RANSAC Compressive Sensing Reconstruction in Noisy Wideband Underwater Sonar Imaging
- Resource Type
- Conference
- Authors
- Stankovic, Isidora; Ioana, Cornel; Dakovic, Milos
- Source
- Global Oceans 2020: Singapore – U.S. Gulf Coast. :1-5 Oct, 2020
- Subject
- Aerospace
Communication, Networking and Broadcast Technologies
Computing and Processing
Engineering Profession
General Topics for Engineers
Geoscience
Power, Energy and Industry Applications
Robotics and Control Systems
Signal Processing and Analysis
Transportation
Sonar
Robustness
Sensors
Noise measurement
Image reconstruction
Compressed sensing
Wideband
Compressive Sensing
OMP
RANSAC
Sonar imaging
Sparsity
Noisy
- Language
The reconstruction of noisy wideband underwater sonar images using the improved techniques of compressive sensing is analyzed in this paper. The received signal is considered with different kinds of additive noise, with some of the samples being affected by a high-impulsive noise. The sonar images are assumed to be sparse, which makes the compressive sensing (CS) theory suitable for their processing and reconstruction. The goal is to localize and reconstruct targets using a combination of the CS technique and a sonar signal sample selection method (RANSAC). Numerical examples confirm the robustness of the proposed method.