Abstract:Advances in laser scanning technology are making 3D acquisition feasible for increasingly large objects. Different from most existing scanning projects, which have been devoted to short or medium range objects, our research focuses on dealing with long range outdoor environment scans. Compared with indoor scanning, the data obtained through outdoor scanning is large, incomplete, complex, inaccurate, and non-uniformly sampled. These properties raise many new challenges for processing outdoor scanned data. The research presented in this thesis aims to model, stylize, and render scanned outdoor environments and deliver interactive navigation. Existing methods are normally inadequate to construct models for outdoor scanned objects such as trees and buildings because of the incompleteness and complexity of such data. In this thesis, we have developed a knowledge-based method to model outdoor scanned trees. With knowledge about the structures of trees, we use certain heuristics to produce full polygonal models from the scanned data. The difficulty in rendering outdoor environments is accommodating the inherently incomplete and inaccurate data sets to deliver a smooth animation without suggesting the underlying data deficiency. Our solution to this issue is to employ non-photorealistic rendering techniques for rendering outdoor environment scans represented by points. In this way, varying importance and accuracy in the scanned data can be naturally illustrated. We have developed a unified framework for producing sketchy, painterly, cartoon, and intermingled styles. We have proposed a continuous level-of-detail data structure, the continuous resolution queue, to promise a coherent and consistent animation. We have also leveraged modern graphics hardware to achieve interactive rendering of large scenes.

3D scene analysis and modeling both for static and dynamic environments

Combining Laser-Scanning Data and Images for Target Tracking and Scene Modeling

3-D Surround View for Advanced Driver Assistance Systems.

Unifying Terrain Awareness Through Real-Time Semantic Segmentation

3D Scene Modeling and Understanding from Image Sequences

3D Indoor Environment Modeling and Detection of Moving Object and Scene Understanding

Analysis and Modeling of 3D Indoor Scenes

Three-Dimensional Geometric-Physical Modeling of an Environment with an In-House-Developed Multi-Sensor Robotic System

Spatiotemporal Analysis of Static and Dynamic Traffic Elements from Road Scenes.

Scene Understanding in a Large Dynamic Environment Through a Laser-Based Sensing.

Vision-Based Moving Objects Detection with Background Modeling

Dynamic 3D Scene Analysis by Point Cloud Accumulation

Towards Scene Understanding with Detailed 3D Object Representations

INDOOR AND OUTDOOR STRUCTURED MONOMER RECONSTRUCTION OF CITY 3D REAL SCENE BASED ON NONLINEAR OPTIMIZATION AND INTEGRATION OF MULTI-SOURCE AND MULTI-MODAL DATA

A survey of Object Classification and Detection based on 2D/3D data

3D Scene Understanding at Urban Intersection Using Stereo Vision and Digital Map.

Automatic 3D Indoor Scene Modeling from Single Panorama

Dynamic Obstacle Detection Through Cooperation of Purposive Visual Modules of Color, Stereo and Motion

3D indoor scene modeling from RGB-D data: a survey

Modeling, Stylization, and Rendering of Three-Dimensional Scanned Outdoor Environments

Human-centric indoor environment modeling from depth videos