Pedestrian localization and tracking system with Kalman filtering

This work presents an implementation of a vision-based system for recognizing pedestrians in different environments and precisely localizing them with the use of a Kalman filter estimator configured as a tracker. Pedestrians, in various poses and with different kinds of clothing, are first recognized by the vision subsystem through the use of algorithms based on edge density and symmetry maps. The information produced in this way is then passed on to the tracker module which reconstructs an interpretation of the pedestrians positions in the scene. An appropriately configured indoor system setup with an accurate measurement of the imposed human trajectory has been realized. This setup has permitted an accurate evaluation of the accuracy of the results, when the new auxiliary tracker is activated.