send an e-mail for copyright permission to contact .
References
- Leo Miyashita, Ryota Yonezawa, Yoshihiro Watanabe and Masatoshi Ishikawa : 3D Motion Sensing of any Object without Prior Knoledge, ACM Transactions on Graphics, Vol.34, No.6, 218 (2015)
