Underwater Object Prediction Using Sonar Waves

Abstract

The detection and differentiation of potentially hazardous objects, such as mines and rocks, in underwater environments are crucial for the safe navigation of submarines during warfare scenarios. This study compares and contrasts different machine learning algorithms to determine whether an object detected by a submarine's sonar system is a mine or a rock. From the comparative analysis, we have proposed the Logistic Regression Model (LRM), which is used to train and test the models. To extract pertinent features like signal intensity, frequency content, and time-domain characteristics, sonar signals are essential to the labeled dataset. This dataset is made up of sonar signals and corresponding object classifications (mine or rock) based on the frequency range. These features served as inputs to the machine learning algorithms and enabled the features to learn the underlying patterns and make accurate predictions. In this paper, we have contributed a novel approach to underwater object detection using LRM, which is applied to sonar data. In summary, this research presents an innovative solution to the challenge of underwater object detection using sonar signals and machine learning algorithms. The developed model exhibits promising results, opening up new possibilities for advancing our understanding of underwater environments and enhancing underwater exploration and security capabilities.