Dissertation > Excellent graduate degree dissertation topics show

Research on Passenger Counting Unit Algorithm and Its Implementation in TMS320DM642 System

Author: LiYinGe
Tutor: BiSheng
School: Dalian Maritime University
Course: Information and Communication Engineering
Keywords: Passenger Counting System Embedded System Binocular Stereo Vision SEED-VPM642 Development Board
CLC: TP391.41
Type: Master's thesis
Year: 2011
Downloads: 54
Quote: 2
Read: Download Dissertation

Abstract


In recent years, the automatic passenger counting system is being developed and gets a preliminary application in the field of scientific research. Many people research on the passenger counting technology based on monocular image processing, but most are limited to obtain two-dimensional information of images to extract head features as the primary means of identification and tracking passenger flow. Then, the number of passenger is counted by model matching. This paper design the passenger counting algorithm based on binocular stereo vision, and be implemented in the embedded systems.This system is based on the SEED-VPM642 development board and two CCD cameras, which uses the binocular stereo method to recovery depth information of moving objects in the image, and combines with some two-dimensional features to realize passenger counting. Two-dimensional features include area of the body target, the ratio of width to height, and centroid coordinates of the target area. Binocular stereo vision technology is roughly divided into several parts as follows:camera calibration, left and right images acquisition, feature extraction, stereo matching and three-dimensional information recovery. Stereo matching is the most important and difficult part of the binocular vision technology.In this paper the passenger counting unit algorithm is designed by analyzing and researching related algorithms, which can be completed by several parts as follows: Firstly, the target area in video sequences are obtained by segmentation; Secondly, feature points within target area are extracted by Harris corner detection, and then the points in left and right images are matched through region-based matching method; Finally, according to the principle of binocular vision, the value of disparity and depth of points matched successfully is calculated, which combines with two-dimensional feature parameter values of human body. Then, the number of passenger is counted.The algorithm of passenger counting in this paper is verified by the experiment, and the statistical average accuracy rate is up to about 92%, because the algorithm combines depth information to improve the detection accuracy and stability.

Related Dissertations

  1. The Electric Power Steering System,U463.4
  2. Gui Design and Application Oriented to Embedded Ultrasonic Detecting System,TP274.53
  3. Data Collecting and Processing of Multi-Linear-CCD Visual Measuring System,TP274.2
  4. Research on Temperature Control System for High-Performance Oven-Control Crystal Oscillator (OCXO),TN752
  5. Design and Realization of GPS Monitoring System Based on WEB,TP311.52
  6. Study on Ethernet MAC Protocol and Its FPGA Implementation,TP393.04
  7. Research of Security Computer Board Testing System Based on Embedded System,TP306.2
  8. Research and realization of dynamic performance analysis of code in the embedded environment,TP368.1
  9. Design and Implementation of GUI in Embedded Linux,TP316.81
  10. Linux Kernel Transplant and the Research and Implementation of USB Wireless Card Driver,TP316.81
  11. Transplant of Windows CE Operation System Based on ARM9,TP316.7
  12. Design and Implementation of Integration of Console Server and IPDU,TP311.52
  13. Design and Implementation of High Performance Encryption/Decryption System Based on FPGA,TP309.7
  14. The Modeling Optimization and the Algorithm of IO Jitter and Latency in the Hard Real-time Task Scheduling,TP316.2
  15. Research and Development of "Virtual Tennis" System Based on Computer Vision,TP391.41
  16. Study and Implementation of Game System Based on Windows CE Platform,TP311.52
  17. Study on Acquisition, Storage and Playback Afterwards of Smart Terminal Real-Time Image,TP391.41
  18. Teaching and Experiment Platform Study Based on ARM and Win CE,TP391.6
  19. Embedded software testing technology - oriented research,TP311.52
  20. Research on Algorithm of Digital Meter Indication Automatic Reading Based on ARM,TP391.41
  21. Design and Development about Embeded Electronic Medical Record Writing System Based on XML,TP311.52

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