Fast 3D Scanning with Automatic Motion Compensation

By Thibaut Weise, Bastian Leibe, Luc Van Gool. 2007 IEEE Conference on Computer Vision and Pattern Recognition, 2007

338 citations recorded by the source corpus.

We present a novel 3D scanning system combining stereo and active illumination based on phase-shift for robust and accurate scene reconstruction. Stereo overcomes the traditional phase discontinuity problem and allows for the reconstruction of complex scenes containing multiple objects. Due to the sequential recording of three patterns, motion will introduce artifacts in the reconstruction. We develop a closed-form expression for the motion error in order to apply motion compensation on a pixel level. The resulting scanning system can capture accurate depth maps of complex dynamic scenes at 17 fps and can cope with both rigid and deformable objects.

Subfields: Computer Vision and Pattern Recognition, Media Technology

View Paper: https://www.semanticscholar.org/paper/16b1167e78e6c57aab84324722225038cf47de1a

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

In Media Technology

Citations in the corpus

Real time head pose estimation with random regression forests (2011, methodology): “5D data output of a range scanner like [27].”
Internet Based Morphable Model (2013, methodology): “Similarly we collected photos of adults making different facial expressions.”
Robust single-view geometry and motion reconstruction (2009, methodology): “Our dynamic acquisition system shown in Figure 4 provides dense depth maps with a spatial resolution of 0.5 mm at 25 frames per second (see [Weise et al. 2007] for a description of a similar acquisition setup).”
Static and dynamic 3D facial expression recognition: A comprehensive survey (2012, background): “Recently, high-speed structure light based systemswere proposed [24–27,34]which use projected fringe patterns.”
Real Time Head Pose Estimation from Consumer Depth Cameras (2011, background): “Such approaches, however, use bulky 3D scanners like [22] and are not useful for consumer products or mobile applications like robots.”
Temporal phase unwrapping algorithms for fringe projection profilometry: A comparative review (2016, background): “The measured object is not motionless during the acquisition process; that is, one point in the projected pattern sequence can map to different points on the object surface, which causes socalled motion artifacts [89,90].”
Random Forests for Real Time 3D Face Analysis (2012, methodology): “We ran our real time system on a Intel Core 2 Duo computer @ 2 GHz, equipped with 2 GB of RAM, which was simultaneously used to acquire the range data as explained in Weise et al. (2007).”
A Survey on Human Motion Analysis from Depth Data (2013, methodology): “The subjects, both male and female, with and without glasses, were recorded using the scanner of [63] while turning their heads around, trying to span all possible yaw and pitch rotation angles they could.”
Example-based facial rigging (2010, methodology): “To evaluate our method, we created two sets of training poses using geometric modeling tools and two sets using a 3D scanner [Weise et al. 2007].”
High-speed 3D shape measurement with structured light methods: A review (2018, methodology): “Following this endeavor, numerous real-time techniques including [78–81] have been developed for 3D shape measurement leveraging both advanced hardware technologies (e.g., GPU) and new innovations on software algorithms.”