Title: An approach into navigation and vision for autonomous fire fighting robots

Authors: Shaun Q.Y. Tan; V.J. Karthik; A. Govind; P.M. Rajasree

Addresses: Electronics and Instrumentation Department, RV College of Engineering, Bengaluru, India ' Electronics and Instrumentation Department, RV College of Engineering, Bengaluru, India ' Electronics and Instrumentation Department, RV College of Engineering, Bengaluru, India ' Electronics and Instrumentation Department, RV College of Engineering, Bengaluru, India

Abstract: Fires are dangerous and are a threat to life, property and the environment, if left uncontrolled. There is thus a need for technology, especially the use of unmanned autonomous wheeled rescue robots to tackle such situations. In this paper, a path planning algorithm is developed using a combination of image processing techniques along with the A* algorithm which is able to find a path within an indoor environment, given a fire escape plan as its input. In addition, a fire recognition and localisation system is also presented, which combines a particle swarm optimisation (PSO) optimised convolutional neural network (CNN) with various image processing techniques to recognise and localise fire sources in input images. The scope and future improvements for this work are also discussed.

Keywords: convolutional neural network; CNN; particle swarm optimisation; PSO; image processing; path planning.

DOI: 10.1504/IJAMECHS.2023.132522

International Journal of Advanced Mechatronic Systems, 2023 Vol.10 No.3, pp.156 - 164

Received: 31 Mar 2022
Accepted: 24 Apr 2023
Published online: 25 Jul 2023 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article