ISSN 2348-1196 (print) International Journal of Computer Science and Information Technology Research ISSN 2348-120X (online) Vol. 9, Issue 3, pp: (39-47), Month: July - September 2021, Available at: www.researchpublish.com
Home Automation using Machine Learning for Television Balu Anush A[1], Archana S Ajith[2], Buthalapalli Lohith Reddy[3], B Sai Rahul Reddy[4], Zafar Ali Khan[5] Presidency University, Bengaluru, Karnataka, India-560064
Abstract: Home automation is building a smart home which will monitor different attributes like light, temperature, appliances, entertainment unit and many more. This paper describes about home automation specialized in the field of the entertainment unit that is television. Our primary purpose here is that when a person comes in front of the television, the television should be switched on and the desired channel which most probably the person would watch at that time will be displayed. So, when a person comes in front of the television at a set distance the Passive Infrared (PIR) sensor which is used to detect motion detects the motion of the person and Pi camera recognizes the person’s face using computer vision, which is integrated with raspberry pi, and according to the algorithm used, the television is switched on and the preferred channel will be displayed on the television screen. The proposed system uses only PIR sensor, Pi camera and raspberry pi hence it is low cost. The proposed system mostly comes in handy to old and disabled people as they can just come in front of the television to watch their favourite show instead of doing everything manually. Keywords: Home automation, Machine Learning, Computer Vision , Television, PIR sensor, Pi Camera, Raspberry Pi.
I. INTRODUCTION This paper aims at minimising the relative effort of an individual in suggesting relevant television channels through the optimal use of Internet of Things and ML techniques. The suggestion employs the use of PIR sensor which is used to detect the presence of a human being, PI camera which is a basic camera module attached to Raspberry PI which is used to identify the person and his/her motion and Raspberry PI which is used to control the state of television to accomplish task. The proposed solution could find its valuable application in a number of use cases, the most prominent of which would be its use to aid especially abled individuals in helping them view/have access to television at limited physical hassle.
II. RELATED WORK [1] John Jaihar, Neehal Lingayat, Patel Sapan, Vijaybhai, Gautam Venkatesh, K. P. Upla use CNN, Fischer Face Classifier, SVM give an insight about how to go about bringing in ML algorithms with IoT. [2] Vishwajeet Hari Bhide, Dr, Sanjeev Wagh use Naïve Bayes and points out the perspective of disabled/handicap people. Use of household data and activity to incorporate the algorithm. [3] L. Mary Gladence,V. Maria Anu, R. Rathna, E. Brumancia use NLP, Logistic Regression, Bayes Classifier, API Protocols Uses NLP. They speak about the connection & need for automation system. Preference of cyber physical system, advancements and different technologies used. [4] Harsh Kumar Singh, Saurabh Verma, Shashank Pal, Kavita Pandey gives information about the end devices used and their technology. [5] Vishwakarma, S. K., Upadhyaya, P., Kumari, B., & Mishra, A. K talk about IoT applications, the system design. provides workflow chart, provides a way to use low power consuming devices.
Page | 39 Research Publish Journals