Latest Innovations in Electronic Tolling Systems & Vehicle Profiling

Solutions to make roadways safer and more efficient

axle counting for multi-lane traffic, Latest Innovations in Electronic Tolling Systems & Vehicle ProfilingElectronic tolling has become widely accepted as the most efficient and cost effective solution for the tolling industry. Electronic tolling systems classify vehicles using a combination of various criteria, including vehicle dimension, vehicle class, number of axles, and number of tires. Using the number of axles has spread worldwide as the basis criteria for many electronic tolling systems.

In the Americas, 63% of tolling systems are axle based. However, when installing high-precision axle counting for multi-lane traffic, many electronic tolling systems require installation within the actual road pavement, which is costly and time-consuming. SICK, Inc. has better solutions that still provide the same level of convenience and operational efficiencies, without doing damage to the actual roadways.

axle counting for multi-lane traffic, Latest Innovations in Electronic Tolling Systems & Vehicle Profiling

Profiling and Counting Vehicles Using Vision-Based Axle Counting

Axle counting for multi-lane traffic locations is now possible without any installation in the road pavement using a SICK profiling system (TIC502, TIC102, or FPS Profiling systems) and the vision-based axle counting system. With these systems, axle counting can now happen from above the lane and from the side of the road with 2D LiDAR sensors and infrared cameras.
The monochrome, near-infrared camera used on this system works well during the day or at night. It has a 1.3 megapixel resolution, with a frame rate of 60 fps (system only requires 10 fps). The housing includes an integrated illumination unit 850nm and functions in a wide-range of temperatures, from –40°F to 140°F.

An above-the-lane mounting position has many advantages, including easier services without lane closures, less contamination due to high mounting position, and less vandalism due to high mounting position. If you’re looking to monitor three lanes of traffic, it’s ideal to have three LiDAR sensors and four IR cameras for complete coverage.

axle counting

This technology is all made possible by using deep learning. SICK recorded hundreds of thousands of vehicles at several sites during different times of day and in different weather conditions. This huge amount of data is the input for a convolutional neural network (CNN) for object detection.

In vision-based axle counting applications, the CNN finds image regions that look like axles, classifies them, and counts them. Up to 15 single pictures can be stitched together to make one overview picture of a vehicle. All counted axles are marked in the picture. With this system, 95% of axles are detected at speeds up to 75MPH (for maximum performance).

axle counting for multi-lane traffic, Latest Innovations in Electronic Tolling Systems & Vehicle ProfilingInterested in learning more about our vision-based axle counting capabilities? Contact us today at info@sick.com.

Print Friendly, PDF & Email