Low Complexity Adaptive Sonar Imaging

The Low Complexity Adaptive (LCA) beamformer is a technique used to improve image quality of e.g. sonar images by adapting to the data in the scene. It can be viewed as either a simplified version of the fully adaptive minimum variance distortionless response (MVDR) beamformer, or as an adaptive extension to the conventional static delay and sum (DAS) method.

This page provides supplementary material (2 movie files + source code) to the article Low Complexity Adaptive Sonar Imaging published in IEEE Journal of Oceanic Engineering, 2016, DOI: 10.1109/JOE.2016.2565038.

The Holmengraa shipwreck imaged with the DAS, LCA and MVDR beamformers.

Media 1 - Typical MVDR window responses

Description:	Typical spatial amplitude responses for the windows a robust MVDR implementation selects in shadow, highlight and speckle regions in an image. The shipwreck is of Holmengraa located outside Horten, Norway.
Size:	28.6MB
File type:	mp4 (MPEG-4)

Media 2 - Effect of adjust number of windows

Description:	Mosaic of LCA images showcasing the effect of the number of windows it is allowed to choose from.
Size:	28.6MB
File type:	mp4 (MPEG-4)

Source Code

Python function for generating array tapers/windows for LCA: calcWindow3dSimple.py

Contact Info

Jo Inge Buskenes
University of Oslo
Oslo, Norway

joibu@ifi.uio.no

By Jo Inge Buskenes

Published May 27, 2016 10:55 AM - Last modified May 27, 2016 6:10 PM