Public Lab Research note

Results: Simple Aerial Photomapping session, Junction Avenue School READY field day

by patcoyle | December 16, 2013 04:21 16 Dec 04:21 | #9888 | #9888





What I want to do

I want to excite youngsters, like the 5th-8th graders of Junction Avenue School, participating in Rotarians Engaged and Directing Youth (READY), about science using the the Public Lab DIY mapping tools. Note the kids forming the letter "R" in the two images above.

My attempt and results

I did a session with the students, their Rotarian mentors and teacher on 12/14/13.

We did a brief overview at the school. We showed the kite and camera rig, discussed using your own personal eye in the sky, like your own satellite, to get a high resolution map of the day; that in some parts of the world, existing satellite imagery isn't very good and how the imagery and maps might be used for monitoring land clearing, environment issues, Engineers Without Borders projects. We looked at prior maps and a blank Mapknitter map of the nearby site at Robertson Park.

After a short drive to Robertson Park, we filled a Public Map Balloon Mapping Kit balloon, started the CHDK intervalometer scripts on our SX260HS (visible) and A490 with Rosco 2007 (infrablue), got our dual juice-bottle rig attached, and launched our balloon. I did a short safety talk and then we started mapping. Everyone who wanted to fly the rig got a chance to, which was a big hit. We captured over 140 infrablue images and 700 visible images.

We drove back to the school and looked at our images and had a good discussion with lots of questions. I also brought an ARParrot and R-10 quad, since one of the students is a skilled RC aircraft pilot. We discussed the pros and cons of kites, balloons and UAVs for mapping. We didn't have an internet connection, so I'll be back with the processed maps and photos from the flight.

Microsoft ICE was used to stitch selected infrablue images which were then processed to NDVI with Ned Horning's ImageJ/Fiji macro.

In addition, Jack Toeppen graciously processed the images using PhotoScan Standard on his more capable machine and I took a screenshot (standard edition doesn't do true othomosaics) from his 3D model and processed it into the following images. Subsequent discussions with Chris Fastie indicate the white balance on the images is such that the vegetation is not adequately differentiated in NDVI. Also the white background is skewing the results from Ned Horning's software, resulting in the vivid pink-hued image.




Visible map

NDVI map (Note, this map requires one to zoom out and maneuver to find the map, both in link above and embedded web viewer. Sorry about that.)

Images: Ground-based photos courtesy of Don Wentz, others from balloon mounted cameras

Questions and next steps

Return to discuss results with students. Follow up to offer other mapping sessions. Consider sessions with other Public Lab tools like the foldable spectrometers.


Login to comment.