This paper presents a new approach to the extraction of the contour of a moving object. The method is based on the fusion of a motion segmentation technique using image subtraction and a color segmentation technique based on the split-and-merge paradigm and edge information obtained from using the Canny edge detector. The advantages of this method are the following: it can detect large moving objects, the background can be arbitrarily complicated and contain many nonmoving objects, and it requires only three image frames that need not be consecutive provided that the moving object is entirely contained in the three frames. It is assumed that there is only one moving object in the image and the objects are not blurred by their motion so that the edges in the image are sharp. The method was applied to road images containing a moving vehicle, and the results show that the contour was correctly extracted in 18 of the 20 cases. We show that this contour extraction method gives good results for other types of moving objects as well. We also describe how the extracted contour can be used to classify a given vehicle into five generic categories. In this study, 19 out of the 20 vehicles were correctly classified. These results demonstrate that integration of multiple cues obtained from relatively simple image analysis techniques leads to a robust extraction of the object of interest in complex outdoor scenes.

Contour extraction of moving objects in complex outdoor scenes

This report describes a study which assessed advanced imaging technologies for potential application to roadway surveillance and detection. Technologies that were considered included ten types of infrared video cameras and one millimeter-wave still-frame imaging system. The different types of imaging technologies are described and included is a discussion of their merits and limitations. The overall conclusion reached by the researchers was that advanced imaging methods are justified only for selected roadway situations such as recurrent fog in combination with hazardous traffic patterns.

Advanced image sensing methods for traffic surveillance and detection

Real-time tracking systems often make use of a reference image and, where processing power is limited, an 'area of interest'. To obtain such information commonly requires user interaction. Typically, a reference image is obtained by the user capturing an image when the objects being tracked are not present in the scene and an 'area of interest' may also be defined by the user; both being specific to a particular scene. An alternative approach that would automatically generate and maintainn an up-to-date reference image is investigated in this paper. This consists of, essentially, 'cutting and pasting' areas of image from a sequence of frames to obtain an image containing no moving objects. Furthermore, a method for automatically generating an 'area of interest' is described. This method identifies areas of movement in a sequence of frames in order to build the 'area of interest'. These techniques have been successfully developed and proven using video sequences of more than one traffic roundabout.

/pdf/automatically-locating-an-area-of-interest-and-maintaining-a-1c2fslup77.pdf

Automatically Locating an Area of interest and Maintaining a Reference Image to Aid the Real-Time Tracking of Objects

The invention proposes a method and an apparatus for identifying through traffic, where entry to and exit from a restricted area are ascertained and the passage time is used to establish whether there has been unauthorized passage through the area.

Method and apparatus for identifying through traffic

This article described research into a method of calculating a wide range of traffic parameters using data obtained by aerial video recordings. A small (stol-type) aircraft was used to collect the video recordings. The aircraft flew up and down the section of road being monitored, at 500-2000 feet above the ground. Pictures were taken using a ccd (charge-coupled device) camera and recording the results on video tape. Each image covered 50-100 metres of ground. Flight speeds were chosen within the following limits: (a) on flights in the direction of the traffic, the speed was higher than that of the fastest vehicles; and (b) on flights against the traffic speed was as low as possible. All video recordings were analysed automatically by computer to obtain the following data for each vehicle observed: (a) time of observation; (b) actual position on road; (c) lane occupied; (d) vehicle speed; (e) dimensions (length and/or width); and (f) colour. The method for calculating traffic parameters from these data was also described, including equations to correct for the bias of moving observations, and the method to match cars observed more than once. Once the vehicle matching has been accomplished, the vehicle time-space-trajectory was calculated. This method was tested on the Brenner autobahn in Austria. Traffic speed and density profiles were obtained and found to be useful. Further tests were required to optimize and calibrate the algorithms in certain situations and to transfer selected images from video tape to computer automatically. (TRRL)

Automatic traffic data collection using aerial video images. second international conference on road traffic monitoring

In this paper an approach to traffic analysis is presented which employs image processing techniques to detect non-conforming behavior of vehicles on roadways in addition to providing the normal traffic statistics required for traffic monitoring. Traffic scenes recorded on video tape were used in the laboratory to test the approach. Binary images were obtained by subtracting each incoming sampled frame from a reference frame and thresholding the result. Centroids were calculated for each of the objects found in the binary images and were used to track the path of each vehicle on successive frames for its duration along the roadway in the region of interest. To minimize the computation time required to match objects from a given frame to the corresponding objects in the next frame, the velocity of each vehicle and inertial constraints on speed and angular deviation were used to predict each object's location in the next frame. The object nearest the target value was chosen as the appropriate match. The trajectory of each vehicle was checked for conformity. The implementation of the approach has been able to identify non-conforming vehicle behavior and issue a message on the monitor describing the detected behavior.

W.L. Brillhart

Papers

Traffic analysis including non-conforming behaviour via image processing