TPU Developer Makes a Smart Parking Monitoring Service

A young researcher of Tomsk Polytechnic University has created an online service for monitoring the availability of parking spaces.

 

“It tends to take a lot of time to find a free parking space. Moreover, improperly parked cars may lead to conflicts and accidents. A smart city infrastructure is being developed but the parking problem has not yet been resolved, ” Ali Ozdiev, developer and junior researcher at the TPU International Laboratory for Non-Destructive Testing, says.

The SPOT (Smart Parking Occupation Tracking) service is uses a video camera aimed at a parking lot and a single-board microcomputer, a kind of AI sensor that analyzes the video stream. Device algorithms use machine learning methods and were developed by TPU scientists.

“The device is an original development using both ready-made free software and our original solutions.

The video stream is processed using machine learning technologies. The system already knows how to recognize cars by analyzing the image for the number of occupied and available spaces. In the future, for instance, the algorithm will be able to find improperly parked vehicles or those occupying two parking spaces.

All this happens in real time,” the developed says.

The device is ready and was tested in parking lots in Karpov street in Tomsk. In addition, a demo version of the system is integrated with the 2GIS map.

You can monitor the test parking lots on the site. Empty parking spaces are displayed with green markers and occupied ones with red.

“The web service is available for smartphones. I used a webcam in my first tests, but quite a few parking lots are equipped with video surveillance, so it is not necessary to change equipment to operate the system,” the scientist underlines.

Photo: testing parking lot

An interesting challenge for the SPOT, according to the developer, could be the improvement of parking spaces between the TPU buildings on Lenin Avenue. This is downtown, so it is difficult to find available parking there.

“Machine learning and computer vision are promising technologies, which can solve numerous issues in the modern world. Therefore, the project is open to proposals from potential partners and investors, ” Ali says.

Find more and join the project on the official website.