Public Lab Research note


Wratten 25 and NDVI A1200 imagery from TerrAvion fly-along

by patcoyle | August 08, 2014 23:08 08 Aug 23:08 | #11024 | #11024

What I want to do

Share initial results of Wratten 25 and resulting NDVI A1200 imagery from TerrAvion fly-along.

My attempt and results

Stewart Long and I met with team at TerrAvion. A startup that provides weekly aerial imagery for agriculture, including NDVI, color infrared, and thermal imagery with next day processing/delivery on a subscription basis.

Robert Morris and Cornell Wright, principals, were very familiar with PublicLab and were willing to discuss their approach and workflow and try to fly our cameras on one of their runs.

They added our Canon SX260HS (visible) and A1200 (Wratten 25), both running CHDK intervalometer script. (To Chris Fastie's comment, below, they were just running in auto mode, unless by chance, I'd managed to set the 1/800 speed Chris noted. The visible images were overexposed and blurry, suggesting too slow a shutter speed.)

Maps: I selected this area from the images along the flight path, because of readily discernible features and a good variety of crops and fallow land. The area spans about eight miles, longest map I've done in Mapknitter. There are over 700 images, so I may look for additional areas of interest to map.

I find you need to zoom out like six click to get the image in the map viewer.

Wratten 25:

NDVI processed with Ned Horning's PhotoMonitoring Fiji tool:

Questions and next steps

I want to explore opportunities to collaborate with Terravion for pro-bono imagery for publications.

Why I'm interested

I'm delighted with their service and receptiveness to Public Lab


2 Comments

Great application. I assume your camera but not you got to go in the plane? How was the camera triggered? Was it mounted on the plane? What was the elevation?

It looks like you locked shutter speed at 1/800 second at ISO 80 and the shots were taken with the aperture all the way open (good choice).

I looked at the eight Wratten 25 photos at Mapknitter. Each one is blurrier on the right side than on the left side. Is this the case with all photos taken by this camera? If so, it might be that the sensor is out of alignment after the conversion. Adjusting those three screws that hold the sensor plate could fix that. I just converted an A1200. It was so easy to open and close that I did it 10 times, each time making fine adjustments to one or more of those final screws. Between adjustments I took a few photos of a distant scene and checked the focus. My last adjustments were tiny turns of one screw. I did not replace the four screws on the outside of the camera each time, just snapped the back cover on between adjustments. So there are only two screws (that hold the LCD) to remove in order to access the sensor screws (and no ribbon cables to unplug). It takes only three minutes to get in and out. The final screw positions were not screwed in quite as far as when they were still glued (about 1/5 of a turn). That could vary depending on what filter replaced the IR block filter.

Is this a question? Click here to post it to the Questions page.

Reply to this comment...


Thanks Chris. As always I appreciate your tips. I'll have to take a look at these points.

Reply to this comment...


Login to comment.