Dissertation > Excellent graduate degree dissertation topics show

Real-Time Image Mosaic Based on CUDA

Author: GuoYiHan
Tutor: ShiMeiPing
School: National University of Defense Science and Technology
Course: Control Science and Engineering
Keywords: Ground mobile robots Big vision remote control navigation Real - time image stitching Target point positioning GPGPU CUDA RANSAC
CLC: TP391.41
Type: Master's thesis
Year: 2011
Downloads: 56
Quote: 0
Read: Download Dissertation

Abstract


Remote navigation technology is a key technology for terrestrial mobile robotics. Remote navigation core issue is how to allow the operator to obtain real-time, large field of view environmental information, in order to ensure real-time remote navigation and transparency. A feasible way to solve this problem is the use of image stitching. Image stitching process, however, a large quantity of data related to the image matching, compute-intensive, so the many serial processing method in the practical application is difficult to meet the real-time requirements. To this end, the remote control ground mobile robot navigation applications for background, from both theoretical and practical aspects based on CUDA (Compute Unified Device Architecture, unified the device computing architecture) the real-time image stitching technology, in-depth study, including the combination of a priori information adaptive image matching method solution, based on the CUDA real-time image stitching algorithm and its application in mobile robot navigation in remote image transformation model based on improved RANSAC (Random Sample Consensus, random sample consensus). The papers completed work: 1) demand for real-time image stitching applications, adaptive image matching method is a combination of a priori information. Firstly, according to the overlapping area range known in advance to be mosaic image, in the overlapping region of the left image uniformity select to be matched point; then adaptive matching by rotational invariance regional similarity measure, in the right image search area search of the best match point. The experimental results show that this approach can significantly improve the efficiency and accuracy of image matching. Point) for image matching process inevitably exist mismatching, in-depth study on the basis of the RANSAC algorithm by introducing random sampling constraints and optimal sample set of evaluation criteria are given based on the improved RANSAC image transformation model algorithm. The experimental results show that the algorithm can effectively improve the correct rate and accuracy to solve the image transformation model. 3) In-depth analysis of the CUDA programming model based on a best match based on the GPU CUDA-based real-time image stitching algorithm (CUDA-RTIM), and take full advantage of the GPU's parallel computing ability, given a point parallel search algorithm, as well as projective transformation model based on improved RANSAC parallel algorithm. The experimental results show that the the CUDA-RTIM algorithm not only has good robustness and real-time, and also has a good adaptability to various environments. On the basis of the above work, the paper design to achieve a ground mobile robot remote control system based on a large field of view, and Pioneer-III-type robot experiment platform, to verify the feasibility and effectiveness of the above methods and algorithms.

Related Dissertations

  1. Study of Binocular Stereo Matching Algorithm and Algorithm Implementation Based on Multi-core.,TP391.41
  2. Research of Finite Element Method on GPU,O241.82
  3. Implement and Research of High-Quality Global Illumination on GPU,TP391.41
  4. Study on Real-time Screen Space Fluid Rendering,TP391.41
  5. White Light Interferometry for Fast Areal Surface Measurement Based on GPGPU,O439
  6. Research groups and evacuation simulation algorithm simulation system,TP391.9
  7. GPU-based book recommendation system Research and Implementation,TP391.3
  8. CUDA-based video fire detection system,TP391.41
  9. SURF algorithm based on improved electronic image stabilization technology,TP391.41
  10. GPU-based reconstruction algorithm for X-rays accelerate research,TP391.41
  11. Exam cheating Intelligent Analysis Method,TP391.41
  12. GPU-based acceleration of neutral gas leak and rescue simulation study,TP391.41
  13. Image Matching Based on Local Invariant Features,TP391.41
  14. CPU GPU -based heterogeneous platforms string matching algorithm and implementation,TP301.6
  15. A Study on Diagonal Computing Model for GPGPU Platform,TP391.41
  16. GPGPU and Image Matching Parallel Algorithm Based on SIFT,TP391.41
  17. High - performance computing systems based on particle simulation of the GPU,TP338
  18. Research of Acceleration Photon Mapping Based on GPU,TP391.41
  19. Research on Accelerating Krylov Subspace Method with GPU,TP391.41
  20. Lunar Soft-landing Vision Navigation and Descending Sequential Imagery Match Research,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