ImageIntegration documentation updated (preview)

[Juan Conejero]

Hi all,

I have just finished a complete revision of the reference documentation for the ImageIntegration tool. I have rewritten many sections, and I have added new material to document new features. Lots of errors and inaccuracies have also been fixed. While I release official updates to the core application, a new documentation system and some new tools, here is a preview of the new ImageIntegration reference document:

http://pixinsight.com/doc/tools/ImageIntegration/ImageIntegration.html

On the theoretical part of this document, I have tried to be much more rigorous and comprehensive with many key topics of the image integration task, with more detailed descriptions of the implemented algorithms. The practical part (the Usage section) is also more extensive, and documents all of the new features in the latest versions of the ImageIntegration tool. There are also more and better usage hints at the end of the document. I hope you like it and find it useful. As soon as I have another "time window", I'll start a complete revision of the documentation for StarAlignment.

The above document has been created with the latest version 1.4 of the PixInsight Documentation Compiler script, which I'll also release very soon. Among many improvements, the new compiler comes with automatic rendering of embedded LaTeX equations, automatic numbering of references, equations, tables and figures through symbolic references, a new document class for the PixInsight JavaScript Runtime, and generation of dynamic contents. Note for example what happens when you mouse over bibliographic and equation references.

[Sean]

Juan,

This is an impressive piece of work, both in terms of understanding the use of Image Integration in PI, and also gaining a deeper understanding of the general principles involved. I look forward to spending some time reading it in detail - it's like taking a course!

Thanks for all your hard work,

Sean

[Philippe B.]

Thank you Juan, nice and very useful !!!
Cheers

[georg.viehoever]

Epic! The printed version is 43 pages. That's more documentation for a single module than I have seen as full documentation for many other "professional" products. Plus it is much more comprehensive than the usual "click here, than drag this..." type of documentation that you often get these days.

Will study it in the bath tub 
Georg

[bitli]

Thank you. As usual it is more than a documentation, it is a piece of art.
I cannot overstate how much I feel that the documentation adds a lot of value to PI. It is both advanced and  by nature very relevant to the product.
Hope to see more,
Your never satisfied :-)
-- bitli

[jkmorse]

Juan,

It is this type of intellectual rigor that makes using PI so damned much fun. 

Thanks for another outstanding piece of work.

Jim

[dayers]

+1

[pk825]

Great documentation!

I find a little error: the link of the SubframeSelector
(http://pixinsight.com/doc/scripts/SubframeSelector/SubframeSelector.html) in the middle of the document doesn't work.

[mschuster]

Hi Juan,

A note on the Sn and Qn Estimators of Rousseeuw and Croux section in your new documentation.

The Sn definition includes normalization factors, a sample size dependent scaling along with an additional standardized dispersion scaling.

It appears the Statistics process implementation does use these factors (at least for large sample size image test I did), but the documentation does not mention them.

Thanks,
Mike

[Juan Conejero]

Hi Mike,

The Sn definition includes normalization factors, a sample size dependent scaling along with an additional standardized dispersion scaling.

My implementation in PCL (and by extension on the whole PixInsight platform) applies correction factors for finite sample sizes. I haven't mentioned them because they are actually irrelevant for images, since in this context we are working with very large sample sizes, where these corrections are immaterial. Specifically, the correction factor c for sample sizes larger than 9 items is:

For Sn:
For odd sample size n: c = n/(n - 0.9)
For even sample size n: c = 1

For Qn:
For odd sample size n: c = n/(n + 1.4)
For even sample size n: c = n/(n + 3.8)

For coherence with already existing statistical estimators in PCL, the normalization factors are not applied by default in my implementation. They are not used in the ImageIntegration tool either, since in this case we are comparing the same scale estimates between pairs of images, and hence the normalization factors cancel out. The factors that make these estimators converge to the standard deviation of a normal distribution are:

For Sn: k = 1.1926
For Qn: k = 2.2219

[Juan Conejero]

Thank you all for your nice words. I'm glad you like this document.