Want to talk about soil? Join us for Open Call every Tuesday until Mar. 30 and our kick-off call on this topic on Feb. 2! Click here for details on how to join!

Public Lab Research note

NDVI with a key chain camera to use with a mini-drone - First steps

by azaelb | April 11, 2015 22:46 11 Apr 22:46 | #11754 | #11754

I am a professor of environmental physics working in Panama, Central America, with Florida State University program there. I understand that Vegetation Index and precision agriculture studies can be quite expensive and out of reach for poor farmers here. After seeing use of drones in these types of studies and other sites using Public Lab Infragram project, I thought why not helping these farmers.

I had a Syma X5C-1 mini drone for quite a while which only comes with a 720p custom video camera, so it is essentially a flying camera. Recently I modified a 1080p spy key chain camera to use it with the drone to get better videos, and thought if it will be possible for it to carry a similar 1080p camera modified for NDVI analysis. This X5C mini drone only weights 110g and can carry up to 50g of payload to a decent height to scan a soccer field. The 1080p camera is about 15g once removed from the original metal case.

I got two 1080p key chain cameras from a mail order company for about $30 ea. These pictures show the camera. They are sold as "night-vision" capable, so their filter does not block completely the infrared. They use microSD cards and come with a LiPo battery that lasts about 1 hour. keychain_camera_orig.jpg

I took the camera out of the metal case -too heavy for the mini drone- and removed the IR filter that comes with it, which turned out to be very thin compared to filters in photo cameras like the ones suggested for NDVI conversion herein. Below there is a sequence of this process. keychain_camera_circuit.jpg

Here is the camera circuit with its lens and battery taken out of the original metal case. keychain_camera_lens.jpg

To remove the lens, some heat was applied with a pencil soldering iron to remove the glue holding the lens, but later had to use the same iron to remove the plastic keeping the filter. What is left is a square indentation were the IR filter was.

The camera circuit was later encased in a old micro-cassette plastic case and sealed to prevent light and dust to get inside, leaving holes for the lens, the microSD card, the power/microUSB connector and the ON/OFF/VIDEO switch. Here is a picture of the two cameras, left one -VIS- with its IR filter to use as reference, and the converted one without the IR filter, tagged 'IR'. Both will be mounted under the mini-drone body. VIS-IR-camera-pair.jpg

Here is a picture frame of the video from the converted camera taking a short clip of a friend´s front yard. IRphoto.jpg

A DIY Infragram filter kit with the red and blue filters has been ordered. Once it arrives, I plan to put a filter holder in front of the converted camera and tests both filters and uploading the video frames to the Infragram.org site for processing. I will show the results here once I got the filters and do the processing.

One inconvenience for this cameras is that I do not have any idea on how to white-balance these type of cameras. No technical data is available. Perhaps another community member has done something about it and will like to share it. It will be most appreciated.

  1. Barrera (azaelb)



Reply to this comment...

Login to comment.