Aural Based Scene Understand for Sonar Applications

Abstract

Investigation of sonar returns using an information theoretic framework allows for feature extraction and scene parsing. We seek to characterize seabed and oceanographic features ranging from nearly stationary to episodic whether spatially or temporally. The approach will examine the information in individual and combined sonar echo returns as it pertains to forming auditory objects that combine to form a scene. In particular, we will investigate the application of summary statistics in auditory perception as a means for representing underlying scene characteristics such as sand ripples, gravel beds, or sea grass. We will explore ways to encode, compactly store, and update streaming auditory data that both establishes a capability for background/object separability yet is sensitive to background changes.

Document Details

Document Type
DoD Grant Award
Publication Date
Aug 20, 2019
Source ID
N000141912679

Entities

People

  • Daniel C Brown

Organizations

  • Office of Naval Research
  • Pennsylvania State University
  • United States Navy

Tags

Readers

  • Coastal Oceanography
  • Computer Vision.
  • Theoretical Analysis.

Technology Areas

  • AI & ML