Vision-Based Vehicle Surveillance and Parking Management Using Multiple Cameras

Abstract

This paper proposes a vision-based vehicle surveillance system for parking management in outdoor environments. Due to the limited field of view of camera, this system uses multiple cameras for monitoring a wide-area parking lot. Then, an affine transformation is used for linking the relations between cameras. Two major components are included in this system, i.e., vehicle counting and parking place management. For the first one, this system integrates three features, i.e., color, position, and motion together for well tracking vehicles across different cameras. Thus, even though two vehicles are occlude together, they still can be well identified and tracked under different lighting changes. For the second one, we propose a grid-based approach to model the color change of floor for determining whether a parking space is vacant. Due to the perspective effects, the visibility of a parking space is often affected by the vehicle parking on its neighbor space. To tackle this problem, we divide a parking space to different grids with different weights. Then, according to the orientation of parking space. an automatic weighting scheme is proposed for measuring the visibility of each grid. By judging the quantity of edge, we can determine whether the parking spaces are vacant. The experimental results reveal that our system works well and accurately in several different conditions.