Dissertation > Excellent graduate degree dissertation topics show

Research on Visual Tracking Algorithm Based on SIFT

Author: LiuYang
Tutor: YaoLiXiu
School: Shanghai Jiaotong University
Course: Control Engineering
Keywords: Visual Tracking Scale Invariant Feature Transformation Random Sample Consensus Mean-shift algorithm Bag of words Support Vector Machine Point Tracking
CLC: TP391.41
Type: Master's thesis
Year: 2012
Downloads: 386
Quote: 0
Read: Download Dissertation

Abstract


With the development of science and technology, visual tracking has attracted the attention of many researchers. Visual tracking is one of the key technologies used in video monitoring, navigation and many other applications. So it is widely used in these applications and also brings great prestige to the society. Although researchers have come up with a lot theories and methods, visual tracking still faces huge challenges since the actual scene is complex. Based on the existed methods, visual tracking still needs to be studied and improved.Modeling is one key technology for visual tracking and it can affect the tracking results directly. Recently, Scale invariant feature transform (SIFT) as a robust feature for scale, rotate, blur and illumination has drawn many researchers’attentions. Some of them have used SIFT to improve the target’s model. The registration based on SIFT can reach sub-pixel accuracy, so we can improve the accuracy of tracking by using SIFT.This paper introduces some common features, especially SIFT and also summarize the tracking algorithms. Based on these theories, tracking algorithm is further studied and applied into two programs in this paper. By combining the target tracking and target recognition, a SIFT based target tracking and identification method is proposed. Tracking algorithm models the target with its appearance model and SIFT feature. Then it uses the probabilistic voting method and mean-shift method to estimate the object’s optimal center. The recognition part adopts the SIFT feature to construct bag of words model. After that, using support vector machine algorithm to get a recognizer for each target by training the sample. The whole algorithm’s first step is extracting foreground by using Gaussian Mixture Model. Then recognize the foreground and track each target to get the optimal position.A point tracking method based on SIFT and HOG is proposed when doing a tracking project. After matching the front and back images’SIFT feature, a homography matrix is obtained. Map the tracking point to the second image to get the candidate point. In order to improve the accuracy, random sample consensus algorithm is introduced to get a more accurate matrix. HOG feature is applied to precisely locate the candidate point. At last, update the point’s SIFT and HOG feature to eliminate the offset caused by the changed environment.By introducing the algorithm of point tracking based on SIFT into the program of detecting object’s speed, this paper presents a speed measurement algorithm based on SIFT. The algorithm gets the target’s moving distance by matching two images.

Related Dissertations

  1. Research on Automatic Detection Algorithm for Substructure Distress of Highway Pavement Based on SVM,U418.6
  2. Research on Autamatic Music Structrue Analysis,TN912.3
  3. Research on Transductive Support Vector Machine and Its Application in Image Retrieval,TP391.41
  4. Fault Diagnosis Method Based on Support Vector Machine,TP18
  5. Process Support Vector Machine and Its Application to Satellite Thermal Equilibrium Temperature Prediction,TP183
  6. The Grid-Connected Wind-solar Hybrid Generation System and Maximum Power Point Tracking,TM61
  7. Research for Infrared Image Target Identification and Tracking Technology,TP391.41
  8. Study on the Road Condition Monitoring Based on Vehicular 3D Acceleration Sensor,TP274
  9. Research of Diagnosing Cucumber Diseases Based on Hyperspectral Imaging,S436.421
  10. Research on Whispered Speaker Identification in Channel Mismatch Conditions,TN912.34
  11. Scholar Resume Automatic Generation Based on Text Mining,TP391.1
  12. Research on Intrusion Detection Based on Feature Selection,TP393.08
  13. Bag-of-Visual Words for Image Categorization,TP391.41
  14. The Research on Intrusion Detection System Based on Machine Learning,TP393.08
  15. Study on Visual Target Detection Based on SVM,TP391.41
  16. Research on the Algorithms of Moving Object Detection and Tracking,TP391.41
  17. The New Privacy Preserving Algorithms for Linear Programming and SVMS,O221.1
  18. Dimensionality Reduction Methods for Gene Expression Data Base on SVM,TP181
  19. Prediction of Binding Affinity of Human Transporter Associated with Antigen Processing,R392.1
  20. Study on Applications of Soft Sensor in Chloromethane Recovery,TQ222.214
  21. Research of Fault Diagnosis Method of Analog Circuit Based on Improved Support Vector Machines,TN710

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