Dissertation > Excellent graduate degree dissertation topics show

3D Reconstruction Based on Binocular Stereo Vision of Robot

Author: LiLiJun
Tutor: NiuZhiGang
School: Taiyuan University of Technology
Course: Vehicle Engineering
Keywords: TUT-CDMR robot binocular vision camera calibration fundamental matrix stereo matching 3D reconstruction
CLC: TP391.41
Type: Master's thesis
Year: 2012
Downloads: 140
Quote: 0
Read: Download Dissertation

Abstract


Coal energies are used largely in our country, and the frequency of ore difficult is very high in the state. The disaster would lead to the second disaster if the rescuers don’t know the specific situation of the underground. Thus, if we can obtain the exact information of the accident site by the robot in coal mine rescue, it will provide the valuable reference information to the rescuer. In the key technology of robot, binocular stereo vision has advantages of quickly detecting obstacles and making sure the object space position, and measuring rapidly and accurately, etc. Therefore, binocular stereo vision has become an important research direction in the research to the key technology of the robot. This paper aims to research robot and design how to implement stereo vision for the3D reconstruction system.The accurate calibration of stereo vision cameras is the foundation of researching stereo vision. Therefore, the camera calibration is further researched, and several calibration methods are discussed, finally the traditional camera calibration is adopted.16photos of calibration board are filmed respectively by the installation of the SONY camera and MIR camera from different point. Then, two cameras of TUT-CDMR robot are calibrated respectively by the camera calibration tool box of Matlab, and the calibration results of less error can be used as the based data of3D reconstruction.The estimation of fundamental matrix plays an important role in the3D reconstruction of binocular stereo vision. It is mainly to eliminate the error match in the stereo matching, which can provide more exact data of match point for3D reconstruction. First, the properties of fundamental matrix are introduced, and then the method of estimating fundamental matrix is elaborated. Finally, the experiment to some estimating method has been done, the distance of the point to the epipolar line and standard deviation and time are considered as evaluation index, and the result is that the robust method of MAPSAC is better than the other methods. The result of MAPSAC method from is satisfactory, therefore, the estimation of fundamental matrix is used that method in3D reconstruction.Stereo matching is the core content of stereo vision system, first, the paper introduced some basic concepts and constraints of stereo matching, then, the main algorithm of stereo matching is elaborated, and finally, the match of feature points based on epipolar geometry constraint is adopted, and a good matching result is obtained.The system of3D reconstruction is designed and accomplished by the use of Matlab GUI.Two images of outdoor scene are photographed by two cameras of the robot, and then, corner detection, stereo matching, epipolar line constraints, and3D reconstruction are experimented on them which demonstrate a fact that the method of this paper is effective.

Related Dissertations

  1. 3D Visualization Research of Medicial Ultrasound Image,TP391.41
  2. The Research of 3D Human Motion Capture Based on Reference Points,TP391.41
  3. Camera Calibration and Position and Pose Detecting on Vision Measurement System of PCB,TP391.41
  4. Approach for Calibration and Orientation of Digital Camera Based on Linear Features,TP391.41
  5. Research on 3D Reconstruction System for RPM Based on Medical Images,TP391.41
  6. The Research of Three-Dimensional Surface Reconstruction Algorithm Based on CT Image,TP391.41
  7. Research on Key Technologies for Binocular Stereo Vision,TP391.41
  8. Research and Implementation of 3D Reconstruction Algorithm Based on Computer Vision,TP391.41
  9. The Research on Object Correspondence Technology Based on Geometric Constraint in Multi-camera Video Surveillance System,TP391.41
  10. Sparse Representation Based Stereo Matching Algorithm and Infrared Targets Detection and Tracking,TP391.41
  11. Underwater 3D Reconstruction Based on Binocular Stereo Vision,TP391.41
  12. Moving Object Detection and Tracking for Volleyball Video,TP391.41
  13. The Construction of Eeg Numerical Calculatio Model Based on Medical Image,TP391.41
  14. Research on the Theory and Algorithms of 3D Reconstruction from Multiple Images,TP391.41
  15. Research on Detection System of Moving Object Based on Binocular Vision,TP391.41
  16. Camera calibration Related Issues,TP391.41
  17. Augmented reality registration methods of virtual objects,TP391.41
  18. HD image license plate location based vehicle tracking and violation detection,TP391.41
  19. Research on Binocular Stereo Vision System for Indoor Mobile Robot on Complex Environment,TP242
  20. Analysis and Research of the Shot-put Measuring Technology Based o Video Image Processing,TP391.41
  21. The Design and Implementation of Integrative Video Speed System Based on the Component Technology,TP391.41

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