The goal of the project was to develop and implement a parking guidance system for vehicles, aimed at precisely determining the occupancy status of parking spots, especially for trucks, based on precise detection and classification of vehicles at entry and exit, as well as a balancing module equipped with an algorithm for improving the result based on a statistical model.
The developed system has the following key features:
- data source – a measurement system based on the integration of lidar-2D scanner readings with image analysis from cameras dedicated to ITS. To achieve high (required) detection quality, it enables three-dimensional reconstruction of the vehicle silhouette (length, height). Unlike available commercial systems, the system does not interfere with the road surface, thus not causing problems during installation and operation. Systems installed in the roadway are exposed not only to strong environmental factors but also to factors related to operation (vibrations, exhaust gases, cleaning agents) and maintenance (salting) of road infrastructure, which will be considered both on the construction/mechanical and software side.
- data quality – a precise vehicle detection, identification, and classification system was developed, based on neural networks and proprietary ANPR/MMR algorithms, including a worldwide innovative classification algorithm based on video image analysis. The system works correctly even in atypical situations based on real driver behaviors (e.g., reversing or parking at the parking entrance). The quality of measurement data ensures that the system does not require frequent calibration (due to measurement errors).
- statistical correction – the system is equipped with a self-learning statistical model (parking system statistical model) based on balancing system indications and periodic calibration results, used to improve parking occupancy determination.
- compatibility – data transmission interfaces and protocols were developed in accordance with international standards TL-S2012, Datex II, allowing the system to be used worldwide.
The system was assessed for quality and functionality in accordance with the requirements specified by the Directorate of Highways of Southern Bavaria (ABDSB). Test results were obtained, among others, at the reference parking lot BAB-A9 Gelbelsee , test parking lot in Wrocław (ZABERD S.A.), and at highway service place at Witowice. Ultimately, it was shown that the functionality and quality of the NeuroPark system meet the requirements defined by ABDSB in Munich.
The project was carried out in cooperation with the Road and Bridge Research Institute in Warsaw.
