CIR and NDVI from Orthorectified Photomap

CIR and NDVI from Orthorectified Photomap

NDVI from 2s Mission NRG Orthophoto SMALL DEM_ortoPhotoCIR DEM_ortoPhoto VISHere is some resized and compressed imagery from part of a local vineyard (this area is about 200x300m). This was taken about 10 days ago, which is close to harvest time so the plants are very far from maximum vigour. Nevertheless, all three images show severe stress and actual plant loss on the East side of the vineyard (left side of the image since it has not been oriented to North. The angle of viewing is in fact from the North East and the path down the middle of the vineyard runs at about 170 deg North To South).

Some things to take note of:
– the sun was to the West and the imagery was taken at about 2pm, so the shadows run to the East of the vines. Shadows give false foliage data in the NDVI and the ‘true’ foliage is thus the line to the West (left in the image), when there are two lines.
– Being Autumn here, no foliage scores are particularly high, but the lawns in the houses on the surrounds are relatively healthy, and compare with the healthy parts of the vineyard in NDVI score.

This imagery was generated primarily using free and open source tools. Thanks to:
– Ned Horning for his Photomonitoring resources.
– Brenden Duffy for resources: video tutorials, links to software: VisualSFM, CMPVSM 6.
– Tools used: PaintShop Pro 8.0, Fiji and Photomonitoring plugin, VisualSFM, CMPVSM 6, Meshlab.

5 Comments

  1. JohnGWS

    No, I haven’t had great results from a single camera setup, either using blue or red filters. These images were generated from a dual camera setup.

    • geobduffy

      Hi John,
      Great work! Can you talk about how you rectified the two images together (NIR,R).
      Other than doubling payload on a single flight, many are concerned with being able to match the data between images. Any suggestions or thoughts you can share on this? Also, did you run a single flight with two cameras or two flights each with a single camera?
      Thanks for sharing!
      B

      • JohnGWS

        Hi Brendan
        The best way is to use two action cameras, which loses you some resolution and results in distortion but this can be dealt with by camera calibration. You can use a GoPro or a Mobius, the GoPro obviously has better resolution and so can be flown higher and map more area.
        The way I do it is to generate two textures – the visible and the NIR. Then I use bUnwarpJ or another registration module in ImageJ to register them. You may need to play around with the parameters, but *any* two images with matching features can eventually be registered by one of the ImageJ plugins.
        Once they are registered, they can be processed into the NRG (CIR) and the NDVI by Ned Horning’s PhotoMonitoring plugin in ImageJ. If you leave out the initial registration step, the registration processing in Ned’s plugin is often not robust enough to register the textures.
        The resulting NDVI and CIR will not be perfectly orthorectified any more, but for NDVI this is not a huge issue.
        Hope this helps.
        John

  2. scampos

    Hello, are you using infragram filter?you generated these images using a single camera?

    • JohnGWS

      Hi S Campos
      Sorry this reply took so long, I haven’t logged in here for a while.
      These images were generated from a two camera setup, see my reply to the next comment for how I did this. The NIR camera uses a 720nm high pass filter and the red bands from the two images are used for generating the NDVI.
      Thanks,
      John

Submit a Comment

Pin It on Pinterest

Let's stay in touch

Would you like to sign up to receive an occassional newsletter from us.  We'd love to let you know about new and wonderful things happening on the site.  Your details are never shared or sold and we hate spam as much as you do. 

You have Successfully Subscribed!