Non-intrusive traffic counting systems are not embedded in the roadway but installed off-road, either at one side or overhead above the roadway or lane. Their service life does not depend on the road surface, but they tend to provide lower-quality data due to certain environmental conditions and actual installation conditions.
LASER TRAFFIC COUNTING
Laser-based traffic counting is the best non-intrusive traffic counting system and, under certain heavy traffic conditions with stop-and-go situations, the best traffic counting system in general. KINEO works with equipment made by SICK, a European manufacturer with whom we have practical experience in Spain and Peru.
We provide multi-lane solutions (2-3 depending on the distance from the sensor support to the road) or single-lane solutions (depending on the quantity and quality of the data required). Installation can be done on a side column, on a bracket that brings the sensor closer to the roadway, or on an existing gantry or structure in the case of solutions with a sensor per lane to be installed overhead and centred on each lane.
We can work with a laser sensor or a combination of two sensors depending on the type of data, the quality required or even the environmental conditions, which can influence the minimum architecture required. In either case, we generate a database with speed, class and dimension data. Accuracy depends on the number of sensors, their location in relation to the vehicles and the flow of vehicle traffic.
RADAR, INFRARED AND ULTRASOUND TRAFFIC COUNTING (DUAL AND TRIPLE TECHNOLOGY)
There are multiple traffic counting options available on the market with doppler, infrared (IR) or ultrasound gauges, and equipment that integrate two of these, such as doppler+IR or IR+ultrasound, and, finally, equipment with triple technology (in versions for 2, 5 and 8 classifications). The 1- and 2-technology options do not cover more than 2 classifications, although some more could be obtained based on length, but in situations where the vehicle class is clearly marked based on length.
They are generally systems that need a power source because the processing of the sensor signals must be done in a processor, which, in the case of fibre optics, can be located in a control centre; however, if there is no Ethernet communication, the process must be done locally and this requires a VAC power supply or a duly sized solar system.
The combination of sensors (dual and triple technology) measures the speed of the vehicle by means of the reflected microwave frequency shift detected by the Doppler Radar, the ultrasound section scans the height profile of the vehicle passing under the detector, while the multichannel curtain of the infrared detector provides information on the position of the vehicle within the lane.
The combination of sensors is mounted on side columns, brackets, gantries, bridges or other raised structures that pass over the lanes to be monitored.
Radar, infrared and ultrasound traffic counting applications range from vehicle classification, counting of all vehicle types, individual vehicle speed, presence and queue detection, occupancy, distance and time between vehicles, wrong direction, etc.
TRAFFIC COUNTING BY VIDEO ANALYSIS
Traffic counting and classification by video analysis is an alternative for permanent and temporary traffic counting in situations where the number of vehicle types is light (2-3 classes) and maximum accuracy in counting and classification is not required.
The software analyses the video recorded preferably by cameras specifically set up for the purpose, because the quality of the result is totally conditioned by the quality of the images recorded and the position of the cameras relative to traffic direction.
Excessive angles generate perspective distortions that cause the automatic analysis to interpret the image of two short vehicles as a long vehicle, so only 50% is counted and, therefore, the vehicle is misclassified. Therefore, the best possible position for traffic counting by video analysis is one that takes images from overhead on the road surface; in other alternatives, the greater the angle, the worse the result.
These systems are influenced by lighting and weather conditions, since good detection is not possible without clear images. Only thermal alternatives would be feasible in such cases.
We have analysis solutions for existing camera networks, simply installing software in the control centre, embedded solutions in each camera, portable video recording systems with up to 72 hours of autonomy with an expansion module for another 96 hours, etc.