In this work, we tried to provide, in the context of crowd monitoring in subway (underground railway) station corridors, a motion estimation method and a motionless detection method which both take into account three difficulties: real time constraint, deformable objects and occultation. The first part of our study concerns motion estimation. Thereby, we chose three motion estimation techniques, each of them belonging to one of the three main classes of motion estimation methods. The chosen techniques are: block matching (primitive matching technique), optical flow (differential technique) and Gabor filter (frequential technique). We then attempted to improve both the speed and the performances of each algorithm. As a result, we got a very efficient optical flow technique suitable for various applications in the field of crowd monitoring and traffic management. The second part of our study concerns the motionless detection. Thus, we developed a method based on motion detection and background image construction.
Crowd motion estimation and motionless detection in subway corridors by image processing
01.01.1997
712816 byte
Aufsatz (Konferenz)
Elektronische Ressource
Englisch
IMAGE PROCESSING - Detection of abnormal stationarities in subway corridors
Online Contents | 1999
|ARTIFICIAL VISION - Detection of abnormal stationarities in subway corridors
Online Contents | 1998
|A Motionless Laminography System
NTRS | 2006
|Intelligent subway carriage with crowd density detection function
Europäisches Patentamt | 2020
|