People - Research Staff / Brad Grinstead

IRIS Research Staff

Brad Grinstead
Ph.D. Student

		Office:	410 Science & Engineering Research Facility The University of Tennessee Knoxville, TN 37996-2100
		Telephone:	(865) 974-9213
		Fax:	(865) 974-5459
		E-mail:	grinstead@imaging.utk.edu
		Personal Web Page:	http://imaging.utk.edu/~grinstead/

Current Work: Detail Enhancing Denoising of Digitized 3D Models from a Mobile Scanning System

Here, work is presented on the development of a method for denoising the 3D data acquired by a mobile scanning system used to digitize large-scale environments. This digitization platform combines a variety of sensors - including laser range scanners, video cameras, and pose estimation hardware - on a mobile platform for fast acquisition of 3D models of real world environments. The data acquired by such systems are extremely noisy, often with significant details being on the same order of magnitude as the system noise. By utilizing a unique 3D surface analysis tool, we intend to develop a denoising algorithm that identifies regions of detail and enhances their appearance, while removing the effects of noise on the overall model. The developed algorithm can be useful for a variety of digitized 3D models, not just those involving mobile scanning systems.

		Detail Enhancing Denoising of Digitized 3D Models from a Mobile Scanning System - Large-scale 3D digitization systems produce a vast amount of raw data that often contains noise that is on the same order of magnitude as important surface geometry. This work is aimed at developing processing methods to identify those areas, enhance the details present in the raw data, and remove the effects of noise from the overall surface.
		Pose From Motion: A Global Approach - There is a need for pose estimation from video sequences that is robust to outliers caused by mixed motion effects (i.e., static and dynamic objects in the same scene). This project is about developing such a pose estimation system.
		Road Scanning for Terrain Modeling - Utilizing VB scanning methods to digitize terrain from ground level scans. A short-range, high-accuracy laser range scanner is combined with pose estimation equipment to acquire high-resolution terrain models.
		Fast Digitization of Reality Using Mobile 3D Scanning - Developing a system to quickly capture real world models using an ad hoc vehicle-borne laser range scanning system.

Publications:

B. Grinstead, A. Koschan, A. Gribok, and M. Abidi, "Outlier rejection by oriented tracks to aid pose estimation from video," accepted for publication in Pattern Recognition Letters, June 2005.

B. Grinstead, A. Koschan, and M. Abidi, "Hybrid Self Localization for a Mobile Robotic Platform in Indoor and Outdoor Environments," in Transactions of the American Nuclear Society, Vol. 92, San Diego, CA, USA, pp. 52-53, June 2005.

B. Grinstead, A. Koschan, and M. Abidi, "A comparison of pose estimation techniques: Hardware vs. Video," in Proc. SPIE Unmanned Ground Vehicle Technology VII, Vol. 5804, Orlando, FL, USA, pp. 166-173, March 2005.

B. Grinstead, A. Koschan, and M. Abidi, "Developing detailed a priori models of large environments to aid in robotic navigation tasks," Proc. of the SPIE Defense & Security Symposium: Unmanned Ground Vehicle Technology, pp. 561-568, Orlando, FL, April 2004.

B. Grinstead, A. Koschan, and M. Abidi, "Fast Digitization of Large-Scale Hazardous Facilities," Proc. of 10th Int. Conf. on Robotics & Remote Systems for Hazardous Environments, pp. 521-525, Gainesville, FL, March 2004.

Last updated:
Webmaster