Near East University engineers developed software that can control the social distance and mask from camera images
Date Added: 09 June 2020, 23:18
Last Updated Date:17 November 2020, 10:35

Near East University Engineering Faculty, Computer Engineering Department and Applied Artificial Intelligence Research Center engineers analyzed the camera images and developed software to calculate the social distance between individuals and check whether they wear masks.

According to the information given by the Press and Public Relations Directorate of Near East University, after the measures taken against the coronavirus, it will be possible to check with the software developed whether or not people are wearing masks and obeying requirements of social distance in places such as sports halls, nurseries, libraries opened during the normalization process.

Thanks to this system developed, the images taken with the camera placed at the entrance of the places are analyzed and the person's body temperature checked and if they wear a mask or not monitored. The images taken from the cameras placed inside the places are analyzed and the distance between the individuals is measured. The system gives a warning if it is detected by the system that the required compliance is not obeyed. Using computer image processing algorithms and deep learning techniques, engineers detect the face of the person before the images taken from the camera and then check for a mask on the face. In addition, the body temperature of the user is measured with the help of a thermal camera. In order to detect the masked and unmasked face of the person, approximately 2000 masked and unmasked face photographs were uploaded to the system with 99% accuracy with deep learning techniques.


Gives social distance warning...
In the system that finds people from in the image using deep learning algorithms and trained object detection techniques, the image is turned into a bird's eye by applying “perspective transform” to this image. Then, the distance between the people is calculated by marking the points where people are located and using the "Euclidean Distance" method. Therefore, those who do not comply with the social distance rule are detected and necessary warnings are given by the system.

From the virtual world to real-life...
Murat Arslan, one of the engineers of the "Applied Artificial Intelligence Research Center", who stated that the tests of the software in the computer environment have ended and that studies are continuing to adapt it to real-life, said, "It was essential to take such measures in the normalization process after the measures taken against the coronavirus. It is very important for every individual in the community to follow these measures for public health. Such auxiliary systems provide more accurate and unbiased analysis without physically exhausting us.”

Studies to serve humanity are being carried out…
Arslan added that they continue to work on many other systems such as autonomous car simulation, head movements for disabled people, computer control, the system that instantly translates the sign language used by the hearing-impaired, and the driver sleep detection system. The application video, which provides social distance and mask control system, can be accessed through https://youtu.be/dzcRvJyjcvk and https://youtu.be/G6GINgBxxFg links.