Enhancement of Fire Detection based on Deep Learning using Manta Ray Foraging Optimization (MRFO)
Smart surveillance, decision making, and automated response systems are trending topics nowadays, especially after performance’s growth for a wide range of control systems, which is a normal result of the noticeable evolution in the areas of artificial intelligence (AI) techniques made dramatically increasing in performance in terms of accuracy and results that have been reached for various systems and algorithms used for different applications in all fields. This development enabled human beings to make a lot of tasks that were done within manual manners much easier than before, not only easier but less time and power-consuming with reducing the numbers of essential operators. In this research, a model was produced for Fire detection, and the data set used was 8000 images, gathered from different sources from the Internet and by making some fires and taking a lot of images with different circumstances. The module achieved an accuracy of about 97.32%. Many tools have been used for various types of necessary tasks which will be parts of the desired application, such as Python 3.7, which was used for building the essential algorithms, KERAS framework, which provided the deep learning algorithms, Visual Studio Code (VSC) as an Integrated Development Environment (IDE), and Anaconda navigator for downloading the different compatible libraries. The proposed models were trained on Processor Intel(R) Core(TM) i7-10750H CPU @ 2.60GHz 2.59 GHz, RAM 16.0 GB, 64-bit operating system, x64-based processor, NVIDIA GeForce GTX 1650 GPU.