A5074871715- Advanced Vision and Imaging
- Robotics and Sensor-Based Localization
- Computer Graphics and Visualization Techniques
- 3D Shape Modeling and Analysis
- Image Processing and 3D Reconstruction
- 3D Modeling in Geospatial Applications
- Advanced Image Processing Techniques
- Image Processing Techniques and Applications
- Advanced Measurement and Detection Methods
- Robotic Path Planning Algorithms
- Model-Driven Software Engineering Techniques
- Evacuation and Crowd Dynamics
- Color perception and design
- Human Motion and Animation
- Traffic control and management
- Data Visualization and Analytics
- Video Surveillance and Tracking Methods
- Urban Design and Spatial Analysis
- Arctic and Russian Policy Studies
- Autonomous Vehicle Technology and Safety
- Software Engineering Research
- Software Testing and Debugging Techniques
- Architecture and Computational Design
- Economic and Technological Systems Analysis
- Environmental Sustainability and Technology
Stanford University
2010-2011
University of North Carolina at Chapel Hill
2006-2010
Brigham Young University
2002-2008
North Carolina State University
2007
We present a viewpoint-based approach for the quick fusion of multiple stereo depth maps. Our method selects estimates each pixel that minimize violations visibility constraints and thus remove errors inconsistencies from maps to produce consistent surface. advocate two-stage process in which first stage generates potentially noisy, overlapping set calibrated images second fuses these obtain an integrated surface with higher accuracy, suppressed noise, reduced redundancy. show by dividing...
We present an interactive furniture layout system that assists users by suggesting arrangements are based on interior design guidelines. Our incorporates the guidelines as terms in a density function and generates suggestions rapidly sampling using hardware-accelerated Monte Carlo sampler. results demonstrate suggestion generation functionality measurably increases quality of produced participants with no prior training design.
We present an interactive furniture layout system that assists users by suggesting arrangements are based on interior design guidelines. Our incorporates the guidelines as terms in a density function and generates suggestions rapidly sampling using hardware-accelerated Monte Carlo sampler. results demonstrate suggestion generation functionality measurably increases quality of produced participants with no prior training design.
We present a method for automated generation of building layouts computer graphics applications. Our approach is motivated by the layout design process developed in architecture. Given set high-level requirements, an architectural program synthesized using Bayesian network trained on real-world data. The realized floor plans, obtained through stochastic optimization. plans are used to construct complete three-dimensional with internal structure. demonstrate variety computer-generated...
We present a method for automated generation of building layouts computer graphics applications. Our approach is motivated by the layout design process developed in architecture. Given set high-level requirements, an architectural program synthesized using Bayesian network trained on real-world data. The realized floor plans, obtained through stochastic optimization. plans are used to construct complete three-dimensional with internal structure. demonstrate variety computer-generated...
The paper introduces a data collection system and processing pipeline for automatic geo-registered 3D reconstruction of urban scenes from video. collects multiple video streams, as well GPS INS measurements in order to place the reconstructed models geo- registered coordinates. Besides high quality terms both geometry appearance, we aim at real-time performance. Even though our is currently far being real-time, select techniques design modules that can achieve fast performance on CPUs GPUs...
Abstract We present a novel method for the synthesis and animation of realistic traffic flows on large‐scale road networks. Our technique is based continuum model flow we extend to correctly handle lane changes merges, as well behaviors due in speed limit. demonstrate how our can be applied many vehicles network at interactive rates show that simulate believable publicly‐available, real‐world data. furthermore scalability this many‐core systems.
Model synthesis is a new approach to 3D modeling which automatically generates large models that resemble small example model provided by the user. extends 2D texture problem into higher dimensions and can be used many different objects environments. The user only needs provide an appropriate does not need any other instructions about how generate model. create symmetric models, change over time, fit soft constraints. There are two important differences between our method existing...
We present a method for automated generation of building layouts computer graphics applications. Our approach is motivated by the layout design process developed in architecture. Given set high-level requirements, an architectural program synthesized using Bayesian network trained on real-world data. The realized floor plans, obtained through stochastic optimization. plans are used to construct complete three-dimensional with internal structure. demonstrate variety computer-generated...
We present a method for procedurally modeling general complex 3D shapes. Our approach can automatically generate models of buildings, man-made structures, or urban data sets in few minutes based on user-defined inputs. The algorithm attempts to that resemble input model and satisfy various dimensional, geometric, algebraic constraints control the shape. These are used capture intent user shapes look more natural. also describe efficient techniques handle highlight its performance many...
We present a novel method for procedurally modeling large complex shapes. Our approach is general-purpose and takes as input any 3D polyhedral model provided by user. The algorithm exploits the connectivity between adjacent boundary features of computes an output that has similar connected resembles input. No additional user needed to guide generation proceeds automatically. In practice, our simple implement can generate variety shapes representing buildings, landscapes, fractal in few minutes.
We present an approach for automatic 3D reconstruction of outdoor scenes using computer vision techniques. Our system collects video, GPS and INS data which are processed in real-time to produce geo-registered, detailed models that represent the geometry appearance world. These generated without manual measurements or markers scene can be used visualization from arbitrary viewpoints, documentation archiving large areas. consists a acquisition processing video sequences off-line but...
We present a method for automatically generating polygonal shapes from an example using graph grammar. Most procedural modeling techniques use grammars with manually created rules, but our can create them example. Our generate graphs that are locally similar to given disassemble the input into small pieces called primitives and then reassemble new graphs. organize all possible hierarchy find matching within hierarchy. These matches used grammar construct every graph. generates converts...
In order for an unmanned aerial vehicle (UAV) to safely fly close the ground, it must be capable of detecting and avoiding obstacles in its flight path. From a single camera on UAV, 3D structure surrounding environment, including any obstacles, can estimated from motion parallax using technique called motion. Most algorithms attempt reconstruct environment optical flow value at each feature point. We present novel method calculating that does not require precise calculation Due effects image...
We present a novel method for procedurally modeling large complex shapes. Our approach is general-purpose and takes as input any 3D polyhedral model provided by user. The algorithm exploits the connectivity between adjacent boundary features of computes an output that has similar connected resembles input. No additional user needed to guide generation proceeds automatically. In practice, our simple implement can generate variety shapes representing buildings, landscapes, fractal in few minutes.
Recent advances in many multi-discipline technologies have allowed small, low-cost fixed wing unmanned air vehicles (UAV) or more complicated ground (UGV) to be a feasible solution scientific, civil and military applications. Cameras can mounted on-board of the for purpose scientific data gathering, surveillance law enforcement homeland security, as well provide visual information detect avoid imminent collisions autonomous navigation. However, most current computer vision algorithms are...
We present a method for procedurally modeling general complex 3D shapes. Our approach is targeted towards applications in digital entertainment and gaming can automatically generate models of buildings, man-made structures, or urban datasets few minutes based on user-defined inputs. The algorithm attempts to results that resemble input model satisfy various dimensional, geometric, algebraic constraints. These constraints are used capture the intent user shapes look more natural. also...
We present an evaluation methodology and data for large scale video-based 3D reconstruction. evaluate the effects of several parameters draw conclusions that can be useful practical systems operating in uncontrolled environments. Unlike benchmark datasets used binocular stereo multi-view reconstruction evaluations, which were collected under well-controlled conditions, our are captured outdoors using video cameras mounted on a moving vehicle. As result, videos much more realistic include...
Structure from motion is a technique that attempts to reconstruct the 3D structure of scene sequence images taken camera moving within scene. can be used on an Unmanned Aerial Vehicle or Ground for obstacle detection as well path-planning and navigation. The estimated using optical flow values found at set feature points image. Typically, this done under assumption all have same level accuracy each value in horizontal vertical directions same. These assumptions are not entirely correct....
For a UAV to be capable of autonomous low-level flight and landing, the must able calculate its current height above ground. If speed is approximately known, can estimated from apparent motion ground in images that are taken an onboard camera. One most difficult aspects estimating lies finding correspondence between position object one image frame new succeeding frames. In some cases, due effects noise aperture problem, it may not possible find correct object's next frame. Instead, only set...