Hi,

I started a project on my own sometime ago mainly to do the automated detection and catalog matching (even in low quality/amateur imaging). Meanwhile I added other futures such as to show asteroids or artificial earth satellites passing through at the time the picture was taken, image tagging, instruments control, resource management, online collaboration etc.

You can check my project at: larryo.org/work/information/unimap/index.html

I am almost ready to release it in a pre-alpha version, however I need some help to test it as well as to see how it behaves in different environments (I implemented it on windows xp). Also, there is still work to be done to improve the user interface and optimize the performace.

Any feedback will be appreciated!

Larry

Hi Larry,
Do you have a link where we can download the software and give it a test run. ?

Hi Dave,

I do have a link I will send/post it once I finish a use/test guide and I must solve some moremory leak propblems I just discovered. I also hope to trim down the archive a little bit as now it stands at 80Mb but I think I can reduce it down to 30Mb.

Is anyone here inetersted in a linux release as well? I am asking this because most of the code I wrote is in c++ with wxwidgets so I can attempt to compile it on different platforms(ie linux and mac) ... well execept for the digital camera support which for the moment is windows only.


Larry

Interesting stuff. Does this actually do plate-solving?

Larry,

I see this does do plate solving. This would have a lot of interest for minor planet observers. Once you have a beta version I would be happy to test it on some real astrometry and compare the results against the Minor Planet Centers data. Are the computed asteroid positions derived using a 2 body or N body solution?

Yes, it does do plate solving. The derived asteroids position is computed using the 2-body solution. I also had a look at a solution for n-body computation but I had to leave that for latter (to much on my plate right now:)

Ain;t research grade yet here:) and a lot yet to test, however you can set the software to use the very latest astrob.dat file which should yell an arcsecond error or less for most not earth approaching asteroids with the 2-body solution.

Larry

Hi larry, Yes I have looked into N body solutions too, still have the pain in my head from it! I have a good book here 'Methods of Orbit Determination - Boulet' It's tough reading, but it descibes some elegant algorithms for n body solutions. However if one is just looking to derive astrometric positions from an image, there is no need for any orbit solutions at all. Just an accurate astrometic position from the image.

Anyway, nice job, look forward to seeing your finished work.

Dave.

Thanks Dave!

I was thinking to be looking into n-body "problem" after I will fine tune the multithreading engine and also add an extra punch with gpu processing to it

Larry

Larry - replying to your response in another thread (I should made my comment in this thread).

I believe an amateur should be able to get down to 0.2" rms residuals or at least very close. Only time I've done it I was close to 0.2 in one axis - but nearly 0.5" in the other - I used 20 stars and one was badly mapped in. I was experimenting with fitting a WCS to the FITS header.

Do you intend to allow the user to fit WCS (I believe astrometrica does this)?

I've installed linux and IRAF in a virtual machine on my laptop now - was a bit painful. I want to see just what I can do when using the same tools as the professional crowd.
I've learned stuff enroute that I didn't expect to. I think its going to be a fantastic learning opportunity.

My next step is to see if I can get WCSTools installed and working. Manually fixing up WCS is well doable with basic IRAF but I'd like to make it easier.

Good luck with your project.

Mark

Hi Mark,

Aaa I presume you calculated residuals for the best 20 of them right? Try with all especially were you have field distortion caused by optical equipment/setup – thing which for the average amateur setup will be hard to subtract. Or maybe we are not talking about the same thing here?
However it’s a good idea to display that piece of info as well, I wrote it down.

I also used IRAF and the “bunch” of tools associated and wcstools and as well when though some pain at the time with iraf .. I don’t recall exactly the reason.

Yes, it “supports” WCS in FITS headers … I quote because right now isn’t plugged in the new image i/o engine. The initial image engine I use did not support enough image formats as I wanted so I scraped that for the current library I am using which is better but then again I am not happy because is not very efficient(memory wise) in reading big files so I will probably go in the future for something better and special formats like fits I think I going to implement separately.

Wow, I see people here going serious about this and I was thinking to scrap some parts from my projects as I thought it will make things look too complicated.

Anyway to clarify things a little, I started this project back in 2004 as I realize that more an more people get into astrophotography and I thought that it will be good for them to have a simple to use tool to allow them to explore and learn about the images they are taking. My idea was to just load it in your image which will be solved and then with a few clicks to learn about the stars, nebula, galaxies, distances, etc by automatically fetching information from various sources.
I think that only a few astrophotographers go further then just making nice pictures of the sky and I believe many will go on and learn more about what they shot if there were to be an easy to use tool to allow them to explore and present the information in an easy to digest way. So for a start I am looking more at this software as an educational tool rather then research grade. Next will be to tune down the application to fit the maximum precision possible and have advanced options in it for higher end users – but in this case probably it will be better two release a separate application …
Then again I don’t want to compete with pro/highend tools since these are out there - stuff like iraf/gaia/etc some of them are a bit of nightmare to put together and learn to use them but for someone who wants to play hard is just a matter of time …

Maybe I am wrong the way I thought about it? Any thoughts?

Larry

I like your last post.

Yes - the easier it is for amateurs to visulaise what they are doing and (where there is interest) do metrics the better. I do very little of either at present. I'm very definitely an armchair astronomer. My scope comes out infrequently. I love the maths and concepts and cloudy weather doesn't hinder me too much.

To do serious stuff requires some degree of competancy - and support - and the support is not really there for amateurs. I therefore totally endorse your initiative in providing tools that suport amateurs. However, sometimes it is education that is lacking in the community and not necessarily the tools. Tools (to do almost anything) *are* freely available.

Though having said that, amateurs generally have a lesser stringent operational environment than professionals (there's less that they can do to high precision) and so the software tools do not need to be so comprehensive. It is this area, to some extent, that Astrometrica and, if I may say, your tool, can fill a niche - they empower amateurs without burdening them.

I do feel that going forward amateur WCS fitting should be important. I say this because I believe that it is important that if one amateur astronomer at one place and time says "I have an image and there's a fuzzy" that we can easily compare images/data of observations by others with disparete setups - WCS facilitates this. But we need a means of sharing the data - with appropriate intellectual property restrictions.

Woudn't it be great if an amateur could image an undefined "fuzzy" and be able to trawl a database of images of similar recent field of view for confrmatory support / denial. The time is right for such a service - and WCS will be needed to do this.

I'm ranting, forgive me (I'm an armchair astronomer)

Mark

Hi Mark,

I agree, that’s why I added to my software the online upload/download/browse and a simple(now) search engine ... so the user can share the data(image and astrometry data) directly from within the tool.
Well only the basics for this are implemented and I am holding the astrometric data in a mysql database (so I though will be best to access/search it easier/faster with quries rather then in the header of the fits) … but I can do both no problem (In any case I use WCS standard)… anyway this and everything else is up for debate. I’m open for any suggestions. Ahh … talking of standards … I hope someone out there will think to create a standard for catalogs because believe me it was a real pain to write code to import over 50 different catalogs all in different exotics formats – and there are thousands of catalogs with all sort of information – it will be some to put all this together …

I was thinking of something else the amateur could do … helping revise catalogs. The big ones like usno b1.0 as you know are generated by means of automated plate solving and prove to be inaccurate in some cases. And here you need the human factor and I believe a distributed project like galaxyzoo only little bit more complex could help build better catalogs. And if you ask yourself why anyone will do such a boring task – well you can make this look like a game of some sort and then the competition spirit will kick in. well .. is late and I am rambling …

Larry

mjc wrote:

To do serious stuff requires some degree of competancy - and support - and the support is not really there for amateurs. I therefore totally endorse your initiative in providing tools that suport amateurs. However, sometimes it is education that is lacking in the community and not necessarily the tools. Tools (to do almost anything) *are* freely available.

Mark, I would say that if amateurs are willing to put the effort ino gaining the required competencies to do research grade work then (at least in my experience) professionals are more than happy to lend a hand. I think where *some* amateurs fall down is that they often want it 'handed on a plate'. I think this is where the pros draw the line. (and I don't blame them). In my own area of minor planet research, I've found on many occassions some of the pros from the big surveys often offer unsolicited insights or assistance. For example last year, I was asking a question about uncertainty maps for near earth asteroids. One of the big surveys took their main scope off the survey to demonstrate a point under discussion.

I take my hat off to Larry for the excellent work. As an amateur astronomer and a professional programmer, I fully appreciate the incredible amount of work and effort that goes into something like this. In many ways as a programmer you try to take a PHD (Push Here Dummy) approach to developing tools. Whereas this is a lot of effort for the developer, it can sometimes oversimplify a process for the end user leading to a situation where the end user can use the tool but has no idea about the hows and whys of what is actually going on.

Often I meet people who are full of enthusiasm for minor planet research until a point comes where they realise that in fact if can be damned hard work and then they disappear. Or some even don't get that far because they couldn't be bothered to read the documentation.

What I'm taking forever to say is that there is a huge amount an amateur can do if he/she is prepared to put in the effort, but alas many want the gain but no pain.

Larry: I like your idea about catalog refinement, but what I'd *really* love to see is a project where the amateur ends up developing a real skillset and some more detailed background knowledge about what it is they're doing. For example a version of galaxy zoo which enables users to learn why galaxies are the shapes they are. Maybe teaching the user some elements of statistical distributions and then letting the user play with the data and draw some conclusions. Of course many would run a mile from something like this, but thankfully many (just like you guys) would see it for the opportunity it presented and grab it with both hands.

Good resposne, Dave

I think I know what larry was driving at re: catalogues - the disparate format and information content. What might improve the situation there is XML tagging.

There appears to be some movement in this direction as indicated in this reference
www.adass.org/adass/proceedings/adass99/O9-04/
which might interest Larry. I've read this page and it interests me - I haven't read any further to determine if there's any actual implementations yet.

With regards to your coments re shapes of galaxies and understanding of same, Galaxyzoo recently announced the launch of a new project where users compare simulated galactic collisions and match the best outcomes against actual galaxy formations. I think very closely related to your suggestion.
See: www.galaxyzooblog.org/2009/11/24/galaxy-...ding-cosmic-mergers/

I'm reasured of your comments regarding support you have had from the MPC arena but I feel I have to really trawl hard for information at times. Though I have to say support in the form of access to a lot of professional resources is there but it's been a bit of a slog to determine what I should be looking for (for whatever I might be interested in at the time), then finding it, then understanding it. Though sometimes I think that it is this that I probably enjoy most.
I'm very tempted to get myself an Athens account because I'm frustrated at times when I can only get the abstract of a paper.

Mark

DaveGrennan wrote:

Whereas this is a lot of effort for the developer, it can sometimes oversimplify a process for the end user leading to a situation where the end user can use the tool but has no idea about the hows and whys of what is actually going on.

Well, in the idea I wrote this, for the main purpose of expanding the knowledge of the amateur astronomer about what is in their picture I am not sure that is so important they know about the maths behind star matching process The simple PHD approach is backed by raw data exported in a plain text format so there are “plugs” for more advanced studies on the images. Also you will find there are a lot of options to tune the parameters which will require a decent knowledge about the process if you want to play harder. Then again I don’t think the amateur astronomer to be proficient in linux (or other unix flavors or in general it) in order to get an image solved

DaveGrennan wrote:

Larry: I like your idea about catalog refinement, but what I'd *really* love to see is a project where the amateur ends up developing a real skillset and some more detailed background knowledge about what it is they're doing.

Yes, I know you mean. I was disappointed as well about how “bold” that galaxyzoo task was – somehow when I heard of them I expected something more orientated towards the amateur astronomers community rather then general public… Probably the best approach will be a “game” where the user can acquire small bits of information in steps (nothing to hard at once to scare them away) by following nicely done interactive tutorials and by this allow them to advance hierarchically - but this will require some resources – but then again we are a small army of amateur astronomers

mjc wrote:

Galaxyzoo recently announced the launch of a new project where users compare simulated galactic collisions and match the best outcomes against actual galaxy formations.

Yes, this projects is slightly better then the initial one. Yet they hold on the idea of being open to general public only – I think they should provide an advanced interface where you can customize your simulations more precisely by modifying all kind of values/parameters of your simulation along with a good tutorial as what every parameter stands for and what impact might have on your simulation. Also, I think these projects should be more open as to provide access to the data resulted from analysis and in other aspects as well.

mjc wrote:

I think I know what larry was driving at re: catalogues - the disparate format and information content. What might improve the situation there is XML tagging.

Yes, this will help, but there is also a lack of standard in the way measurements are presented – like for position in decimal and sexagesimal or the units used for various measurement or the way magnitude is expressed in various catalogs or naming conventions. I know that in most cases there is a reason behind the way that catalog was built but I think having a better standard will ease the task of centralizing all these thousands of catalogs with various measurements or verify/compare them.

Larry

I finally managed to build the pre-alpha version – with a slight delay as my life has been complicated lately.
You can download and test the very first pre-alpha release from here:

larryo.org/work/information/unimap/download.html

I know there are still bugs and room for of improvements, optimization, etc. "Unfortunately" I have built this application so big for so little time I’ve got that I should probably be focusing on getting the a core functionality working right and here any suggestions and ideas will really help.




Larry

Anyone tried the application yet? Any thoughts?

Larry

I am wondering if people here find this application useful? At the concept level what do you think I should change/add to improve it?

Hi Larry,

I haven't used it but browsing through it (at a glance) it's a smart piece of application software.

Seanie.

There tons of smart things out there ... many not really useful ... or practical

That's true. How good are your software skills at making a similar app for Android smartphone users? I know there are lots out there, but supporting those a little closer to home are more my thing.

Seanie.

Version 0.0.2 pre-alpha is now available for download from:

larryo.org/work/information/unimap/download.html

Larry

Version 0.0.3 pre-alpha is now available for download from:

larryo.org/work/information/unimap/download.html

Larry

I have also added two small video tutorials (basic) here:

larryo.org/work/information/unimap/support.html

I am not sure if this format is ok, feedback will help before I go on to make better tutorials.

Larry