Detailed description of the service 

Researchers and local authorities alike need traffic data to understand, plan and manage future traffic. In road traffic, this requires, among other things, traffic counting at intersections and cross-sections of roads. In Measure G-4, the OTGroundTruther was developed, a tool that can be used to carry out manual traffic counting at intersections or cross-sections based on video material. It is possible to capture traffic flows in detail and to distinguish between classes of different road users. The OTGroundTruther enables the rapid counting of traffic flows if automated recording is not possible with sufficient certainty and without costs for the software. This tool thus provides an alternative to costly counting personnel on site or automated counting methods that are not accurate in every scenario. In addition, this tool can be used to generate ground truth data for benchmarking the counting accuracy of automated systems such as OpenTrafficCam. Since it is licensed under the GPL-3.0 license, it is open source and can be used and changed by everybody in any way. This makes the tool very transparent and open for changes for a use on slightly different tasks.

Classes that are provided by the OTGroundTruther are pedestrian, bicyclist, bicyclist with trailer, cargo bike driver, scooter driver, motorcyclist, car, car with trailer, private van, private van with trailer, delivery van, delivery van with trailer, bus, truck, truck with trailer, truck with semitrailer, train and other. The tool is designed to be as easy to use as possible and instructions for how to use it can be displayed by clicking on the help button. You can jump back and forth between different steps in the video. Traffic objects that have already been counted are displayed directly in the image so that you can check them without further ado and to ensure that they are not counted again. In the right-hand area of the tool, there is an overview of all counted traffic objects, which can also be used to jump directly to the traffic object and to delete it. It should be noted that the gates required for the count are loaded in the form of an otflow file or otevents file and these can be created by the OTAnalytics tool.

With the GroundTruther, this data can be created using video material.

The concept of the digital twin, including quality assurance and evaluation (‘Quality Check’ & ‘Evaluation’), has been implemented in the OpenTrafficCam software and can be used directly by stakeholders (traffic researchers).

Traffic data plays a crucial role for both researchers and local authorities when it comes to understanding, planning, and managing current and future traffic patterns. This information is essential for a variety of purposes, including traffic regulation, urban planning, and improving road safety. One of the primary ways to collect traffic data is through traffic counting, which is often done at key locations such as intersections and cross-sections of roads. Accurate traffic counting allows experts to evaluate traffic flow, determine the capacity of roads, and monitor the behavior of different types of road users. Traditionally, traffic counting has been done either manually at the location or through automated systems, but each method comes with its challenges, including high costs or lack of accuracy in some cases.

In response to these challenges, Measure G-4 saw the development of the OTGroundTruther, a manual traffic counting tool designed to simplify and improve the process of gathering traffic data. The OTGroundTruther can be used to perform manual traffic counts by analyzing video footage of intersections or cross-sections of roads. The tool allows users to track traffic flows in great detail and distinguish between various types or classes of road users, making it highly versatile. This capability to differentiate between different vehicle types and road users is particularly useful in complex traffic environments where automated systems might struggle to provide accurate counts. The OTGroundTruther stands out as a valuable solution when automated systems fail to record data with adequate certainty or precision, offering an efficient and cost-effective alternative.

One of the key advantages of the OTGroundTruther is that it enables fast and accurate counting without the need for costly on-site personnel or expensive automated systems. This makes it an appealing tool for cities or organizations with limited budgets or for scenarios where automated systems are prone to errors. Additionally, the OTGroundTruther can serve as a means to generate “ground truth” data. Ground truth data is essential for validating and benchmarking the accuracy of automated traffic counting systems, such as OpenTrafficCam. In other words, by using the OTGroundTruther, traffic researchers can establish a reliable dataset against which they can measure the performance of automated systems, ensuring their accuracy and reliability in real-world applications.

Another important feature of the OTGroundTruther is its open-source nature. The tool is licensed under the GPL-3.0 license, which means that it is freely available to the public and can be modified and adapted to fit specific needs. This level of openness and flexibility makes the OTGroundTruther highly customizable, allowing users to adjust it for slightly different tasks or scenarios. For example, traffic researchers who need to focus on a specific type of vehicle or road user can easily modify the tool to meet their needs. This transparency encourages collaboration and continuous improvement, as users can share their modifications or enhancements with the broader community.

The OTGroundTruther provides a variety of classifications for different types of road users. These classes include pedestrians, cyclists, cyclists with trailers, cargo bike riders, scooter drivers, motorcyclists, cars, cars with trailers, private vans, private vans with trailers, delivery vans, delivery vans with trailers, buses, trucks, trucks with trailers, trucks with semitrailers and trains. By offering such a wide range of classifications, the tool allows users to gather detailed traffic data that can account for the various types of vehicles and road users that make up the traffic flow. This level of detail can be invaluable for traffic analysis and urban planning, especially in areas with diverse traffic compositions.

The OTGroundTruther has been designed with user-friendliness in mind, ensuring that it is easy to navigate even for users who may not have extensive technical expertise. The tool includes a help button that provides instructions on how to use it, making it accessible to a broad audience. Users can jump back and forth between different points in the video to review or adjust their counts. The counted traffic objects are displayed directly in the video image, allowing users to quickly verify that an object has been counted and avoid double-counting. On the right side of the interface, there is an overview of all the counted traffic objects. This overview allows users to jump directly to a specific traffic object or delete an entry if necessary. This feature makes it easier to manage the counting process and ensure accuracy.

The OTGroundTruther requires gates to be loaded in the form of either an otflow file or an otevents file, which are created by the OTAnalytics tool. These gates are necessary for the counting process and help define the points in the video where traffic objects are counted as they pass through. By using video material and digital tools like OTGroundTruther and OTAnalytics, traffic data can be collected and analyzed more efficiently than through traditional methods.

The concept of the digital twin has also been incorporated into the OTGroundTruther, ensuring that quality assurance and evaluation processes, such as Quality Check and Evaluation, can be applied seamlessly. This concept has been fully implemented in the OpenTrafficCam software, which can be used directly by traffic researchers and other stakeholders. The digital twin approach allows for continuous monitoring and assessment, helping researchers ensure that the traffic data they collect is accurate and reliable.