Social Distance Predictor Using Deep Learning

Social Distance Predictor Using Deep Learning

Author Name : D. Vigneshwar Rao, K. Ankith Sai Reddy, V. Prashanth, Ch. Dilip Goud

ABSTRACT  

The objective of the model is to make the detection of social distance between people who are roaming on the roads and the footpaths or anywhere outdoors or in the society easy and more reliable. We know that when a naked human eye monitors the, accuracy is considerably low and also a human eye cannot monitor more than one group at a time. Therefore this system is introduced to make the process easy, accurate and consistent in monitoring people at all times by using 24/7 surveillance cameras. This system is using the YoloV3 (You Only Look Once) which is the fastest object detection algorithm and contains several classes of objects for object detection. This algorithm is not only used for object detection but also used in finding the position of the object and tracking the movement of the object from one position to another. This feature will be used in finding the distance of the objects whose coordinates will further help in finding out the distance between each object in real time and determine whether they are following social distancing or not.

Keywords: Social Distance, Object Detection Algorithm, YOLO.