Dissertation > Excellent graduate degree dissertation topics show

Real-time Three Dimensional Information Obtaining System Based on Camera Array

Author: ZhouYing
Tutor: YuHuiMin
School: Zhejiang University
Course: Information and Communication Engineering
Keywords: Stereo matching Demarcate Camera array TOF Multithreading GPU
CLC: TP391.41
Type: Master's thesis
Year: 2010
Downloads: 137
Quote: 2
Read: Download Dissertation

Abstract


In recent years, real-time three-dimensional access to information technology has become a very hot research topic in the field of computer vision. Digital three-dimensional model of the object in the the archaeological artefacts modeling, city models digitized Entertainment figures scene modeling, reverse engineering precision molds and other areas have broad application value, quickly and accurately obtain the object space is three-dimensional information object modeling the necessary conditions. There are two main ways to obtain three-dimensional information of the object currently: to strike a the images match point parallax and depth stereo matching method as well as the recent TOF technology based on infrared sensors. Both techniques have their advantages and shortcomings: TOF technology can generate real-time depth map and texture missing area can get good results, but the low resolution image. Stereo matching can generate high-resolution depth map, but requested that the scene richly textured. These two technologies are done in-depth research and analysis of the characteristics of each of the two technologies, heterogeneous construct an ordinary camera with TOF camera camera array, and a combination of the two technologies through data fusion the merits of a complete system to obtain three-dimensional information in real-time, the main work is as follows: (1) construct a heterogeneous camera array contains more than ordinary CCD sensor camera with a TOF camera, and acquired by plane template between the parameters within each of the cameras and the CCD camera external parameters, using the fusion method of calibration of the depth information obtained TOF camera and a CCD camera between external parameters, thereby completing the entire calibration of the camera array. (2) mapping the original depth-map of the TOF of the camera to the CCD camera view, constructed under the specified camera vision initial depth map, combined with the initial depth and color images, the use of segmentation method of locking the outline of objects in the scene to strike the edge of the edge, by a stereo matching method parallax depth at the edge of the contour of the object. (3) the initial depth map and the edge depth map integration CCD camera view of the complete depth map and using bilateral filtering method to obtain a final quality depth map. (4) the algorithm described in the text on the system a realization, a complete three-dimensional information access to the system is constructed, including image acquisition, processing, and final results in Fig multi-perspective display section, and the use of multi-threading technology and GPU parallel processing The method of acceleration so that the system can be output in real time three-dimensional information of the scene to reach the subject indicators expected.

Related Dissertations

  1. Implementation and Research of Illegal Websites Detection System Based on Comparison Methods,TP393.08
  2. Visual Feedback and Memory Behavior Based GPU Parallel Ant Colony Algorithm,TP301.6
  3. Without contrast agent -dimensional TOF MR angiography with France multidetector CT angiography in the diagnosis of lower extremity arterial disease comparative study,R816.2
  4. Detection of modified nucleosides in urine of bladder transitional cell carcinoma and its biological behavior of,R737.14
  5. Mushroom C 91-3 fermentation broth preliminary identification of protein purification and anti-tumor mechanism,R730.5
  6. Studies on Screening Determination of Multiclassification Multipesticide Residues in Vegetables,S481.8
  7. Analysis of Antenna Protein of Silk Moth by Two -dimensional Electrophoresis,Q966
  8. Study of Binocular Stereo Matching Algorithm and Algorithm Implementation Based on Multi-core.,TP391.41
  9. Multi-view 3D Reconstruction and Application Based on Depth Fusion and Surface Evolution,TP391.41
  10. Spatial Temporal Refinement on Depth Video Processing,TP391.41
  11. Research on Carrying-Saddle Measure System Based on Structured Light 3D Measurement,TP274
  12. Research of Finite Element Method on GPU,O241.82
  13. The Design and Implementation of C/S-Based Vehicle Remote Monitoring System,TP277
  14. Research of Image Segmentation in Web Image Search Based on GPUs,TP391.41
  15. Dynamic analysis based on multi-core acceleration method,TP332
  16. GPU-based SIFT algorithm,TP391.41
  17. GPU-based image search Chinese Research on key technologies of the retrieval,TP391.1
  18. Based on GPU / CPU multi- level parallel CFD Optimization Strategy,V221
  19. Ffmpeg-based high-performance high-definition streaming media player software design,TN919.8
  20. Flame effects Augmented Reality System Key Technologies,TP391.9
  21. Efficient multi- GPU based acoustic wave simulator and its application,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