Dissertation > Excellent graduate degree dissertation topics show

Research on Track-before-Detect Algorithm Based on Particle Filters

Author: ChenZuo
Tutor: JiHongBing
School: Xi'an University of Electronic Science and Technology
Course: Signal and Information Processing
Keywords: Particle filter Dim target Pre-test track Marginalized Quasi - Monte Carlo
CLC: TP391.41
Type: Master's thesis
Year: 2009
Downloads: 369
Quote: 3
Read: Download Dissertation


Weak target detection and tracking infrared early warning system, a key technology for precision guidance system, satellite remote sensing systems. Under the influence of the long-distance attenuation and strong noise, the sensor receives the target very low signal-to-noise ratio, the traditional target detection and tracking method has been very difficult to meet the requirements. Recent years Pre-test track (TBD) an effective way to solve this problem, this method in one set of detection and tracking, and take full advantage of sensor data without threshold processing based on accumulated through time goal energy, thereby improving the signal-to-noise ratio, dim target detection and tracking. Based particle filter (PF) the TBD algorithm superior performance, but the particle filter generally requires a lot of random samples to ensure their performance, while a large number of random samples of the forecast, updated and re-sampling to calculate the particle filter is difficult to meet the requirements of real-time on the project . This paper focuses on using a variety of techniques to reduce the computational burden and improve the real-time nature of the algorithm based on particle filter TBD algorithm. First, through the analysis of infrared dim target model proposed a marginalized particle filter based TBD algorithm. The algorithm is characterized by marginalization method, the target speed linear Gaussian target state features state separation, and its use linear optimal Kalman filter, target position, intensity and nonlinear state is still with the particle filter processing. This not only reduces the dimension of the particle filter estimated state, greatly reducing the amount of computation, but also improve the detection performance and tracking accuracy of the algorithm in the low signal-to-noise ratio. Second, the faster convergence with an error intended to replace the traditional Monte Carlo (MC) particle filter integration method Monte Carlo (QMC) integration, and propose an improved algorithm: The proposed Monte Carlo-based Gaussian particle filter (QMC- GPF). Structured distributed sample points of the QMC integral with fewer the MC integrator of precision, the algorithm can guarantee the accuracy of the premise saves a lot of computational burden. Finally, in the QMC-GPF algorithm based on the convergence properties of the filter state covariance matrix tracking process to build judgment logic to achieve target detection. The algorithm is simple, small amount of calculation, simulation and experimental display of measured data, the algorithm has a good track of more than 3dB target detection capability.

Related Dissertations

  1. Research on Cooperative Orbit Determination in Satellite Network Based on Multi-Agent System Theory,V474
  2. The Maneuvering Target Tracking Research Based on VRPF,TN957.52
  3. The Research of 3D Human Motion Capture Based on Reference Points,TP391.41
  4. Research on the Theory of Visual Object Tracking Based on Particle Filter for Autonomous Robot,TP242
  5. Space Infrared Target Simulation and Application of Target Trace,TP391.41
  6. Detection and Tracking of Moving Object in Complex Background,TP391.41
  7. Moving Objects Detection and Tracking Using Fisheye Camera,TP391.41
  8. The Research on the Target Localization and Tracking Based on WSN,TN929.5
  9. The Study of Moving Object Detection and Tracking Algorithm Based on Image Information,TP391.41
  10. The Research on Localization and Target Tracking in Wireless Sensor Network,TN929.5
  11. AUV Integrated Navigation Algorithm Study and System Implementation,U666.1
  12. Architectural Optimizations for Particle Filters,TN713
  13. Wheeled humanoid robot navigation and path planning,TP242
  14. Petri net -based network intrusion detection system Research and Implementation,TP393.08
  15. GPU-accelerated particle filter PET image reconstruction algorithm,TP391.41
  16. Particle filter based multi-component FM fixed distance reconnaissance signal separation and parameter extraction,TN911.7
  17. Based on mid-level visual features and high-level structural information complementary target tracking model,TP391.41
  18. Target Tracking Based on Particle Filter Algorithm and DirectShow realized,TP391.41
  19. Based on Particle Filter and reliable visual tracking technology,TP391.41
  20. Distributed Microphone Array Tracking Algorithm,TN912.3
  21. Based on the theory of multi-target tracking FISST Research,TN953

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