A measurement method of pedestrian traffic flows by use of image processing and its application to a pedestrian traffic signal control

We propose a method to estimate macroscopically the situations of pedestrian traffic flows on a crosswalk. This method evaluates the spatial distribution patterns of pedestrians obtained with image processing by using entropy. By use of entropy, a pedestrian traffic demand can be estimated from a pedestrian traffic image. We also propose a pedestrian traffic signal control method which adjusts a pedestrian green time using entropy every signal cycle.