Home Security Using IOT and Machine Learning

EasyChair Preprint 3212

Abstract

Now a day’s home security has become an important issue in real world. Although there are good quality security systems available, there is scope for improvement in the area of cost effectiveness and utility features. Current advances in the field of internet–of-things (IOT) including machine learning can bring about improvements in the home security systems. The main input in this project is from infrared (IR) sensor also other indoor sensors like fire, gas, level sensors and it is fully controlled by microcontroller embedded on NodeMCU board while the output is buzzer to startle the theft and also alert the neighbours around and lastly Android application communicating through MQTT server to the hardware components accessed by admin to monitor the activity around house without physical presence. We propose a multifeatured intrusion and hazard detection alarm system using IOT and object detection.

Keyphrases: Internet of Things, Tensor Flow, YOLO, machine learning, object detection

@booklet{EasyChair:3212,
  author    = {Apurva Amrutkar and Samruddhi Mistari and Krutika Thambave and Rajshree Pandhare},
  title     = {Home Security Using IOT and Machine Learning},
  howpublished = {EasyChair Preprint 3212},
  year      = {EasyChair, 2020}}