Dissertation > Excellent graduate degree dissertation topics show

Multi-part Cooperative Tracking Algorithm Research on Human Target in Aquatic Sports

Author: LiFei
Tutor: ZhangQi
School: Dalian Maritime University
Course: Pattern Recognition and Intelligent Systems
Keywords: Multi-part Cooperative Tracking Immersion Occlusion Meanshift Algorithm Aquatic Human Target
CLC: TP391.41
Type: Master's thesis
Year: 2012
Downloads: 23
Quote: 0
Read: Download Dissertation

Abstract


As an important branch of the computer vision,moving target tracking is widely used in security project,taffic surveillance,etc.In recent years,due to the popularity of aquatic sports and the enhancement of safety consciousness,more and more experts and scholars pay attention to the aquatic human tracking.Because the background environment and the target movement forms in aquatic human tracking is different from the human tracking on the ground,its occlusion problem has distinctive features.The tracking effect is not ideal in the water environment which uses the tracking algorithm apply to the land environment.The focus of this paper is how to deal with the interference in the complex water environment,solves the problem of immersion occlusion,guarantees a good tracking effect.According to the complex water environment and the characteristic of Meanshift algorithm,this paper presents a number of preprocessing method of the image pixels and discusses the parameter selection of Meanshift algorithm thoroughly through the comprehensive analysis of water video.On this basis,according to a large number of experiments,this paper obtains the law of immersion occlusion and provides a new thinking for the aquatic human tracking.For the situation of inaccurate tracking with the Meanshift algorithm,this paper adopts the method of tracking the 3 parts of human to solve the problem.According to the analysis of the law of immersion occlusion,this paper obtains 7 conditions of influencing the tracking.This paper uses the constraints of spatial position and the time domain to give the correction method respectively,solve the effect of immersion occlusion and improve the tracking accuracy.This paper takes the software realization of the multi-part cooperative tracking algorithm after several tests,it proves that the multi-part cooperative tracking algorithm compares with the Meanshift algorithm has improved in the tracking length of time and accuracy.The multi-part cooperative tracking algorithm can solve the problem of immersion occlusion effectively and improve the tracking effect.

Related Dissertations

  1. Moving Object Recognition and Opto-Electronic Tracking Location Technology,TP391.41
  2. The Research of Human Video Tracking Based on MeanShift Algorithm,TP391.41
  3. Semi-Supervised Color Image Segmentation Research and Application,TP391.41
  4. Multi-moving Targets Tracking Based on the Video,TP391.41
  5. Image processing technology based mobile robot target tracking,TP242.62
  6. Research on Algorithm of Object Tracing and Detecting for Video Surveillance System,TP391.41
  7. Research on Tracking for Cell Based on Video Technology,TP391.41
  8. Detection and Tracking of Objects for Inspection Robot of Overhead High Voltage,TP242
  9. Detection, Tracking, and Orientation Analysis of Human Body Based on Video Image,TP391.41
  10. Complex Background for the Rapid Detection and Tracking of Targets,TP391.41
  11. Design and implementation of intelligent monitoring system based on face detection and tracking,TP391.41
  12. Object Detection and Tracking in Video Image,TP391.41
  13. Research on the Algorithms of Moving Object Detection and Tracking,TP391.41
  14. The Research and Implemention of Image Retrieval Based on User Interested Feature,TP391.41
  15. Research of Image Mosaic Technology,TP391.41
  16. Research on Infrared Image Simulation for Aerial Objects and Background,TP391.41
  17. Research on Transparent Security of JPEG Image,TP391.41
  18. Research on Image Recognition Algorithm in the Forest Fire Prevention System,TP391.41
  19. Research on Hyperspectral Image Compression Method Based on Information of Interest,TP391.41
  20. Research on Feature Extraction and Classification of Tongue Shape and Tooth-Marked Tongue in TCM Tongue Diagnosis,TP391.41
  21. Study on Virtual Detector of Infrared Hyper-Spectral Image,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