With the introduction of “smart homes” and
the integration of automation systems to control a given
space’s functioning, we see an evolving need to optimize
the integrated automation systems to meet the intelligence
and controllability level of the client. The most appealing
automation systems are the ones based on power usage
and control. Current systems available in the market for
automating electric switches are primarily based on
either infrared or proximity sensors, which are either
inefficient or expensive in more extensive system settings.
Our proposed system uses image processing to control
switches of electrical appliances in the room. Our system
first captures the images of the room and further
processes it before applying the Tiny Faces Algorithm on
it, which will determine the positions of the individuals
present in the room. Using this obtained data will
generate control signals to the raspberry pi, which will
further trigger the relay circuit, thus triggering an
automatic ‘switching ON’ of the electrical appliance
connected to this system. The result of the proposed
system is nearly 92% accuracy. This simple, efficient, and
cheaper system helps to optimize the power consumption
of the house and helps to save energy.
Keywords : Image Processing, Home Automation, Face Detection, Tiny Faces algorithm, Raspberry pi.