Dissertation > Excellent graduate degree dissertation topics show

Investigation of Image Matching Algorithm Based on Feature Points

Author: TangYongHe
Tutor: LuHuanZhang
School: National University of Defense Science and Technology
Course: Information and Communication Engineering
Keywords: Feature point extraction Image matching DOG SIFT Imp-MOPs Feature descriptor
CLC: TP391.41
Type: Master's thesis
Year: 2007
Downloads: 1137
Quote: 17
Read: Download Dissertation

Abstract


The image matching is to having the two or more images of the same scene is aligned in space , and thus determines the process of the transformation relationship between them , and these images may be at different times and with different sensors , taken from different perspectives down . The small number of feature points target tracking , image navigation , text recognition , resource analysis , face recognition , and computer vision field , one of the key issues that need to be addressed is the image matching using contains image important structural information to determine the transformation between images the relationship is an effective method to solve this problem . Difficulties of this approach is how to extract a stable feature point , and Constuction rotating , deformation , obscure, feature descriptor of the noise and other factors . On the basis of this paper to analyze and summarize the feature point extraction algorithm and the characterization of the sub- generation method , a fast image matching algorithm based DOG feature points . First simplify DOG feature point extraction algorithm , and to set the appropriate parameters to ensure that the number of feature points and stability , then build a 25 -dimensional rotation invariant feature descriptor , then the descriptor similarity determining initial match point , and according to the corresponding the affine transformation of the segment invariance excluded error matching points, in order to achieve the matching of the two images . Finally, in terms of image transformation , noise , image blurring , image compression , brightness variations , smaller viewing angle changes and scale scaling of the proposed algorithm , SIFT Imp - MOPs simulation experiments , and their properties were compared , the results show that match the performance of the new algorithm with SIFT basically the same , but has greatly improved than Im-MOPs , the processing speed is faster than the other two algorithms about doubled .

Related Dissertations

  1. A Study on the Technology of the Infrared Image Matching,TP391.41
  2. Effects of Chinese Medicine Herbs Jiegusan on Fracture Healing in Dogs,S858.292
  3. The Establishment and Application of cPL Double Anti-Body Sandwich Elisa Method for the Diagnosis of Canine Acute Pancreatisis,S858.292
  4. Study on the Anesthestic Quality, Blood Concentration and Clinical Application of Shumianning Administered Intravenously by Continuous Rate Infusion Using Microinjection Pump in Dogs,R965
  5. The Genetic Diversity of 12 STR Loci in Four Dog Breeds and the Application in Individual Discrimination,S829.2
  6. GC-MS Analysis of Endogenous Metabolites in Plasma of Police Dog after Aflatoxicosis,S858.292
  7. Study on Configuration, Function and Histopathology of Compensatory Kidney in Dog,S858.292
  8. Expression and Significance of HIF-1α on the Insulin Resistance during the Injury of Reperfusion of Ischemic Myocardium in Dog Undergoing Cardiopulmonary Bypass,R654.1
  9. Research on Image Matching Based on Graph Cut,TP391.41
  10. The Research of Image Matching Method Based on Feature Descriptor,TP391.41
  11. Research on Subimage Selection and Mathching Method for Synthetic Aperture Radar(SAR) Target Recognition,TN957.52
  12. The Research of Phishing Detection Technology Based on Nearest Neighbor and Similarity Measurement,TP393.08
  13. SAR image can be matched study,TN957.52
  14. GPU-based SIFT algorithm,TP391.41
  15. SIFT-based digital watermarking algorithm,TP309.7
  16. BoW-SIFT -based model and the level of the three-dimensional mesh feature retrieval system,TP391.41
  17. Based on the geometry of multiple images and textures automatically rebuild,TP391.41
  18. Helmet based on image processing spatial position measurement,TP391.41
  19. Saliency -based image retrieval keying Research and Implementation,TP391.41
  20. SVM Based on SIFT and scene classification,TP391.41
  21. Image registration based on the rotation angle calculation methods sand,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