Smart Technology

Fire outbreak has been a common issue in forests and large buildings. In this research we are using a drone to detect forest and building fires which uses the techniques of image processing and video processing. Our research can be divided into three main modules. The first module focuses on the cloud service amazon cloud to perform data analysis.
The Data collected by the drone, with the help of flame sensor attached to Aurdino, is transferred to amazon cloud.This is achieved through the Aurdino. The second module is based on image processing. With the help of inbuilt camera module, the drone captures real time images and image processing is done on that image using google API. The algorithm uses content based image retrieval and detects whether fire is caught inside the image or not, along with its intensity. The algorithm also detects the objects present nearby the fire.
Hence if people are present or trapped nearby the fire we can detect them and prioritize the areas which needs immediate attention. The third module focuses on video processing. The algorithm detects fire based on the color of the pixels captured. All the three modules are together integrated in this research to boost up the accuracy of fire detection by the drone. Uniqueness of this research is that it comprises of 3 different modules each having different but inter-related functions. All three modules boost up the fire detection accuracy and can thus help in saving lives.