Photo tagging

[CommGeek]

Well to my understanding it was hosted on a personal server by the creator.    From What I was told the member that built it had to remove it from the server because some 'real' paying clients 'moved in'. The Wing IT staff  dosent have a clue on how to host an application, and due to security concerns on the Wing server they are looking for a new home.

If its not web based how do you get the images to the client?  if they cant see them on a map, in my real world experience with the Guard, and State...they don't want anything to do with them.   Yes the software is free, the server script is free....it just needs a competent IT person to install it on a server thats not sitting in someones living room.    If we acre going to advertise that we can produce arial imagery we need to step up  and support it.

[CommGeek]

There was an idiot proof, fully automated sloution that was proposed to national that would cost NOTHING! The state of Florida, FEMA, and the Florida National Guard went crazy over the product...they were in love.   National turned it down...

I can think of plenty of programs, technology, and software, that one group loves and defines as "perfect" when in fact they are kludgey, personality-based, or simply don't meet the need in the eyes of the person who is empowered with the decision.

As an example, any platform or tool adopted these days which is not web-based and OS agnostic is a poor choice.

You guys are free to use it....were trying to find a new server for it.  The server we had isnt available anymore.  When we find a new home to host it ill post a link.

And thus perhaps one reason it wasn't adopted - it requires hosting somewhere, which is not "free".  That doesn't mean its not a useful
tool, but bandwidth, hosting, and administration for a tool that is to be used nationally is not a small proposition.

The "best" solution for this is likely something tied to WMIRS that does the tagging as part of the upload.

Oh yes...becouse WMIRS dosent need hosting, bandwidth, and admin.. either!!!       

[Eclipse]

Well to my understanding it was hosted on a personal server by the creator.    From What I was told the member that built it had to remove it from the server because some 'real' paying clients 'moved in'. The Wing IT staff  dosent have a clue on how to host an application, and due to security concerns on the Wing server they are looking for a new home.

If its not web based how do you get the images to the client?  if they cant see them on a map, in my real world experience with the Guard, and State...they don't want anything to do with them.   Yes the software is free, the server script is free....it just needs a competent IT person to install it on a server thats not sitting in someones living room.    If we acre going to advertise that we can produce arial imagery we need to step up  and support it.

Bingo - home grown hosted under someone's desk - this is specifically what I referred to as "personality based" - people depending on an asset not in control of the corporation and running at the whim of a single member.  A recipe for disaster.

There are lots of ways and places that these types of things can be run inexpensively, but the key is what the scope of user is and who has the keys.

In my AOR I've kiboshed more than one IT project that was "sliced-bread-level" but not scalable without benevolent hosting and administration.  For better or worse, that's a requirement.

That's why the Googlesphere is so compelling for most CAP needs.

[CommGeek]

let me clarify....the server was a real hosted server in a data center.... the developer is a REAL IT guy...

I wish i could say the same for the Wing IT Dept...

[Capt Rivera]

Commgeek,

got any screen shots?

[Ops Guy]

I own a computer company with servers here in the US and abroad. What do we need, just space and a website to make it available or space for the processing also, let me know.

[Capt Rivera]

Commgeek?

... anyone have a clue what he was talking about?

[CommGeek]

The system has actully found a home at national.  In a few weeks it will be available for all to use.

[Capt Rivera]

The system has actully found a home at national.  In a few weeks it will be available for all to use.

ARGUS?

Whatever it is its not a state secret...  please give useful information. I'm starting to think you actually don't know any more then anyone else...

[Eclipse]

let me clarify....the server was a real hosted server in a data center.... the developer is a REAL IT guy...

I wish i could say the same for the Wing IT Dept...

And is he willing to sign a real contract reviewed by real lawyers to provide the level of service and uptime required by a real national
organization involved in real search and rescue?

Is the hosting center prepared to donate several real years of real hosting for free, with a contract?

I've had plenty of conversations with people willing to "stash a server" in an open rack slot - until they leave CAP or get mad
about their promotion being delayed and then the whole thing goes away.

We don't create national systems based on a handshake.

[Ops Guy]

If its free why don't you share it with those of us who might use it, now, rather than waiting for a long drawn out server placement.

[tinker]

I'm looking for a program that will tag photos per 1AF standards. I'm looking for one that will accept .gpx or .kml as the input track.

Take a look at this: http://www.scc.net/~mitty/Phyllis/Phyllis_Intro.pdf

Phyllis can be considered to be either a stand-alone program (no internet required) or a front end for Argus.  Phyllis acquires and organizes the photos from the cameras, which Argus does not.  Argus provides means to view the photo locations on Google maps and provides a distribution mechanism, which Phyllis does not.  Both support photo manipulation with visual and EXIF tagging.  Both use identical visual and EXIF syntax.  If the photo work is done within Phyllis, a complete (all photos and track files) zip file for uploading to Argus can be generated at the user's option.

Re track files, Phyllis will take GPX or NMEA syntax.  Argus will take only NMEA, so Phyllis automatically translates GPX files to that format for the upload.  Neither program takes KML files.  (At least that is the Argus status as I know it.  Troy may have plans to add other track file formats in the near future; I have not discussed it with him lately.)

[davidsinn]

I'm looking for a program that will tag photos per 1AF standards. I'm looking for one that will accept .gpx or .kml as the input track.

Take a look at this: http://www.scc.net/~mitty/Phyllis/Phyllis_Intro.pdf

Phyllis can be considered to be either a stand-alone program (no internet required) or a front end for Argus.  Phyllis acquires and organizes the photos from the cameras, which Argus does not.  Argus provides means to view the photo locations on Google maps and provides a distribution mechanism, which Phyllis does not.  Both support photo manipulation with visual and EXIF tagging.  Both use identical visual and EXIF syntax.  If the photo work is done within Phyllis, a complete (all photos and track files) zip file for uploading to Argus can be generated at the user's option.

Re track files, Phyllis will take GPX or NMEA syntax.  Argus will take only NMEA, so Phyllis automatically translates GPX files to that format for the upload.  Neither program takes KML files.  (At least that is the Argus status as I know it.  Troy may have plans to add other track file formats in the near future; I have not discussed it with him lately.)

I'll play with that. What's Argus?

[tinker]

What's Argus?

It's an internet-based photo handling system that's being rolled out for NCR in a couple of weeks.  I'm not sure what the general availability for use is right now.  Its strong points (my take) are:

1. Once the photos are uploaded and geotagged, the sortie path can be viewed on Google maps with pushpin symbols indentifying the camera location for each photo.  Very sexy.  Long term, even on non photo missions, being able to review sortie paths should become available and IMHO will be very useful.

2. Once the photos are tagged with target name, etc. then a user-customer with appropriate permissions can select photos to be emailed to himself, specifying the pixel size he'd like.

3. At the point of emailing, the photos are visually tagged by adding a data strip at the bottom with all the 1AF stuff.  (The data strip IMHO is superior to the "olden days" visual tagging where the added information irreversibly damaged the image itself.)  The info required to automatically remove the data strip (i.e., the pixel height of the original image) is stored in the EXIF block so the removal can be done automatically as needed.

4. All tagging information is stored in the EXIF block --- within the photo itself.  So no separate files or records, easily separated, are necessary.  (If you're not familiar with the EXIF spec, hit Wikipedia.  The germane parts are also discussed in the Phyllis documentation.)

[Robborsari]

Will this tool handle tagging and processing of 800 to 1600 images from a sortie?  We are currently processing that many using a few different methods.  Nikons capture NX2 does a wonderful job of removing haze from the photos and it does have a batch mode.  We are trying various ways to get them tagged in the time we have, mostly with success but I am always looking for ways to improve the process.

[tinker]

Will this tool handle tagging and processing of 800 to 1600 images from a sortie?  ...  always looking for ways to improve the process.

Well, the design is such that the tool can be used from a USB flash drive plugged into any computer that is running Access or the free Access runtime.  All the Phylls code lives on the flash drive; no installation is required.  The photos are stored back onto the flash drive using a directory tree structure and file naming convention based on the mission and sortie numbers.  So you can process on as many computers as you like, then just grab all the flash drives and merge the subtrees into one mission tree.  Divide and conquer, in other words.

Probably the best way to deal with the workload is to cull the duplicates and poor photos out of the herd before you get too far into processing.  Phyllis provides preview facilities that will help with that.  At each processing step you select the photos to be processed by checking boxes.  So it is pretty easy, for example, to avoid carrying three photos of the north side of the same target through all the dog work.  Just uncheck two of the three boxes before you hit "Process."  If you really have 800-1600 images from a single sortie, it seems inevitable that you will have major redundancy.  Even if you misspoke and meant "mission" there may still be a lot of redundancy.

In my experience the most time-consuming portion of the processing is the typing in of target description (name, city, state, etc.) information and lat/long (if there is no track file or in-camera geotagging).  Phyllis allows the user to do this in an Excel spreadsheet (if the host computer has Excel), so you can do cell copies and use other Excel capabilities to deal more quickly with repetitive tags across several photos/spreadsheet rows.  Phyllis also does some rudimentary checking of manual data entries too.  For instance, whether the lat/long coordinates entered by the user are in the USA or not.  This eliminates the common error of transposing lat and long.  Also, the two letter state code must be one that the post office would recognize.  Stuff like this should reduce the re-dos, oopses, and aw-s#|ts that also waste time.

You can upload the photo files and track files to Argus individually or you can zip each sortie's photos and track files together and upload the one file.  If you're using Phyllis that zip file can be generated during the visual tagging process, so you don't have to create it as a separate manual task.  Once the photos are in Argus, though, I think you have to deal with them individually for entering target descriptions, etc.

[Robborsari]

Not a typo.  Sortie.  3 of them a day.  Every day for 2 months.  Not our standard tasking for sure.  We are doing the processing with no human intervention as no one could type that much   This is not common but it is real world stuff.  In this case we dont have to put individual target names on the photos so there is no manual handling of each photo.  Even clicking on a box 1600 times is not practical.   

[tinker]

Not a typo.  Sortie.

Ack!  I guess I was assuming that was a typo.  All the tools I am aware of, including Phyllis and Argus, are basically hand shovels.  You are trying to dig the basements for a subdivision.  Agreed -- you need power tools.

Can you say what you need this extraordinary volume of images for and maybe a little more about what processing you do automatically?

[Robborsari]

The pictures are used to update maps.  The image processing is done by capture NX2 in batch mode.  And we are tagging with robogeo and a member written tool from Alabama wing called AIMM.  They are definately power tools.  I would love to find a way to automatically indicate individual targets given a list with target locations.   AIMM is very promising  but still needs some development work.  Robogeo is doing a good job of tagging and it does cover all the required elements.  One of the issues we have been dealing with is that the nikons with GPS attach filter out the heading info.  Robogeo lets us tag with heading and also allows offsetting the heading to the left or right window.