Real Time Human Detection and Counting

Abstract

Detecting humans in images and videos is a challenging problem, the camera and the background and to variations in pose, appearance, clothing, illumination and background clutter. We have developed a detector for standing and moving people in videos, testing several different motions coding schemes and showing the best overall performance. Use of human modelling to recognize and monitor human activity in the scene such as human walking, running etc is tracked. In addition to videos, detection from a static image can also be carried out by providing image as an input instead of a live feed from a CCTV footage respectively. Human detection in videos (i.e., series of images) plays an important role in various real-life applications (e.g., visual surveillance and automated driver assistance). The task of human detection in a series of images is challenging due to various reasons. One of these reasons is the variation of human size in the video frame. This results from changing the altitude of the platform that the camera is attached to during the task. Accuracy and short training time are the two important factors that should be taken into consideration to get a robust human, nonhuman classification system. The current Covid-19 Pandemic has altogether increased the need of such sustainable system that is Real time human detection to avoid any mishap and limit the spread of the virus, with the help of detected persons required actions can be taken by the concerned authorities respectively.