New research on People Counting and Human Detection

Abstract

People counting methods and human detection methods can be considered in determining the number of people in a room. Lately, computer vision in artificial intelligence (AI) is gaining an increasingly important place by interpreting infrared images according to certain characteristics, namely groups of connected pixels in an image that share some common property. This paper starts from three existing systems, two counting the number of people using two PIR sensors, and two ultrasonic sensors and a human detection method, using the Grid-EYE camera. As the Grid-EYE camera resolution is only 8 x 8 pixels, it is not accurate enough and a better solution is being sought. The Lepton infrared camera is chosen, with a resolution of 80 x 60 pixels. Measurements are made in parallel with the two cameras and the results are compared and analyzed. Finally, conclusions are drawn, and applications are exemplified in which it is useful to determine the number of people in a room.