Logo of the Physikalisch-Technische Bundesanstalt

Measuring the speed and the distance of vehicles on challenging road sections

21.09.2012

So far, video-based measurement of the speed and distance of vehicles has only been possible on practically straight and flat road sections. However, accidents more frequently occur on hilly motorway sections with tight bends and cross slopes. Since the type approval for the VIDIT traffic-monitoring system (VKS) has been extended, the speed and distance can now be measured in exactly such accident-prone sections.

Figure 1 shows such a measuring point which is located at the motorway interchange "Köln-Ost" where the traffic volume reaches 240,000 vehicles per day. The VIDIT computer-aided traffic-monitoring system "VKS 3.0" is based on the single-image analysis of a video recording of the traffic situation. To determine the speed, the distance covered by a vehicle between two imaged traffic situations is measured. This distance measurement is based on the acquisition of the two vehicle positions recorded in the video image by manual focusing with a measuring line. The appurtenant time interval between the two traffic situations results from the refresh rate of the camera (which was checked at verification) and the number of full video images determined automatically by the traffic-monitoring system.

Figure 1: Main photo: 3D measurement point along one of Germany's busiest road sections with 240,000 vehicles daily; insert: view into a measuring bus equipped with a VKS traffic-monitoring system.

The designs of the VKS 3.0 traffic-monitoring system approved to date can be used only at measurement points whose surface, as an ideal plane, is sufficiently describable (2D measurement points). Hereby, the ideal plane is defined by four markings which are located on the road and are clearly visible in the video image (so-called "ground control points"). The newly approved design also allows measurements at 3D measurement points whose road surface configuration considerably deviates from an ideal plane. Hence, it is possible to carry out speed and distance measurements at measurement points with:

  • crest curves and/or sag curves,
  • bends with radii smaller than 500 m,
  • or bends with an alternating cross slope (left-right slope).

Hereby, the contour data of the road surface configuration (surface pattern) are acquired in a special measurement point file and saved to the analysis PC.

When analyzing the measurements performed at 3D measuring points, first the position of the measuring line in the video image is, similar to the procedure used so far, converted to the position on the ideal ground control point plane. In addition to this mathematical procedure called "perspective transform", a so-called "3D transform" is performed with the aid of the individual measuring point file.

In order to check whether the new traffic-monitoring system measures distances correctly, a reference section consisting of accurately aligned traffic cones was set up at a 3D measurement point and analyzed with the system under test. In addition, comparison measurements were carried out with a GPS/INS device (on board a test vehicle) as a reference. For these measurements, the two systems had to be temporally synchronized. For this purpose, an obstacle was positioned on the road surface at the measurement point (see Fig. 2). First, the time of the traffic-monitoring system when the vehicle passed the obstacle was determined by video recording. When passing the obstacle, the acceleration sensors of the GPS/INS device provided a characteristic signal for vertical acceleration; this allowed the system's time of the GPS/INS device to be determined (see Fig. 2). The two systems' times could then be synchronized with each other. Based on selected time differences, the distance data determined by the traffic-monitoring system were compared with the indications provided by the GPS/INS device. The comparison of the distances measured yielded very good agreement. These comparison measurements clearly confirmed that the new design of the traffic-monitoring system measures distances and speeds correctly at 3D measurement points.

Figure 2: Large photo: Vehicle passing an obstacle, 1) marking of the ground control point, 2) obstacle; diagram: vertical acceleration curve; the red arrow indicates the signal level generated by the rear wheels of the vehicle when passing the obstacle.

Contact person:

Johannes Kupper, Dept 1.3, WG 1.32, e-mail: johannes.kupper@ptb.de