Mobile and tablet applications meant to detect pedestrians in traffic and fog to prevent car crashes are being developped by researchers at the Cluj Technical University, inside an European project, with first outcome due in 2015. The applications could be used by all drivers possessing a tablet or a smartphone.
”COMOSEF project has a RON 1.5 M European financing and aims at developping driving assistance applications on mobile phones or tablets and detecting pedestrians with the help of infrared images. There are two more components on the mobile devices, one capable to detect pedestrians in the traffic, mostly in the city traffic, and the other which detect fog from the images in order to inform the driver on the distance between his car and the other ones and to adjust speed to traffic conditions. All these applications are meant to prevent smash-ups”, Cluj researcher Mihai Negru told Mediafax.
According to him, the algorithm detecting fog has already been portated on tablets and is on the testing-board.
The other application detecting pedestrians with the help of infrared cameras is directly proportional to the heat emitted by their bodies. ”The camera is placed outside the car and can detect a pedestrian from a distance of 50 metres. The application is useful when it is fog outside and the headlights cannot catch the pedestrians profile , but also when it is raining or snowing. The camera can be connected to the car display and thus can detect a pedestrian whom the driver could not see otherwise and could hit him. The application is calculating the distance of the pedestrian to the car, and the system will throw out a sounding warning and place the pedestrian inside a red rectangle, if he is in danger”, other researcher Raluca Brehar explained.
The Cluj researchers team is working at other pedestrians detection application with the help of a tablet or smartphone fixed up on the windscreen of the car. ”We need to find a real-time detection solution. We succeeded in implementing a soft that runs rapidly also on tablets. We are trying to improve timing and accuracy in order to detect as many pedestrians in difficult traffic conditions as possible. There could be some cases when the pedestrians are staying behind parked cars, ready to cross the street. Another target is to identify pedestrians when the lighting conditions are not so good”, researcher Arthur Costea also said.