Photogrammetry use case, now what? :)

Hi!

Darn you guys... so, you got me nuts with KAP AND photogrammetry... =)

Last summer I had this test project with a local golf club. I thought that it would make both visually interesting and useful to have a 3D model of the course lanes, instead of a static map. I know there's other approaches, like the visually stunning skyx.fi 360x180 degree virtual tours, but I think after the visual WOW effect, that doesn't have any real value for the golfer.

So, there I went to fly a kite, on the golf club. Took some photos, and made these demo versions, which are embedded in a client's web page. Sorry it might take some time for the model to load completely, I didn't optimize the textures in any way yet. Works with WebGL enabled browsers (most recent versions of each brand, practically) ("Yleiskuva" == general view, "Tiiausalue" == teeing ground)

Course model with vector overlay
Course model with texture painted overlay
Another larger course

I left the free rotating possibility out, so that the user isn't confused, if not accustomed to rotating 3D objects with a mouse. So use the buttons over the model to change views.

I also marketed a possibility to produce "ortho-rectified" maps from these models, for print purposes.

image

The club CEO was interested and happy with the results, but still I couldn't get the discussion to proceed to actually making some deal with him. He was interested in helping to develop this further into something bigger, and had some ideas, but unfortunately I don't have the resources for it, since I already have a day job. And I didn't quite feel that much "pull" or "demand" at this point, that I would immediately quit my job and start a business. I tried to get some of my local business oriented friends together, and while also interested, they already had their own thing going on. So now this thing is just lingering as an idea and demo, and I don't quite know what should I do.

So, I'm asking you fellow KAPers, do you see potential in this kind of activity? What should I do, try to market this to some other golf club? Try to develop this into a platform? Or just go and fly a kite. ;)

Comments

  • edited May 2015
    KAP should be a low cost alternative to other aerial platforms for this kind of work. The startup costs for commercial survey by other aerial methods is very steep in the UK. Hobbyist quads can't get the reso we can. ..and stay legal. In practice I have found reliability a problem, getting a predictable fly date is impossible and clients get impatient.

    KAP orthophotos are great:

    06022015_orthophoto"

    And topomapping works well too:

    image

    ...getting it to pay is another matter- I'm up against archaeologists who do this for free. Maybe golfers have a more realistic outlook on the value of these things?


    B
  • Yes, that predictable fly date is one of the biggest issues I've encountered too. It's a short notice or 'come as you please' type of arrangement at best. On the other hand, lighting conditions apply for drones as well. In practice, I got the best photos for 3D model in slight overcast, when it's bright, but still no heavy contrasting shadows, so that was the day I was aiming for. Fortunately the wind was co-operative as well. :D
  • edited May 2015
    I fly on the old golf course of Minchinhampton Golf Club, an important site of the Dobunni Tribe in Gloucestershire. Here is an aerial shot of the course taken by Hamish Fenton:
    https://www.facebook.com/139846896185869/photos/a.139847746185784.1073741827.139846896185869/184627315041160/?type=1
    I intend flying there over the next year or so. The area falls within the remit of the National National Trust for which my daughter Heidi does KAP. Having flown there recently, I was invited to the AGM of the Club at the main course, listened to talks and introduced myself. My interest is purely archaeological. Here is a near infra-red shot taken with an auxilliary keychain camera.
    image
    There were two primary concerns raised at the meeting. One related to the maintenance of bunkers, a veritable science these days, and the other related to the health of the grass. Dry periods and disease can affect the health of grass. The use of fungicides is heavily regulated and used only when necessary.
    Having since thought about the use of KAP, one possible application is in identifying areas of the course where the health of the grass is being compromised, well before any visible signs are apparent. Remedial action can then be taken before the damage takes a foothold. Near infra-red KAP is the way to go. Unhealthy grass, or areas of excessive moisture, appear darker than elsewhere.
    image
    Here are some near IR aerial shots of Bathgate Golf Course in West Lothian, taken by Jim Knowles and myself:
    http://www.armadale.org.uk/bathgatecastle.htm
    At Bathgate, the problem can be too much water!

    Such health monitoring would involve regular flights and would be best managed commercially under contract.
  • I copied this link across to:
    https://www.facebook.com/pages/Kite-Aerial-Remote-Sensing/139846896185869
    where there is a comment from Mark Fonstad.
  • Thanks John! I think Mark's comment is very valid to take into consideration.
  • you might consider looking at single-camera or dual camera NDVI setups if you're getting into plant monitoring.
    Ned Hornings' ImageJ plugin is good for analysis. Dorn Cox in the public lab community is doing field trials of different grasses using single-camera NDVI, he'd know more too.
    http://publiclab.org/wiki/near-infrared-camera
  • @mathewL: That plant monitoring might be something special too that would attract interest. Though, I'm so bad at tinkering with camera internals, that I'd need ready converted NIR camera. :D
  • If you decide to leap into this, see what you can do with a single camera setup. Scott Armitage warned me ages ago that really accurate registration of two images is tough at best. I've been playing with a two camera setup, and registration of the images is tough, just as Scott warned. Issues I've run into are slight differences in plate scale caused from differences in filter thickness and using optics at wavelengths they're not designed for, and issues with focal plane tilt because I didn't have a setup for dialing the focal plane of the converted camera back in. I'm going to address that second one in the next couple of weeks, but I don't have a fix for the first one.

    That being said, when it works it works GREAT. I'm tempted to pick up another one of these cameras to do a single camera conversion on, just so I can play with that, too.

    On to other use cases!

    I've done some KAP archaeology, but most of the time I'm doing scientific data collection with a kite I'm working over a garden. Ortho imagery is good for area analysis, planning for future work, planning for irrigation needs, etc. NDVI is interesting, but isn't of immediate utility because the gardeners themselves typically know which plants are healthy and which aren't. All of the garden work I've done has been for school gardens, so it's all been donated work. I don't know if you could make a business case out of this.

    Farms might be another case altogether, but I can say first-hand that using KAP as your primary data gathering method for even a moderate farm is a lot of work. It's a lot of walking, and if you don't have permission to walk around on adjacent properties it's tough to position a camera over every square inch of a farm.

    I tried to work with a local architect to provide high resolution DEMs of a site prior to construction, but he pointed out that nine times out of ten a site is bulldozed to a particular profile that may have nothing to do with how the site was beforehand. Same goes for landscaping.

    If you don't mind travel, I think there's a strong case to be made for being part of a relief effort team. Natural disasters can take out bridges, roadbeds, culverts, etc. A good set of photos and photogrammetry can be a real benefit for the civil engineers who are planning emergency repairs.

    Tom
Sign In or Register to comment.