Dissertation > Excellent graduate degree dissertation topics show

Research on Methods for Pedestrian Association Across Non-overlapping Camera Views

Author: LiBo
Tutor: LiGuoHui
School: National University of Defense Science and Technology
Course: Systems Engineering
Keywords: Non-overlapping cameras Target detection Feature matching Target association Feature Fusion
CLC: TP391.41
Type: Master's thesis
Year: 2011
Downloads: 15
Quote: 0
Read: Download Dissertation

Abstract


The target tracking non-overlapping monitoring system is an intelligent monitoring system is an important research direction , and it refers to the integration of multi- surveillance cameras to track the trajectory of the target . Goals related issues in a non - overlapping surveillance cameras is a core issue for target tracking in intelligent monitoring system . This paper studies the target associated with the non - overlapping surveillance cameras , and conducted in-depth research goals associated characteristics extraction and matching . Target associated technology status , analyze its current problems and a breakthrough design the necessary pre-processing and post-processing of the target associated with the matching process , and a goal-based appearance features describing the model matching method . The main content of the paper include : 1 . Still camera moving target detection , based on background subtraction methods designed based color clustering model of the target body area to extract the basic framework by introducing color segmentation method solved the problem of a target detection in the method of the background subtraction incomplete , and improve the effect of the target feature extraction . Target feature extraction and matching step for disadvantage of traditional MCSHR method , proposed distribution model based on color space , the better to deal with the appearance of the differences between the different goals . Meanwhile, the reconstruction of the target based on texture pattern matching method , more in line with the characteristics of human cognition , and design a specific matching algorithm can effectively improve the accuracy of target matching , rich characterization of the target appearance . 3 . Traditional goals associated method , did not make good use of the shortcomings of monitoring the video continuous image of the target characteristics , based on Adaboost the continuous target image sequence target associated methods . The use of a continuous target image feature training Adaboost combination of classifier, and for surveillance cameras in the non-overlapping target association , improving the accuracy of the target associated . Combination of the above methods , through experiments associated with moving target surveillance video . Choose a different sight surveillance video moving target feature extraction to match , and the associated results verify the rationality and effectiveness of the proposed method .

Related Dissertations

  1. Research on Joint Target Detection for Dual-Sensor Image and System Implementation,TP391.41
  2. Feature Extraction, Selection and Combination in Lipreading,TP391.41
  3. Moving target trajectory analysis based Intelligent Traffic Monitoring System,TP277
  4. The Study of Moving Object Detection and Tracking Algorithm Based on Image Information,TP391.41
  5. Research on Key Technologies of Monocular Visual Odometry Based on DSP,TP391.41
  6. Research on Moving Object Detected Technique for PTZ Cameras,TP391.41
  7. 360 degree rotation based on single- camera tracking system,TP391.41
  8. Monitoring image ROI extraction and target detection technology,TP391.41
  9. Distributed target detection in wireless sensor network research,TP212.9
  10. Face recognition algorithm based on feature fusion research,TP391.41
  11. The Compression and Fusion Technique Research of Underwater Target Feature,TN911.7
  12. Research on Two-dimensional Feature Extraction Methods,TP391.41
  13. Research on Color Face Recognition Based on Fusion of Components Feature Extracted by Spectroface and FLD,TP391.41
  14. Human Emotion Recognition Using Fused 2D-Fractional Fourier Transform Features Based on CCA,TP391.41
  15. Research on Moving Object Detection and Tracking Algorithm Based on Local Feature Analysis,TP391.41
  16. X-ray Medical Images Stitching Based on Corner Matching Using Two-way Asynchronous Voting Strategy,TP391.41
  17. Based on face and fingerprint multi- modal identification system research,TP391.41
  18. Camera-based network monitoring system,TP277
  19. Moving Target Detection Research on Key Issues,TP391.41
  20. Research on Multiple Objects Tracking Based on Detection in Video Sequence,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