Deep Learning and Machine Vision based Robot for Fire Detection and Control

Abstract

Robotics and deep learning algorithms have emerged as a crucial part of our everyday life. Most of the recent image recognition innovations use deep learning, big data analysis, and artificial intelligence as their core platform. Hence in this work, a firefighting robot that uses deep learning is introduced to detect fire. On detection, the type of fire is also identified with the help of a combination of Imagenet and Alexnet. A Raspberry Pi is used to execute the proposed work and is programmed to function efficiently to detect and classify fire. Based on the results obtained, it is concluded that classification accuracy stands at 92%, while the detection of fire is accurate up to 97.75%. The proposed work shows a dynamic improvement in the accuracy of detecting fire with the help of robots and further classifying them using machine vision.