skip to main content
OSTI.GOV title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: Edge Detection to Isolate Motion in Adaptive Optics Systems

Abstract

Adaptive optics uses signal processing techniques and deformable mirrors to minimize image degradation caused by phase aberrations. In the case of telescope imaging, the atmosphere causes phase aberrations. In the case of satellite imaging, errors due to the ultra-light-weight characteristics of the primary mirror cause phase aberrations. Scene-based Shack-Hartmann Wave Front Sensing takes the correlation between successive wavelets to determine these phase aberrations. A large problem with the scene-based approach is that motion, such as a moving car, can cause the correlation of two lenslets to peak, not where the scenes align, but where the moving object in each frame aligns. As such, the continued use of scene-based Wave Front Sensing necessitates successful isolation of moving objects from a stationary background scene. With the knowledge of which pixels are immobile, one should avoid the problem of locking onto a moving object when taking the correlation of two successive frames in time. Two main requirements of isolation are consistency and efficiency. In this document I will discuss the different edge detection algorithms explored for moving object isolation and how I came to the conclusion that, for our purposes of scene-based Shack-Hartmann WFS, edge detection is too inconsistent to be of anymore » use. Because the Shack-Hartmann lenslets limits us to low resolutions, edge detection that works on higher resolution images will not work on our images. The results of each algorithm will show that with so few pixels per subaperature, edge detection is a poor method of identifying moving objects.« less

Authors:
Publication Date:
Research Org.:
Lawrence Livermore National Lab. (LLNL), Livermore, CA (United States)
Sponsoring Org.:
USDOE
OSTI Identifier:
15004551
Report Number(s):
UCRL-ID-154782
TRN: US201015%%719
DOE Contract Number:  
W-7405-ENG-48
Resource Type:
Technical Report
Country of Publication:
United States
Language:
English
Subject:
42; ALGORITHMS; AUTOMOBILES; DETECTION; EFFICIENCY; EYES; MIRRORS; OPTICS; PROCESSING; RESOLUTION; ROADS; SATELLITES; TELESCOPES

Citation Formats

Chan, C W. Edge Detection to Isolate Motion in Adaptive Optics Systems. United States: N. p., 2003. Web. doi:10.2172/15004551.
Chan, C W. Edge Detection to Isolate Motion in Adaptive Optics Systems. United States. https://doi.org/10.2172/15004551
Chan, C W. Fri . "Edge Detection to Isolate Motion in Adaptive Optics Systems". United States. https://doi.org/10.2172/15004551. https://www.osti.gov/servlets/purl/15004551.
@article{osti_15004551,
title = {Edge Detection to Isolate Motion in Adaptive Optics Systems},
author = {Chan, C W},
abstractNote = {Adaptive optics uses signal processing techniques and deformable mirrors to minimize image degradation caused by phase aberrations. In the case of telescope imaging, the atmosphere causes phase aberrations. In the case of satellite imaging, errors due to the ultra-light-weight characteristics of the primary mirror cause phase aberrations. Scene-based Shack-Hartmann Wave Front Sensing takes the correlation between successive wavelets to determine these phase aberrations. A large problem with the scene-based approach is that motion, such as a moving car, can cause the correlation of two lenslets to peak, not where the scenes align, but where the moving object in each frame aligns. As such, the continued use of scene-based Wave Front Sensing necessitates successful isolation of moving objects from a stationary background scene. With the knowledge of which pixels are immobile, one should avoid the problem of locking onto a moving object when taking the correlation of two successive frames in time. Two main requirements of isolation are consistency and efficiency. In this document I will discuss the different edge detection algorithms explored for moving object isolation and how I came to the conclusion that, for our purposes of scene-based Shack-Hartmann WFS, edge detection is too inconsistent to be of any use. Because the Shack-Hartmann lenslets limits us to low resolutions, edge detection that works on higher resolution images will not work on our images. The results of each algorithm will show that with so few pixels per subaperature, edge detection is a poor method of identifying moving objects.},
doi = {10.2172/15004551},
url = {https://www.osti.gov/biblio/15004551}, journal = {},
number = ,
volume = ,
place = {United States},
year = {2003},
month = {7}
}