Study on the vision-based detection and tracking of cyclists and pedestrians

This paper proposes a vision-based system dealing with the detection and tracking of moving objects, especially for cyclists and pedestrians. The system is mainly composed of two units: the detection unit and the trackingunit. The background frame differencing technique was applied in the detection unit for segmenting the moving object. The detected result was then inputted into the tracking unit and later feedback as measured by Kalman filter. Finally, the test experiment indicated that the proposed model was able to successfully track the trajectory of moving objects in consecutiveframes under the real-world circumstance of occlusion. For the covering abstract see ITRD E140665.

  • Authors:
    • ZHAO, Y I
    • SHAO, CHUNFU
    • CHEN, XIAOMING
  • Publication Date: 2007-10

Language

  • English

Media Info

Subject/Index Terms

Filing Info

  • Accession Number: 01148909
  • Record Type: Publication
  • Source Agency: Transport Research Laboratory
  • Files: ITRD
  • Created Date: Jan 25 2010 9:01AM