Machine Vision for SmartCities

Our Computer Vison solutions are used by different cities and self-government units (JST). Our solutions address specific issues which include the following:

Enforcing restrictions on Long Tracks
A mid-sized city wants to get rid of disruptive traffic of long truck vehicles. Many drivers of Long Tracks are not adapting to road signs which impose restrictions on entry of heavy vehicles to the city. Heavy traffic increases the risk of serious car accidents. It also puts a strain on local infrastructure. City law enforcement unit wants to investigate each case of traffic violation and impose penalty fees on violators.
Information about parking space
One of large Polish cities wants to set up an on-line system providing information on the number of available parking places. Existing solutions require infrastructure, significant investments and yet not always provide reliable information. The city management is interested to use Machine Vision to address this challenge.
Monitoring access to restricted areas
A historical Polish city wants to set up a system of monitoring access to restricted areas. They have a list of registration numbers of cars which are entitled to access restricted area and want to get rid of all cars which are violating restrictions.
Monitoring streetlights
A city wants to monitor the operation of streetlights. They want to have on-line reports about start / end-time and reports about lamp failures.
Video-identification of system users

Our Computer Vision systems relay on proven neural network algorithms for object detection and classification. The system comes with self-learning module which is easy to operate by users. By using this module users can adjust algorithms to their individual needs and requirements.
Computer Vision projects delivered by BD Analytics have different features regarding: algorithms used, source of image data, point of processing or power supply. They include the following:

Our systems are using state-of-the-art object detection and classification algorithms. We also provide self-learning modules which can be applied by users to improve accuracy.

Computer Vision solutions require image data from a given source. This source can be an existing CCTV infrastructure or it can be an IP Camera with build-in Computer).

Image data can either be processed on the Client’s server or in a remote location on the computer integrated with IP Camera.

When image data is processed on the Client’s server the Client has to provide a cable connection between the camera and server. The server, depending on technical specification, can process signal from several cameras.

When image is processed in a remote location on the computer integrated with IP Camera then there is not a need to provide a cable connection between the camera and server. The IP Camera sends a narrow stream of text or image data by means of LTE.

The majority of installations have access to constant power supply. They involve high resolution IP cameras with night vision systems and built-in computers.

There are installations which are equipped with battery which provides power supply for a camera during the day. At night the camera takes power from streetlight systems. This is also the time when battery is being charged for operation during the day.