Square Root Compression and Noise Effects in Digitally Transformed Images

We report on a particular example of noise and data representation interacting to introduce systematic error into scientific measurements. Many instruments collect integer digitized values and apply nonlinear coding, in particular square root coding, to compress the data for transfer or downlink; th...

Full description

Saved in:
Bibliographic Details
Main Authors: Craig E. DeForest, Chris Lowder, Daniel B. Seaton, Matthew J. West
Format: Article
Language:English
Published: IOP Publishing 2022-01-01
Series:The Astrophysical Journal
Subjects:
Online Access:https://doi.org/10.3847/1538-4357/ac7f3d
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850173813049786368
author Craig E. DeForest
Chris Lowder
Daniel B. Seaton
Matthew J. West
author_facet Craig E. DeForest
Chris Lowder
Daniel B. Seaton
Matthew J. West
author_sort Craig E. DeForest
collection DOAJ
description We report on a particular example of noise and data representation interacting to introduce systematic error into scientific measurements. Many instruments collect integer digitized values and apply nonlinear coding, in particular square root coding, to compress the data for transfer or downlink; this can introduce surprising systematic errors when they are decoded for analysis. Square root coding and subsequent decoding typically introduces a variable ±1 count value-dependent systematic bias in the data after reconstitution. This is significant when large numbers of measurements (e.g., image pixels) are averaged together. Using direct modeling of the probability distribution of particular coded values in the presence of instrument noise, one may apply Bayes’ theorem to construct a decoding table that reduces this error source to a very small fraction of a digitizer step; in our example, systematic error from square root coding is reduced by a factor of 20 from 0.23 to 0.012 count rms. The method is suitable both for new experiments such as the upcoming PUNCH mission, and also for post facto application to existing data sets—even if the instrument noise properties are only loosely known. Further, the method does not depend on the specifics of the coding formula, and may be applied to other forms of nonlinear coding or representation of data values.
format Article
id doaj-art-07d8fd82452e4bcdae55c0016388c1d4
institution OA Journals
issn 1538-4357
language English
publishDate 2022-01-01
publisher IOP Publishing
record_format Article
series The Astrophysical Journal
spelling doaj-art-07d8fd82452e4bcdae55c0016388c1d42025-08-20T02:19:47ZengIOP PublishingThe Astrophysical Journal1538-43572022-01-01934217910.3847/1538-4357/ac7f3dSquare Root Compression and Noise Effects in Digitally Transformed ImagesCraig E. DeForest0https://orcid.org/0000-0002-7164-2786Chris Lowder1https://orcid.org/0000-0001-8318-8229Daniel B. Seaton2https://orcid.org/0000-0002-0494-2025Matthew J. West3https://orcid.org/0000-0002-0631-2393Southwest Research Institute , 1050 Walnut Street, Suite 300 Boulder, CO 80302, USA ; deforest@boulder.swri.eduSouthwest Research Institute , 1050 Walnut Street, Suite 300 Boulder, CO 80302, USA ; deforest@boulder.swri.eduSouthwest Research Institute , 1050 Walnut Street, Suite 300 Boulder, CO 80302, USA ; deforest@boulder.swri.eduSouthwest Research Institute , 1050 Walnut Street, Suite 300 Boulder, CO 80302, USA ; deforest@boulder.swri.eduWe report on a particular example of noise and data representation interacting to introduce systematic error into scientific measurements. Many instruments collect integer digitized values and apply nonlinear coding, in particular square root coding, to compress the data for transfer or downlink; this can introduce surprising systematic errors when they are decoded for analysis. Square root coding and subsequent decoding typically introduces a variable ±1 count value-dependent systematic bias in the data after reconstitution. This is significant when large numbers of measurements (e.g., image pixels) are averaged together. Using direct modeling of the probability distribution of particular coded values in the presence of instrument noise, one may apply Bayes’ theorem to construct a decoding table that reduces this error source to a very small fraction of a digitizer step; in our example, systematic error from square root coding is reduced by a factor of 20 from 0.23 to 0.012 count rms. The method is suitable both for new experiments such as the upcoming PUNCH mission, and also for post facto application to existing data sets—even if the instrument noise properties are only loosely known. Further, the method does not depend on the specifics of the coding formula, and may be applied to other forms of nonlinear coding or representation of data values.https://doi.org/10.3847/1538-4357/ac7f3dAstronomy data reductionMeasurement error modelCoronagraphic imagingDirect imaging
spellingShingle Craig E. DeForest
Chris Lowder
Daniel B. Seaton
Matthew J. West
Square Root Compression and Noise Effects in Digitally Transformed Images
The Astrophysical Journal
Astronomy data reduction
Measurement error model
Coronagraphic imaging
Direct imaging
title Square Root Compression and Noise Effects in Digitally Transformed Images
title_full Square Root Compression and Noise Effects in Digitally Transformed Images
title_fullStr Square Root Compression and Noise Effects in Digitally Transformed Images
title_full_unstemmed Square Root Compression and Noise Effects in Digitally Transformed Images
title_short Square Root Compression and Noise Effects in Digitally Transformed Images
title_sort square root compression and noise effects in digitally transformed images
topic Astronomy data reduction
Measurement error model
Coronagraphic imaging
Direct imaging
url https://doi.org/10.3847/1538-4357/ac7f3d
work_keys_str_mv AT craigedeforest squarerootcompressionandnoiseeffectsindigitallytransformedimages
AT chrislowder squarerootcompressionandnoiseeffectsindigitallytransformedimages
AT danielbseaton squarerootcompressionandnoiseeffectsindigitallytransformedimages
AT matthewjwest squarerootcompressionandnoiseeffectsindigitallytransformedimages