Dissertation > Excellent graduate degree dissertation topics show

Research on Video Object Tracking Technology Based on Mean-Shift Algorithm and Kalman Filter

Author: HuBo
Tutor: ChenKen
School: Ningbo University
Course: Signal and Information Processing
Keywords: Target tracking Mean shift Kalman filter Bhattacharyya coefficient Camshift
CLC: TP391.41
Type: Master's thesis
Year: 2010
Downloads: 438
Quote: 1
Read: Download Dissertation

Abstract


Video target tracking is a hot issue for current computer vision research areas, and has a wide range of applications in many aspects of the video surveillance, robotic vision, navigation, military guidance, medical diagnostics, and human-computer interaction. Video tracking is to detect the moving object in the video sequence, extraction, recognition and tracking process, it can provide a basis for the next video analysis and understanding. In the actual scene, target tracking is often subject to their own state of motion, obstructions, target deformation, illumination changes, background noise and other factors become very difficult. Although many scholars studied the video target tracking, but there are still more difficulties for any complex environment of video object tracking algorithm to develop a set of. This article first study of the moving target detection technology frame difference method and background updating method. Focus on target detection based on Mean Shift algorithm and Kalman filter tracking of moving targets, and improved algorithm according to the actual situation. 1. Tracking method presents a combination of Kalman filtering theory and adaptive Mean Shift algorithm to solve the target deformation, partial occlusion, motion too fast. First, in the initial frame, to determine the tracking target and calculate the H-component histogram, the probability of each frame image is converted into the histogram projection view. In the starting position of the current frame using Kalman filter forecast target, then the adaptive Mean Shift algorithm tracking video object. The rigid, non-rigid body and multi-target tracking has good adaptability. 2. Proposed video object tracking method Bhattacharyya coefficient maximization and joint airspace information. Time domain by Kalman filtering prediction target movement information, airspace Camshift algorithm exact match video goals. Due to strong mobility of moving targets, the position and the real position of the Kalman filter predicted larger error, easily lead to the next step tracking failure. In this paper, based on the the Bhattacharyya coefficient from coarse to fine kernel matching search method, adaptive search window at the position on the basis of the Kalman filter forecast to widen the search to determine the initial matching window Bhattacharyya coefficient maximization, and then Camshift algorithm exact match video goals. The method for maneuvering fast moving target having a high tracking accuracy.

Related Dissertations

  1. Algorithm Research on SINS/GPS Integrated Navigation Syste,V249.328
  2. Satellite Attitude Determination Based on Gyro and Star Sensor,V448.2
  3. Research on Transfer Alignment of the Missile Onboard the Aircraft,V249.322
  4. Maneuver Detection and Tracking with the Radar Rate Measurement,TN953
  5. The Maneuvering Target Tracking Research Based on VRPF,TN957.52
  6. Research on Multi-Sensor Embattling and Continious Tracking Technique Applying for Anti-Stealth and Anti-Interference,TN953
  7. Research on Visual Detection and Tracking of Mobile Robots,TP242.62
  8. Design of Gyro Stabilized Pod Control System,TP273
  9. Tracking Cells in High Density Image Sequences Based on Mean Shift Algorithm Combined with Topological Constraint,Q25
  10. Vehicle detection based on machine vision and vehicle distance measuring method,TP274
  11. Research and Application on Detection and Tracking of Object Based on Video Stream,TP391.41
  12. Research on Object Detection and Tracking Method in Active Vision System,TP391.41
  13. AUV Integrated Navigation Algorithm Study and System Implementation,U666.1
  14. Depth map and color images based on treadmill game interaction system,TP391.41
  15. Structural constraints of the dynamic PET image reconstruction,TP391.41
  16. Based on embedded image tracking system,TP391.41
  17. Smart Home Monitoring Robot,TP273.5
  18. Driverless smart car moving target detection and tracking,TP391.41
  19. Video Mean-Shift tracking algorithm applied research,TP391.41
  20. Mobile robot vision -based target search and tracking technology,TP391.41
  21. Based on a combination of multi-sensor automated guided vehicle systems research and application,TP242.2

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