Dissertation > Excellent graduate degree dissertation topics show

Research on Segmentation-Based Stereo Matching Algorithms

Author: ChenChangBing
Tutor: LiuZuo
School: Shandong University
Course: Communication and Information System
Keywords: Stereo matching Image Segmentation Adaptive weighted Energy function Plane fitting
CLC: TP391.41
Type: Master's thesis
Year: 2011
Downloads: 90
Quote: 0
Read: Download Dissertation

Abstract


The three-dimensional structure of stereo vision as an important branch in computer vision, is responsible for recovering from the two-dimensional image of the camera acquisition of natural scenes, 3DTV, multi-view / free viewpoint video, automotive intelligent driver assistance, military, virtual reality, there are very extensive background, has a high commercial value and use value. Get information about the depth of the objects in the scene is the core issue in stereo vision, this work is mainly done by the stereo matching algorithm. The stereo matching technology originated earlier, but still a research focus in the field of pattern recognition and artificial intelligence. Learning stereo matching principle and summarize existing algorithms based on the two improved stereo matching algorithm, the main results of the study are as follows: 1) proposed a local stereo matching algorithm based on adaptive weighted and adaptive window . The local algorithm support windows as matching primitives, enhanced pixel can be distinguished. According to the magnitude of the contribution to the supporting window of the pixel assigned with different weights; contribution greater the higher the weight, the smaller the contribution weight lower. In this paper, a new weight calculation functions, integration of the adaptive support windows and image segmentation information, and can effectively extract the \within these regions the parallax; given the specific steps of the algorithm, and in the final simulation experiment. 2) an improved global stereo matching algorithm. The global energy function typically contains a data item and smooth entry. The image data item represents a pixel in a parallax image with another one corresponding to the pixel points of the match consideration. Distinguishable characteristics of the pixels in order to better meet the pixels on the true degree of similarity matching cost, the need to enhance, support the right of the value of the pixels within the window has not, which is connected with the thinking of the third chapter. The smoothing term integration of a priori knowledge of the disparity map segmentation smooth. The parallax from the same surface is generally smooth transition, while the parallax is not a continuous boundary is mostly distributed on the edge of the object. Assumptions based on image segmentation, segmentation within the block of pixels belonging to the same object, the edge of the distribution of the object edge of the divided blocks, and so to increase the the divided block parallax discontinuous punishment. This paper constructs a new energy function, Chapter of the use of data items matching the cost of the smooth fused image segmentation information, given the multi-label binary label conversion process and build a directed graph model. Finally, the method of detection and improve the parallax diagram singular point of.

Related Dissertations

  1. Research on Image Recognition Algorithm in the Forest Fire Prevention System,TP391.41
  2. Application Research of Digital Image Processing on Container Inspection,TP274.4
  3. Study on the Design and Biomechanical Behaviors of Individualized Artificial Knee Joint,R318.1
  4. Based on statistical methods of magnetic resonance imaging of the human brain image segmentation and three-dimensional analysis of the data,R445.2
  5. Road extraction algorithm based on region segmentation of remote sensing image,TP751
  6. Tobacco Diseases Auto-Recognition Research Based on Image Processing Technology,S435.72
  7. Research for Infrared Image Target Identification and Tracking Technology,TP391.41
  8. BF-FCM Clustering Algorithm and Its Application in the Image Segmentation,TP391.41
  9. Three pairs of atomic and molecular systems potential energy function and spectral properties of,O561.3
  10. Research on Image Matching Based on Graph Cut,TP391.41
  11. Segmentation of cDNA Microarray Image Using Fuzzy C-means Algorithm Optimized by Particle Swarm,TP391.41
  12. The Study of Human Face Detection Method Based on OpenCV,TP391.41
  13. Computer-assisted hip replacement surgery systems research and development,TP391.41
  14. The Depth Estimation Based on Monocular Night Vision Image,TP391.41
  15. Research on Image Segmentation Method Based on Measure of Medium Truth Degree,TP391.41
  16. The Research on Determination Technology of Cotton Maturity Based on Image Processing,TP391.41
  17. Study of Binocular Stereo Matching Algorithm and Algorithm Implementation Based on Multi-core.,TP391.41
  18. Spatial Temporal Refinement on Depth Video Processing,TP391.41
  19. Integration of Spatial Information Bag of Feature in Image Annotation,TP391.41
  20. Research on Pavement Crack Detection Algorithom Based on Digital Image Processing,TP391.41
  21. Research of Image Segmentation in Web Image Search Based on GPUs,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