Dissertation > Excellent graduate degree dissertation topics show

Pump Identification Based on Harris Corner Detection and SIFT Algorithm

Author: YuWeiLu
Tutor: XuXiao
School: South China University of Technology
Course: Optics
Keywords: Pump body identification SIFT Harris corner Remote maintenance
CLC: TP391.41
Type: Master's thesis
Year: 2011
Downloads: 62
Quote: 0
Read: Download Dissertation

Abstract


This topic is a sub-project of the vacuum pump remote maintenance system , mainly through image recognition technology to automatically identify the pump body object from a picture or video . A vacuum pump a wide range of different types of vacuum pumps have a completely different common fault , therefore, the pump identification is a vacuum pump remote maintenance system in the first step , is also a key step . Vacuum equipment is made ??of rigid material , is not easy to produce flexible deformation , the pump has a relatively stable edge feature Therefore , the subject of an edge extraction based on the combination of the Harris corner and SIFT features pump body recognition algorithm all excellent characteristics, i.e. , this method can use the edge can be overcome background interference , but also has the SIFT (Scale Invariant Feature Transform) characterized the rotation, scale scaling invariance . The entire recognition process introduced in detail : ( 1 ) has excellent characteristics of low error rate and high-precision positioning Canny operator edge extraction , can overcome the impact of the light on the identification of the feature points . (2) established on the basis of the edge information of each pixel point of the eight directions FIG extracted Harris corner . (3 ) a series of processes by Gaussian pyramid , Difference of Gaussian pyramid detect SIFT feature points . (4) using SIFT descriptor SIFT feature point will be described , to complete an exact match of the small number of feature points between the images of the templates, and to be identified and solved rotation scaling translation matrix . (5 ) Harris feature point matching rough translation matrix to achieve the final match , by rotating the zoom . (6) by matching feature points accounted for a percentage of all the characteristics of the template points purpose to develop a criterion for judgment . Finally, experiment results show . Through a series of generic experiments to prove reasonable criterion , the correct recognition rate up to 94 %, indicating that the availability of the method employed in this topic .

Related Dissertations

  1. Research of Image Mosaic Technology,TP391.41
  2. The Research of Image Matching Method Based on Feature Descriptor,TP391.41
  3. The Research on Mammographic Images Registration,TP391.41
  4. Research on Target Tracking Algorithms Based on Particle Filter Method,TP391.41
  5. The Research of Phishing Detection Technology Based on Nearest Neighbor and Similarity Measurement,TP393.08
  6. GPU-based SIFT algorithm,TP391.41
  7. SIFT-based digital watermarking algorithm,TP309.7
  8. BoW-SIFT -based model and the level of the three-dimensional mesh feature retrieval system,TP391.41
  9. Based on the geometry of multiple images and textures automatically rebuild,TP391.41
  10. Saliency -based image retrieval keying Research and Implementation,TP391.41
  11. SVM Based on SIFT and scene classification,TP391.41
  12. Image registration based on the rotation angle calculation methods sand,TP391.41
  13. Image Rendering Research of Image Guidance Simulator,TP391.41
  14. Multi-spectral remote sensing image registration and fusion research,TP751
  15. Research on Mobile Manipulator Localization Based on SIFT Algorithm Combining Binocular and Monocular Visual,TP241
  16. Moving Object Detection and Tracking in Dynamic Scenes,TP391.41
  17. Geometrically Robust Image Watermarking Algorithm,TP309.7
  18. Research on Algorithms to Reassign Labels to Regions,TP391.41
  19. Research on Video Object Tracking with C-SIFT Agorithm,TP391.41
  20. The Research and Application of the Unmanned Aerial Vehicle Image Matching Based on Feature Point,TP391.41
  21. Remote Sensing Image Retrieval Based on Radiation and Spatial Information,TP751

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