Underwater acoustic target signal enhancement algorithm optimized by feature preservation and noise update Article Swipe
YOU?
·
· 2025
· Open Access
·
· DOI: https://doi.org/10.1088/1742-6596/2939/1/012004
· OA: W4407033354
With the development of shock absorption and noise reduction technology, it is difficult to use existing aquatic target feature extraction techniques to extract weak signal features from complex marine environments. Therefore, underwater acoustic signal enhancement has become more important. The overlapping characteristics of underwater acoustic target signals and the variability of the ocean acoustic field make the application of Non-negative Matrix Factorization (NMF) in enhancing underwater acoustic target signals ineffective. To this end, the classic NMF algorithm is adaptively improved by combining the characteristics of underwater acoustic target signals, and an improved NMF underwater acoustic signal enhancement algorithm based on feature preservation is proposed. The algorithm first applies divergence constraints and similarity detection to the NMF feature base matrix to eliminate redundancy, optimizing the NMF features with size invariant characteristics while avoiding coefficient dispersion caused by base vector similarity and resulting in base vector loss. Then, the real-time environmental noise received by sonar is used to improve the adaptability of NMF noise basis vectors, achieving noise reduction and enhancement of underwater acoustic target signals. The experimental results show that compared with classical NMF, manifold-constrained NMF, and other algorithms applied to underwater acoustic target signal enhancement, the method proposed in this paper achieves a better signal enhancement effect.
