KAP for intertidal landscape mapping

edited May 2014 in General
A park ranger friend alerted me to this scientific paper presented at a conference he was attending.
http://www.plosone.org/article/info:doi/10.1371/journal.pone.0073550
Using a delta coyne for high quality multi spectral mapping of intertidal landscapes.

Comments

  • edited May 2014
    Thanks for the link Andrew,

    Excellent abstract!

    Great aplication for KAP

    Fig 11 shows just how good KAP is at getting detail resolution imagery of habitat.

    Fig 6 shows a common failing of the KAP photomap: incomplete cover. I have the same greif: no matter how much time you put into a swath the kite just won't let you get a clean sweep!
  • That's one of the best discussions I've seen on the application of NDVI imagery for analyzing the landscape. Up 'till now I've seen it treated more as a "less, more, more, most" kind of thing. It was cool that they were able to say where in the NDVI scale different plants showed up. That's AWESOME! I seriously need to play with this. Very cleanly written article. Thanks for posting!

    Tom
  • Mitch has other papers in the pipeline and I have added another to the link above.
  • Thanks for this! his NDVI calibration procedure is much appreciated.
  • FCB
    edited May 2014
    I think the simplicity of the biological system is a real advantage in this project in making NDVI interpretation and calibration look easy. In most terrestrial systems, comprehensive classification based on NDVI alone (or any other simple ratio) doesn't have a chance because you have so much variability in terms of species diversity and plant nutrient/water/disease status.

    These guys are actually pretty close to comprehensive classification of the main cover types if they used the available visible light channels. The three algae are brown, olive brown, and red (see reflectance differences in Figure 3). You just have to have the camera close enough to the ground (yeah! KAP!) and do some multivariate work by hand or with remote sensing software like ENVI to create derived parameters for classification. (It's probably the next publication).

    Very cool study, and a lot of hard work involved. I like to see projects that are well thought-out before the field work.

  • Hi Guys,

    FCB: Yes as you were saying the next work we are trying to do is look at automatic classification which requires all the available bands (not just derived NDVI), and what we are finding is that the real advantage of KAP here is the spatial resolution allows you to classify things (automatically) based on spatial and textural cues along with the multispectral reflectance spectra.

    Cheers for the comments, I've just joined up for the "notes" forum; lots of interesting reading.
  • Thanks for the work Mitch (et al). I look forward to the texture/spatial work. KAP has already moved past my initial hopes when I got started 8 years ago.
Sign In or Register to comment.

Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!

In this Discussion