Dissertation > Excellent graduate degree dissertation topics show

Point cloud based reconstruction system Research and Implementation

Author: WangErZhu
Tutor: TangHaoXuan
School: Harbin Institute of Technology
Course: Computer Science and Technology
Keywords: Three-dimensional surface reconstruction Point cloud simplification Mesh Generation Mesh optimization
CLC: TP391.41
Type: Master's thesis
Year: 2011
Downloads: 181
Quote: 0
Read: Download Dissertation

Abstract


With games, simulation graphics technology continues to evolve, as well as industrial design, simulation, three-dimensional television, and other complex objects in three-dimensional model of applications continues to increase, on the reconstruction of three-dimensional model of research more and more people's attention. And because the point cloud data easy to access, easy to store, so using point cloud reconstruction technique is also developing rapidly. A complete three-dimensional reconstruction system typically includes the following aspects: First, the point cloud pretreatment; Then the processed point cloud triangular mesh surface reconstruction, get initial mesh; Finally, the initial mesh optimization to get the final triangulation grid, it can be used by other modeling software. This article focuses on the three stages of in-depth study: the point cloud simplification phase, discussed in detail on the basis of the existing methods, we use non-uniform grids and curvature-based method of combining the original point cloud streamlined. First, the point cloud curvature values ??calculated for each point, and the point cloud mesh subdivision, so that the midpoint of each grid cell number is not greater than a threshold value M, then the midpoint of the curvature value according to the cell size and Cell size in accordance with a rule to streamline its so large curvature region retain more points, and will not appear on the surface area flat holes on the subsequent rebuilding of great benefit to the grid. In the mesh generation stage, the first of its original point cloud generated covering the surface of the adaptive ball set, and then calculate the intersection between the ball of to generate triangular mesh generation by controlling the size of the sphere radius to control the accuracy of the grid, generating different levels of fine mesh model; also proposes a vertex-based regional extension method, first select a seed vertex be extended to form a ring around this vertex of the triangle or triangular fan, while the new vertex Jia Rubo before the point of the list , followed by the list from the wavefront point out that can be extended to form a triangle ring until the wavefront point list is empty. This method can greatly reduce the mesh triangles and holes selfing generation. Optimization phase of the grid, mesh optimization element by defining the operation, each time you select a shape factor of the smallest grid mesh elements to perform optimization operation, reduce the complexity of the grid, mesh quality optimization. Finally, based on the proposed algorithm based on the design and implementation of a three-dimensional point cloud reconstruction system, capable of a variety of different formats point cloud processing, has good scalability. By testing different point cloud model to verify the feasibility and effectiveness of the algorithm. Based on the above method developed three-dimensional reconstruction of the system can be well used in the human body modeling, object modeling game scene development, able to quickly and accurately complete point cloud directly generated by the computer model. Also can be applied in other areas such as simulation, medical, cultural recovery display and so on.

Related Dissertations

  1. System design and application of a two-dimensional grid Automeshening,TP391.41
  2. Point Cloud Simplification Based on Affinity Propagation Clustering,TP391.41
  3. Application of Finite Volume Method in Melting Ice of Conductor Wire by Large Current,TM75
  4. Research on Point Clound and Grid Model Construction and Shape Distribution Retrieval,TP391.41
  5. Based on quadric surface approximation of the point cloud model segmentation,TP391.41
  6. Numerical simulation of complex boundary crossing within the three-dimensional turbulence,O357.5
  7. The Study of FEM for Internal Force Analysis of the Large Pumping Station,TV31
  8. Structural-electromagnetic Analysis of Large Mesh Reflector Antenna,TN823.27
  9. A Geometry User Environment of High End Digital Prototyping--Architecture Design and Software Implementation,TP391.72
  10. Numerical Simulation of Hypersonic Aircraft Gas Thermoelastic Multi-Field Coupling,V211.3
  11. Dynamic Prediction of the Road Status Based on Artificial Immune Network,TP18
  12. Based on a sample study of the finite element mesh model optimization,TB114
  13. Studies on Mesh Feature Editing Techniques for Reverse Innovative Design,TP391.7
  14. High-Quality and Controllable Quadrangulation Methods,TP391.41
  15. Study on the Triangle Mesh Translate into Quadrangle,TG381
  16. Parallel Unstructured Grid Generation and Numerical Computation,O35
  17. Two-dimensional area meshing algorithm,O241
  18. Research on GCS-Based Surface Reconstruction and Mesh Optimization,TP391.7
  19. Research on Constrained Delaunay Triangulation and Some Other Related Algorithm,TB115
  20. A Study on Adaptive Surface Mesh Generation,TP391.7

CLC: > Industrial Technology > Automation technology,computer technology > Computing technology,computer technology > Computer applications > Information processing (information processing) > Pattern Recognition and devices > Image recognition device
© 2012 www.DissertationTopic.Net  Mobile