carpooling logo par Cyclope.ai

Video-based traffic
analysis solution

Cyclope.ai has developed a traffic characterization solution based on video analysis which allows, via roadside cameras, to detect all vehicles on the road and to classify them according to predefined classes.
This solution has been designed to meet the growing needs in terms of rich and precise road data of traffic managers and enginnering firms. The aim is to assist them in the implementation of new and cleaner Mobility policies (dedicated lanes, Low Emission Zones , suppport of Soft Mobilities, etc.).

Contact us

Main features

Cyclope.ai Comptage des véhicules
Vehicles counting

Our counting algorithm detects all vehicles on the road, with a high reliability.
As a replacement for counting loops, intrusive and difficult to maintain.

cyclope.ai Classification paramétrable
Customizable classification

Our solution automatically classifies the traffic flow according to the main typologies: light vehicles, trucks, motorcycles, buses, etc.
The classification can be defined according to specific customer needs.

Génération auto. de masques par cyclope.ai
Mask generation

The solution enables the establishement of masking zones for each camera, providing a specific count per zone.
This functionality is particularly adapted to the processing of dedicated lanes.

Ask for a demo

Key values proposition

Non intrusive

Data collection only by camera sensors on roadside. Therefore, no intervention on the roadway is required, neither for installation nor for maintenance.

Open system

Compatible with most cameras - few specific constraints on the equipment used, facilitating system maintainability and scalability.

Multifunction sensor

Possibility to progressively integrate additional video detection functionalities, for example: detection of certain types of incidents like congestion, stopped vehicles, etc.

GPDR compliant

Data confidentiality and respect of road users privacy are at the heart of all our developments. Our Counting solution is therefore, as all Cyclope.ai products, fully compliant with the GPDR.

Application fields

Highways and expressways

For analysing traffic travelling on major road axis, with vehicle speeds between 90 km/h and 130 km/h.

Major urban roads

To understand the uses on axes with a dense and varied traffic, in particular for metropolitan ring-roads or beltways.

Crossroads and intersections

To prioritize soft mobilities and fluidify congestion zones, through better management of the traffic light cycle.

This solution is based on advanced proprietary technologies in Deep Learning and Computer Vision, developed by our Data Science expert team.

Our references

Lima Expresa DIR Méditerrannée DATACITY

Deployed for LIMA EXPRESA
to analyze the traffic on the Lima ring-road

LIMA EXPRESA is in charge of operating the Linéa Amarilla, a network of 25km expressways around the city of Lima, capital of Peru.
For maintenance reasons (planning and prioritization) and a better understanding of uses which will then enable them to re-adapt their offers, LIMA EXPRESA called on Cyclope.ai to deploy a video-based only traffic analysis solution.

1260h

of processed videos per week

4

Classes to identify:
light vehicle, truck, motocycle, bus

90%

precision rate by class

15

camera streams treated

lima-expresa-analyse

Impact study of public transport restricted lanes

The Aix-Marseille-Provence metropolitan area and its partners have drawn up a mobility plan, which one of the major measure is the implementation of several public transport dedicated lanes on expressways. By 2023, it is already planned to set up 100 kilometers of reserved lanes between Aix-en-Provence and Marseille. At the end of a public Call for Tenders procedure, Cyclope.ai was selected by the DIR Méditerranée in July 2019, to participate in outsourcing to Ingerop in the evaluation of the impact of these reserved lanes.

2 months

Of traffic analysis
for each assessment mission

9

Sites on
expressways

3 à 5

Cameras
per site

lima-expresa-analyse

Green mobilities detection,
through the DATACITY program participation

Since 2014, the PCE Lutèce (traffic monitoring center of Paris City) has launched several experiments that promoted the “active” and clean modes of travel - pedestrians, bikes, scooters - used more and more by Parisians. Indeed, in the actual traffic lights management rules, sensors are mainly taking into account the motorized vehicles flow. This very often leads to dangerous situations for the new road users, particularly vulnerable. Thus, within the framework of Datacity (a data innovation program dedicated to Smart City), the Paris City Hall launched a challenge on "How to integrate soft mobilities into traffic light intersections management ?", in which Cyclope.ai participated as the main technology supplier.

1 crossroad

equipped with a
real-time detection
system in the 13th arrondissement

9 camera feeds

treated continuously
for more than 6 months

250 ms

of treatment time

lima-expresa-analyse

Technical specifications

Fixed cameras

context or thermal imaging cameras

Daytime treatment

with possibility of night treatment on thermal camera

Adaptable infrastructure

on-premise or cloud-based intelligence, according to the use case

No image storage

to optimize the associated technical infrastructure

Complementary
modules

Carpooling detection

Detection of the number of occupants
into the vehicle, for pedagogical
applications as for sanction control

Congestion detection

Possibility to configure an alarm
based on a preset threshold number of vehicles detected in a specific zone

DAI plein air

Détecter automatique de certains
incidents sur la voie : présence
piétons, véhicules arrêtés, etc. -
en aide au gestionnaire de trafic.