Information about pedestrian and public transport users activity, including volumes, walking speed, and trajectories, are used by transportation agencies and researchers for planning, design and analysis purposes. There exist a number of technologies for automatic pedestrian data collection; however all have inherent limitations either in functionality or in monetary cost. Also, most technologies only provide counts. The idea we wnat to proposes is the use of an inexpensive motion sensing device: the Microsoft Kinect, which is able to track multiple people in low-light conditions and could be combined with existing video based daytime tracking. The tracking software and speed estimation methodologies are described, and indoor and outdoor studies show the system’s effectiveness at determining pedestrian volumes and walking speeds. The accuracy of speed data is very satisfactory, with correlation of 98% or more with respect to video data validation speeds. The accuracy of pedestrian volume data varies with traffic conditions, however in low to moderate traffic conditions its performance is accept able with an under counting error of about 8%. The different applications of the sensor and its complementarity with other sensors is discussed, this being the first step towards a multi-sensor system.
In this Pdf. is possible to see a very interesting example of the Kinect technology for the control and the analysis of the customers of one or more shops. The same idea expressed in this document is what we are thinking to applicate in the the bus and railway station of the city
Link of 3D descriptor for people re-identification: http://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=6460401