Dissertation > Excellent graduate degree dissertation topics show

Research on Key Technologies of Auto-stereoscopic 3D Augmented Reality

Author: LiuLingShan
Tutor: ZhangMing
School: Zhejiang University
Course: Information and Communication Engineering
Keywords: Augmented Reality Feature extraction and matching Simultaneous localization and mapping Cutout
CLC: TP391.41
Type: Master's thesis
Year: 2012
Downloads: 126
Quote: 0
Read: Download Dissertation

Abstract


Augmented reality technology through computer technology the virtual information to the real environment, so the real environment and virtual objects superimposed in real time on the same screen or space. Naked eye 3D augmented reality is committed to provide users with a free three-dimensional display of the actual situation fusion screen to bridge the communication gap in the real world and the virtual world, allowing users to experience the full interactive effect of creating a sense of the actual situation. Camera tracking, image feature extraction and matching, as well as virtual viewpoint image rendering technology is a key technology in the naked eye 3D augmented reality. Natural three-dimensional television system captured by binocular camera three-dimensional natural scene and extract depth information via the encoding, transmission and decoding by multi-view reconstruction module to map out multi-channel virtual viewpoint image left and right eye for the user to provide a three-dimensional image in order to achieve the naked eye 3D visual enjoy. This article is committed to achieve natural 3D TV experiment platform multi-view autostereoscopic display 3D augmented reality feature, first learn SLAM (Simultaneous localization and mapping) technology to construct the augmented reality experiment platform, and then focus on the study of feature extraction and matching algorithm and proposed algorithm based on Sobel edge fast SIFT (Scale-invariant feature transform), and finally for the multi-view reconstruction module virtual viewpoint image rendering process, the edge of the depth map is inaccurate optimization method. The results achieved include the following three aspects: structures augmented reality experiment platform. The experimental platform draws on the robot research areas SLAM technology, real-time building scene map and calculates the camera position and orientation parameters. Feature extraction and matching technology and proposed SIFT algorithm based on Sobel edge. Firstly, the status of research, feature extraction and matching algorithms. SIFT algorithm exists high redundancy, slow disadvantage Sobel algorithm extracted image edge set and edge set image scale space, to detect extreme points in the scale space constraints. The experimental results show that the proposed algorithm can substantially reduce the redundancy of feature points to improve the algorithm speed while ensuring a higher matching rate. As the Sobel edge detection threshold value increases, the decrease in the number of feature points, reduced running time, and the matching rate. Depth map for the stereo matching imprecise reduce the problem of cause viewpoint synthetic quality, this paper introduces the the matting algorithm with quasi-depth map and color image edge, improve the quality of the virtual viewpoint image. Can only distinguish between of global disadvantage of foreground and background, matting technology proposed a localized matting algorithm to solve complex scenes and more depth level distinction and edge alignment. The experimental results show that, after the registration edge of the depth map have a significant contraction of the edge of the virtual viewpoint image has improved significantly than before the alignment. Local Cutout algorithm can the some global matting algorithm can not distinguish the edge pixels distinguish, virtual viewpoint image quality is further enhanced.

Related Dissertations

  1. Research of Reconstruction and Camera Pose Estimation on Mobile Phone,TP391.9
  2. Augmented Reality Virtual Surgery Dental Research on key technologies,TP391.9
  3. Outdoor target recognition and application in augmented reality,TP391.9
  4. Flame effects Augmented Reality System Key Technologies,TP391.9
  5. Museum of mobile augmented reality navigation Applied Research,TP391.9
  6. ARToolKit augmented reality interaction scenarios based on research,TP391.41
  7. Stereoscopic three-dimensional reconstruction -related technology research and implementation,TP391.41
  8. Augmented reality registration methods of virtual objects,TP391.41
  9. Real-time modeling spatial augmented reality technology research related,TP391.41
  10. For Intelligent Cleaning Robot Path Planning,TP242
  11. Research on the Key Technology of Iris Recognition,TP391.41
  12. Research on Key Technologies of Monocular Visual Odometry Based on DSP,TP391.41
  13. A Music Retrieval System Base on Humming,TN912.3
  14. Research on Automatic Measure Technique for Image Points of LMCCD Imagery,TP391.41
  15. 3D Reconstruction from Image Sequences,TP391.41
  16. The Research on Real-Time Performance of Character Recognition Technique Based on Machine Vision,TP391.41
  17. On-machine 3D Measurement of Large-scale Workpiece Based on Binocular Vision,TP391.41
  18. Research and Design of Drinks after Filling Impurities Inspection System Based on Machine Vision System,TP274
  19. The Matting Technology Base on Dynamic Light Field,TP391.41
  20. The humming songs retrieval system - based research and design,TP391.3

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