Related Experiment Video
Updated: Jul 2, 2026
05:57
Long-term Video Tracking of Cohoused Aquatic Animals: A Case Study of the Daily Locomotor Activity of the Norway Lobster (Nephrops norvegicus)
Published on: April 8, 2019
BEML-sonar: a bio-inspired echolocation and machine learning-enhanced SONAR for underwater object detection and
C Kishor Kumar Reddy1, Vijaya Sindhoori Kaza2, P R Anisha1
1Department of Computer Science and Engineering, Stanley College of Engineering and Technology for Women, Hyderabad, India.
Scientific Reports
|June 30, 2026
Summary
This study introduces SonarNet, a novel machine learning-enhanced sonar system inspired by biological echolocation. SonarNet significantly improves underwater object detection accuracy and reduces energy consumption compared to traditional sonar systems.
Area of Science:
- Marine Technology
- Artificial Intelligence
- Signal Processing
Background:
- Traditional SONAR systems face limitations including high energy consumption, noise interference, and signal degradation in diverse aquatic environments.
- Biological echolocation offers a model for robust and efficient underwater sensing.
Purpose of the Study:
- To develop a novel machine learning-enhanced sonar model inspired by biological echolocation.
- To improve accuracy, robustness, and energy efficiency in underwater sensing applications.
Main Methods:
- The proposed SonarNet model integrates bio-inspired echolocation principles with dynamic acoustic pulse adjustment.
- Deep learning techniques, including Convolutional Neural Networks (CNNs) and Long Short-Term Memory (LSTM) networks, are used for echo classification and object detection.
- Digital signal processing (DSP) methods such as Butterworth filtering, wavelet decomposition, and adaptive thresholding are employed for noise mitigation and signal enhancement.
Main Results:
- SonarNet achieved 92.7% classification accuracy, surpassing conventional SONAR methods (85.4%).
- Adaptive signal processing resulted in a 20.8% reduction in energy consumption and a 15.3 dB improvement in Signal-to-Noise Ratio (SNR).
- The model demonstrated an 18.6% reduction in false positive rate and a 12.4% improvement in depth estimation accuracy via Sound Velocity Profile (SVP) correction.
Conclusions:
- The proposed SonarNet model offers a significant advancement over traditional SONAR systems for underwater sensing.
- The integration of machine learning and advanced DSP techniques enhances performance and efficiency in challenging aquatic conditions.
