A multi-classifier image based vacant parking detection system

Abstract

With the rapid expansion of major cities, the tremendous increase in the number of vehicles, and the construction of huge buildings and parking areas, there is a need to develop smart parking systems to assist drivers finding the nearest available parking spots. Such systems have witnessed substantial research efforts in developed countries. There are mainly four categories of car parking management systems: counter-based, wired-sensor-based, wireless-sensor-based, and image-based. In this paper, we develop, implement, and test an image-based system for the detection of vacant spaces in a parking area. Following an initial edge detection stage, we combine edge density, closed contour density, and foreground/background pixel ratio, at every car parking spot, to identify whether a car is present or not. Combining the features above results in a robust vacant space detection system at low computational cost.