Dissertation > Excellent graduate degree dissertation topics show

Research of 3D Reconstruction Based on Image Sequence

Author: QianLei
Tutor: YeXiaoLing
School: Nanjing University of Information Engineering
Course: System Analysis and Integration
Keywords: Image preprocessing Feature matching Camera Calibration From the movement to restore the structure Bundle adjustment
CLC: TP391.41
Type: Master's thesis
Year: 2011
Downloads: 194
Quote: 0
Read: Download Dissertation

Abstract


Based on the three-dimensional reconstruction of the image sequence is a combination of computer vision and computer graphics technology, through a series of images of the scene or target shooting restore the camera parameters and the scene of the three-dimensional information. Simple modeling to update fast image sequence based 3D modeling techniques in archeology, architecture, geology, virtual reality, robot navigation, precision industrial measurement, object recognition, site investigation, and military field has a very broad application prospects. How three-dimensional reconstruction as automatically as possible is a leading issue, the study of the three-dimensional reconstruction of each part of the method considered automatically issues. This article describes a multi-view geometry, the process of establishing the basic concepts of projective geometry and camera model, image pre-processing characteristics of the detection and matching, as well as camera calibration work based on the three-dimensional reconstruction of the image sequence depth. The paper's main work is divided into the following sections: The first part is the detection and filtering of image noise, median filtering and its various improvement methods, based on BP neural network adaptive switching noise detection filtering method . This method makes full use of the pixels itself and neighborhood information, high noise detection efficiency, automatic filter. The second part of the image feature matching, the article describes the feature detection and matching techniques applicable in the three-dimensional reconstruction, and the use of random sample consensus (RANSAC) algorithm and the fundamental matrix, homography constrained optimization match. For camera calibration using the template, a description based on the the projective invariance template characteristics. The third part is the camera calibration, the dot set plane template-based camera calibration program, automatic camera calibration accuracy, suitable for multi-angle three-dimensional visual medium and small calibration target. The same time, a fine calibration algorithm for multi-view stereovision image downscaling methods and bundle adjustment optimization to achieve more accurate calibration of the camera parameters. Finally, the use of the actual shooting of the image sequence, the camera image sequence parameter estimation and scene three-dimensional information to restore binding part of the methods discussed in this article. In addition, the combination of the technical realization recover from sports structure (SFM), multi-view stereo vision (MVS) and bundle adjustment (BA).

Related Dissertations

  1. The Research of 3D Human Motion Capture Based on Reference Points,TP391.41
  2. Camera Calibration and Position and Pose Detecting on Vision Measurement System of PCB,TP391.41
  3. Research and Realization of License Plate Character Recognition Algorithm Based on SVM,TP391.41
  4. The Research and Implement on Camera Calibration Technology Based on Trifocal Tensor,TP391.41
  5. Single Camera Calibration Method Based on Dual-mirror Imaging,TP391.41
  6. Raster-based three-dimensional projection of dental measurement system,TP391.41
  7. Based on Machine Vision Industrial robots crawling technology research,TP242.2
  8. Applied to narrow gap SAW image acquisition and pre-processing System,TG445
  9. Blends of different fibers and quantitative analysis of image recognition,TP391.41
  10. Based on the visual system of unmanned autonomous helicopter landing,V279
  11. Based on digital image processing meter image recognition technology research and implementation,TP391.41
  12. Dollar bill image detection of critical information and number identification,TP391.41
  13. Camera calibration Related Issues,TP391.41
  14. Light-section method vision measurement system, the key technology research and implementation,TP391.41
  15. Stereoscopic three-dimensional reconstruction -related technology research and implementation,TP391.41
  16. Augmented reality registration methods of virtual objects,TP391.41
  17. HD image license plate location based vehicle tracking and violation detection,TP391.41
  18. Immersive Virtual Shooting Training System in Human-Computer Interaction Research,TP11
  19. Research on Binocular Stereo Vision System for Indoor Mobile Robot on Complex Environment,TP242
  20. Research on Calibration Technology for Multiple Camera Field Stitching Measuring System,TP391.41
  21. Research on the Key Technology of Iris Recognition,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