Application of HOG Algorithm for Automated Room Control System

Abstract

This study aims to develop a prototype of automated room control systems. The histograms of oriented gradients (HOG) algorithm is used to detect and count the people entering or coming out the room so that the use of electricity is more appropriate. The raspberry Pi microprocessor is used as a controller since it is efficient and can be expanded for monitoring over the web. The system testings show that the prototype can detect and count the number of incoming and outgoing people enough accurately with error rate of about 1.7%, while people entering or coming out the room not simultaneously, but can not detect and count accurately with error rate of about 43.3%, while two person walk close together through the door. Using the automated room control, the electronic equipments will be active when there are people in the room and the  temperature below 20⁰C.