10D and GPS (a serious question)

[HeatherJL]

Bob Gross posted in the other thread about the potential of GPS data being tied to the EXIF data of an image.

I work for an engineering firm and we really could use this feature. We take pictures of bridges, highways, constructions sites, etc and often need to know our exact position.

Anyone have any info?

[iwatkins]

Some blurb below that I posted to one of the magazine forums a while back:

----------------------------------------------
I've been looking into this myself for a while now and there are many options, none of which are really perfect. I still simply write in a note book my location from a map when out walking and taking landscape shots, but I really must join the 21st Century at some point :-)

Options:

1. See if you can buy a new camera:
----------------------------------

a) Ricoh have announced a camera in Japan, which for now, will probably stay as a domestic market only item. It is called the Capilo Pro G3 and allows the addition of a Compact Flash GPS card for the recording of GPS information alongside the image. Web link follows: http://www.imaging-resource.com/NEWS/1054244063.html

b) Kodak DC265 allows the connection of a GPS unit via a cable that will either record the Lat/Long onto the image itself or embed within the image data. Even so, it does sound like a tedious way to do it. Web link follows: http://www.techtv.com/freshgear/products/story/0,23008,2304680,00.html

Later EXIF standards have a section for storage of location based information but I believe this was added for the surveying/GIS community rather than consumers.

2. Use a consumer GPS unit separate from the camera:
---------------------------------------------------

a) Buy a normal Garmin (etc.) GPS unit. Every time your take a shot simply set a waypoint on the GPS unit. This is usually done by simply holding down a certain button on the unit. You can then get a list of waypoints (or where-I-shot points) and download to your computer when you return from the trip. By matching the time of the photos in your camera with the time recorded with the waypoints you can tally these up. There is even some software to automate the final process available called "GPS Photo Link" that can be seen at: http://www.geospatialexperts.com/gpsphotolink.html


3. Use a logging GPS unit separate from the camera:
--------------------------------------------------

a) Obtain a GPS Data Logger and a GPS unit. The latter only needs to be a GPS unit, it doesn't need maps or even a display. These are often called a GPS Mouse. Both items can just sit in the top of your daysack and just record away without any user intervention. Simply start the logging as you go out on your shoot. The Data Logger will record your location at a user settable frequency, say once every 5 seconds. Once you return from the shoot, simply download from the logger and tally locations and photos via their respective time stamps.

Example Loggers:
http://www.prairie.mb.ca/gpsdatalogger.htm
Example GPS Unit: http://www.royaltek.com/proditem.asp?hotitem=yes&hid=13
Example All-In-One units: http://www.laipac.com/gps_gpscube_eng.htm
http://www.laipac.com/gps_g30L_eng.htm


4. Use a logging GPS unit connected to the camera:
--------------------------------------------------

a) Use the equipment in 3 above but rig it so that the logging only occurs when you take a shot. How this is to be done I guess depends on the camera etc. being used. Ideally, you could trigger a GPS log when you press the shutter release fully down. This could be accomplished by having a sprung button that you stick on top of the shutter release button that only signals the GPS logger to log when you fully depress it at the same time as actually pressing down the camera shutter button that is below it.

Another option would be to trigger the GPS logging via a PC Sync socket (if it triggers every shot) or even some kind of hot shoe adaptor. This would need work to ensure signalling voltages etc. were correct.

Of course, a wireless method would be better, e.g. WiFI, BLueTooth etc. That way the logging equipment can be kept separate from the camera.


5. For camera manufacturers to build one for us:
----------------------------------------------

By far the best option would be an add on unit, maybe in the form of something like a vertical grip that mounts to the camera and talks to the camera. It would simply pass on location information for the camera to write into the EXIF data. This would mean that every image taken would actually contain the location information with no post processing required.

----------------------------------------------

Cheers

Ian

[robertwgross]

I was thinking that it was a certain Nikon camera that hooked up to a certain GPS receiver, but I don't remember exactly. Of course, nobody in the Canon group would have a Nikon.

I think it was intended for some professionals in some type of surveying capacity.

I use my GPS receiver independently of my camera. A year ago, I was taking one day to shoot landscapes in one of the big national parks, and I had planned five locations to shoot from. I finished the five, and then got back to location number one. Then I noticed that my wallet of 2GB of compact flash cards was gone. So, I needed to backtrack to the exact locations where I had just been photographing to try to find it. Fortunately, I had each of the five exact locations stored in the GPS receiver (within 25 feet or so). I backtracked and found the wallet at the fifth location, right by the tripod marks in the mud.

When doing nature photography, occasionally I will stumble upon some unique flower specimen. I'll shoot the shot and store the location in the GPS receiver. Upon my next return in the following year, I can go to the exact spot and look for the same flower. It would be nice to have that stored along with the EXIF data.

---Bob Gross---

[iwatkins]

Petition to Canon then Bob ? :)

Does sound as though that GPS unit has paid for itself already.

Seriously, I don't think it would be hard to link a GPS unit to the 10D. Ideally, it would be via the USB link, that way the GPS unit could "grab" the image off the CF card, change the EXIF data to include the GPS values and then write it back to the card.

Of course, the ideal method would be to build into the camera itself a GPS unit, but but that would only be possible on new models rather than retro fit to current ones.

I think by far the best method at the moment is the current one people use, i.e. buy a standard GPS unit and mark waypoints as you shoot or to data log your locations while you are out shooting.

Cheers

Ian

[robertwgross]

Well, I've used my own GPS receiver for job purposes (in another life), and I have taught lots of classes on the subject. I believe that it would not be too technically difficult to integrate standard NMEA data from a receiver to a good camera. However, my guess is that there is limited demand for such a gadget. It would be confined to photographers in an outdoor setting only, since GPS works pretty badly, if at all, indoors.

Someplace else, somebody suggested that the lat/long and time/date data should be supplemented with directional data, like a bearing of 15 degrees magnetic. That way, some professional would have proof of when and where he shot the photo, and exactly where the camera was pointed. That starts getting messy. Then somebody else thought that there should be the addition of data for up/down, relating to the horizon. Well, that starts getting very messy. Then somebody else thought that there should be sensor data for the degree of twist, comparing the lens axis to the horizon. This is too messy for me.

I think I will hold off from demanding that from Canon.

To think about it differently... If the new camera that you wanted to buy had a cost of $1500, and then there was an option for it to store something nifty with GPS data from your external GPS receiver, then would you pay extra for the camera option? If it was only $1 extra, then you probably would, even if you didn't think you needed it. If it was $1000 extra, then you probably would not unless you were damned sure that you needed it. How much would you pay? $10 ? $100 ? This is just a conceptual exercise.

---Bob Gross---

[iwatkins]

Well, if all you needed was a NMEA outputing GPS unit and the camera was supplied with the connection and the necessary cable, I would suggest that I would be willing to pay an additional $50 at most for this ability. I would probably pay $200 for it to be all built in. But it would be of value to me as I do shoot lots outdoors, mainly landscape work and I currently log my location (from maps) in a notebook as I shoot. However, I do not *need* it.

But I bet if you ask, 98% of the people with a 10D wouldn't need it. By need I mean actually need it rather than just desire it as an additional gadget.

On the flip side there are probably a small minority that would actually kill for this ability to have been built into their 10D. These are people like Heather who are shooting for work and do realy need accurate location information. I would suggest that $500 on top of the base camera wouldn't be too much for these people.

----

There are cameras that will do all this. However, they tend to hang out of the bottom of twin engined airplanes that run lanes up and down the countryside taking aerial survey shots. The GPS component is a tiny part of their overall cost.

I saw one of the camera setups on the GetMapping.com aircraft. It was a beautiful bit of kit and did indeed store very accurate differential GPS data along with the images and included heading, speed etc. data along with laser alitimeter data. Clever stuff.

Cheers

Ian

[robertwgross]

I'm not sure, but I think that the camera body would have to be GPS-enabled. By that, I mean that the body could have the GPS data happening, but it is temporarily ignored until the instant of the shutter press. Then the body says "now" and grabs the nearest bytes of NMEA data within a time window.

Yup. 98% of photographers would not care. A police investigator might.

---Bob Gross---

[mjordan]

It doesn't sound like it would be hard to be able to use a Pocket Wizard (or other wireless device) to trigger the way button on the GPS device. Transmitter on the camea and button wired into a cable that plugs into the receiver. Take a picture, fire the GPS receiver. You could even do it with multiple GPS units if you wanted to trianglate an area around the picture.

Another way would probably be a fiber optic link that would "see" the GPS unit face at the same time it took a picture so that the GPS reading would be part of the picture. Sort of like a "Head's Up Device" in a military aircraft.

Mike

[bnpndxtr]

Most people who would want a GPS receiver would want to be able to use the GPS receiver as a GPS receiver with no apologies. I mean a camera with a GPS receiver is a nice way to augment a high-end camera, but not a great way make a GPS more usable. I would argue that the better synergy is a low-end camera included a high end handled GPS. Easier to make rugged, low-power, etc.

Heck, this same argument has essentially already been validated in the cell phone market. Throw a cell phone in with a high end camera- naw. But including a low-end camera in a cell phone seems to be selling well even now. On this forum things are viewed as camera-centric, but I think the opposite emphasis is probably more sellable in this case.

[Motorsports Photo]

I think a previous poster was headed in the right direction. Adding GPS capabilities to the camera would be difficult. There are other things I would much rather see stuffed into the digic chip (like some accurate exposures!) that a GPS.

It would be nice if somebody could whip up a piece of software that after the image files and the GPS track were uploaded into a computer, the software could extract the times from image and GPS files and add the location to the EXIF.

This SOUNDS easy, but then again I dont write software for a living either.

Ahh more cool things to dream about!

-Pete

[robertwgross]

Motorsports Photo wrote:
It would be nice if somebody could whip up a piece of software that after the image files and the GPS track were uploaded into a computer, the software could extract the times from image and GPS files and add the location to the EXIF.

You mean...? The software would figure out where the camera was along the GPS track when the image time was recorded, despite whether there was a waypoint stored or not? Cute.

That means you would have to have the clock in the camera and the clock in the GPS following each other pretty closely. GPS clocks are easy to get right, since they sync to the atomic clocks in the birds, but the clock in the camera is pretty loose. Maybe we need a GPS-equipped camera just to keep the camera clock accurate.

Ha!

---Bob Gross---

[CoolToolGuy]

I read somewhere about Canon working with the law enforcement community on a way to verify digital photographs as evidence. It would seem that GPS data could be one aspect of that effort. The 'inside the building' issue previously mentioned could be a factor, since many crime scene photos come from an interior location.
Any thoughts?

Have Fun

[robertwgross]

Lat/long data doesn't have so much importance in an urban setting. First of all, most buildings have a unique street address anyway. But an apartment address on the first floor might have the same lat/long as an apartment address on the tenth floor, and GPS would struggle to distinguish between the two apartments.

No, the situation where this would be most meaningful is outdoors. For the police application, it would be good at a rural crime scene or rural traffic collision. Or a professional surveyor. Possibly a botanist finding rare plants in the jungle.

I guess they will have to carry their Canon camera around their neck and the GPS stuck in their pocket.

---Bob Gross---

[Longwatcher]

Just for trivia, GPS can provide a 3 dimensional location if you have the right chip. Lat/Lon and height above mean-sea-level. The height is not as accurate as Lat/Lon, but can nail you to a particular floor.

Having a optional GPS locator for the Professional cameras would make sense to me, but I would not build it internal to the camera, I would build so it it feed its data into the EXIF data of the image when it is attached.

Having the EXIF plug-in port could be very useful for some applications without adding significantly to the "basic" camera. I can think of several applications where an ability to feed data to the EXIF metadata fields of the image would be extremely handy when needed. 3rd parties could build the plug-in devices.

Just my opinion,

[robertwgross]

Longwatcher wrote:
Just for trivia, GPS can provide a 3 dimensional location if you have the right chip. Lat/Lon and height above mean-sea-level. The height is not as accurate as Lat/Lon, but can nail you to a particular floor.


Nearly all GPS receivers will output three dimensional position. It is an inherent function of GPS. But it takes four satellite signals to do it. In an outdoor setting, it can give you moderately good elevation, but that can be inconsistent. Unfortunately, in an urban setting, it gets poor. When you get to the "steel and glass canyons" of a city like New York, multipath interference screws up overall reception somewhat, and the elevation reading is affected the most. So, it might tell you that you are 1000 feet up and one block away. In the typical outdoor setting, it might tell you that you are 15 feet away and 25 feet up, and that would likely be good enough for camera data.

---Bob Gross---