Abstract

In this paper we discuss the issues that need to be resolved before fully automated outdoor surveillance systems can be developed, and present solutions to some of these problems. Any outdoor surveil- lance system must be able to track ob jects moving in its field of view, classify these ob jects and detect some of their activities. We have de- veloped a method to track and classify these ob jects in realistic scenar- ios. Ob ject tracking in a single camera is performed using background subtraction, followed by region correspondence. This takes into account multiple cues including velocities, sizes and distances of bounding boxes. Ob jects can be classified based on the type of their motion. This prop- erty may be used to label ob jects as a single person, vehicle or group of persons. Our proposed method to classify ob jects is based upon de- tecting recurrent motion for each tracked ob ject. We develop a specific feature vector called a ‘Recurrent Motion Image’ (RMI) to calculate re- peated motion of ob jects. Different types of ob jects yield very difierent RMI’s and therefore can easily be classified into difierent categories on the basis of their RMI. The proposed approach is very eficient both in terms of computational and space criteria. RMI’s are further used to de- tect carried ob jects. We present results on a large number of real world sequences including the PETS 2001 sequences. Our surveillance system works in real time at approximately 15Hz for 320x240 resolution color images on a 1.7 GHz pentium-4 PC.