The overall objective of 3D-INSIDE is to develop novel techniques to get up to date information about the status of pedestrian behaviour at buildings of large public transport infrastructures. This information can be used to identify bottlenecks where dangerous situations for pedestrians can happen or areas where advertisements should be placed so that a lot of people will come across them. In 3D-Inside techniques will be developed for automatically and rapidly obtaining 3D models of the indoor environment of an infrastructure based on digital images. In a next step motion data of pedestrians is analysed by using video surveillance data. Based on this analysis and the 3D model a simulation of the pedestrian walking behaviour will be performed. The main task of JOANNEUM RESEARCH in this project is the development of vision based people detection and tracking algorithms.
Sponsor: FFG under the BMVIT, FIT-IT Visual Computing program