Instructables

Turn a Canon camera into a Plant Health Analyzer using Public Lab's DIY Infragram

Featured
Picture of Turn a Canon camera into a Plant Health Analyzer using Public Lab's DIY Infragram
NDVIexampleS.png
The "Infragram" project is a near-infrared imaging platform developed by the Public Lab community.   In this Instructable, we'll show you how you can use a cheap filter (available through Public Lab's recent Kickstarter) in order to modify an inexpensive point & shoot camera, turning it into a device capable of capturing capture "NDVI" imagery -- the same technology that NASA and farmers currently use to assess plant health.

This how-to focuses on a camera that is particularly easy to source and modify: the Canon A495 (the instructions will be identical for the A490).  You can buy a used version of this camera for around $50 or less on Amazon or Ebay.  

The main steps we'll need to accomplish are: a) remove the near-infrared blocking filter from in front of the camera's sensor, and b) add on a cheap "red-blocking" filter. 

Tools:

- A Canon A495 or A490
- A small phillips-head screwdriver
- A small strip of tape (most any type will do)
- A piece of Rosco #2007 filter paper (available from Public Lab right now via their Kickstarter, for $10)

Note:  if you'd rather not modify your own camera, Public Lab is also going to be producing a pre-assembled "point and shoot" camera -- you can sign up to get one by supporting their Kickstarter

Also note: this material is also explained very nicely in a Public Lab how-to video -- it'll be very useful to watch that video once before starting, and to use it as an accompanying guide for this Instructable. 

Warnings before you begin:

#1: There is a capacitor located deep inside the case of the camera.  Don't stick the screwdriver in random crevices of the camera -- you could receive a bad shock!

#2: It is very possible to ruin your camera's functionality by making a small mistake.  Don't do this with a camera that you'd really miss if something goes wrong ...

Okay, if you're ready  to modify your own Canon A495/0, let's begin!

 
Remove these adsRemove these ads by Signing Up
dad_a_monk1 year ago
This title is misleading or the author does not understand the science. You can't make a digital camera do what it's not designed to do. This is mearly a filter trick. It will only show near IR in a VERY short wavelength. With less than a micron you will only not get anything but pretty color patterens and no true spectrum to give you any information about plants. Some plants reflect IR some do not. You are not see much more than you can see with your own eyes with a pair of blue blocker glasses. This will not do much but perhaps help you pick out whats dead ONLY because of the color difference and not because of the filter giving the camera powers it's not capable of.

Like it was said in another comment, this is just another advertisment for a project to solicit donations for a project that does nothing near what it claims and hopes the public will be too uninformed to understand.
This will pick up much more than you can "see with your own eyes with a pair of blue blocker glasses". This will create a NIR (Near-Infrared) camera. I have tested it with our baby monitor that has Infrared LED's. I converted the same camera with the same filter. Check out a video I made of the test (video is pink because I didn't White Balance the video): http://youtu.be/6xx0FcBT3Uw
jywarren1 year ago
@dad_a_monk - I'm sorry you're upset about this, but I think you misunderstand. If you read about the work done at Public Lab (which is not only a non-profit, but an open community of collaborators working to develop these open source tools), you'll see that we've been building these cameras for several years, and the science is quite well established:

http://publiclab.org/tag/ndvi (the development of the technique, with lots of working tests)
http://publiclab.org/tag/superblue (the filter switch technique)

NDVI, or Normalized Difference Vegetation Index, is typically calculated from near infrared in the 750-1000 nanometer range. You can read more about the science behind the technique here: http://publiclab.org/wiki/ndvi and especially here: http://publiclab.org/wiki/ndvi-plots-ir-kit

You are right that it is a "filter trick" -- but a really useful one which makes use of all CCD and CMOS cameras' ability to detect near infrared light.

Criticism is good! But please don't bash our work just because you don't understand it :-)
neffk1 year ago
I'm not seeing the plant-health part. It's a pain to find articles so mislabeled. This should be called something like "How to remove the IR filter on a Canon A490". Instead of being a useful instructable, this is a thinly veiled advertisement. Come on, people---SEO.
chuckmerja1 year ago
as both previous answers indicated, no, this won't do thermal imaging. most CMOS imagers are sensitive out to a wavelength of about 1 micron. Heat leaking from a house or emitted from a mammal is typically in the 8 -14 micron range and takes a much more sophisticated (expensive) sensor.
Millchard1 year ago
I've always used an ice-cube tray to hold screws in the order they were removed. You then just have to note which order you actually used. On really bad devices, like laptops, I've even resorted to having my video camera mounted on the shelf above the workbench looking down at what I'm doing.
zenderon1 year ago
*****CAUTION ****** MAKE SURE YOU LOCATE THE CAPACITOR THAT FIRE'S THE FLASH***** THIS SHOULD BE DISCHARGED FIRST***** IF YOU TOUCH THESE TERMINALS YOU WILL GET A VERY NASTY SHOCK ....NORMALLY ABOUT 300 VOLTS .......SO BE WARNED.....
RangerJ1 year ago
Neat project!
Will it show heat sources, e.g.: gaps in wall insulation, poor quality windows, etc?
Unfortunately not, the camera's CCD is only sensitive to near-infrared. To view thermal sources, you need to be viewing far-infrared.
donwblair (author)  DavidKaine1 year ago
Hi Ranger J,

Great question; and DavidKaine is spot on with his comment -- this device won't allow you to do "thermal imaging". But it so happens that Public Lab has also been working on a "thermal flashlight device" intended to do the sort of thing you mentioned -- e.g. check for heat leaks in a room:

http://publiclaboratory.org/tool/thermal-photography

The design relies on the same sort of sensor that is used in "non-contact thermometers", along with some other inexpensive electronics; it's set up so that a red LED lights up when the temperature is above a user-set threshold, and a blue LED lightsup when the temp is below the threshold; the user then waves the device over objects in a room while taking a long exposure image of the process, building up a "thermal light painting" that can reveal e.g. poor insulation and leaky windows in the way you suggested ... check it out!
Couldn't you just replace the IR filter with the "infrablue" filter internally?
armstk1801 year ago
Very very awesome job
Good job !
is it possible to do this with Sony camera too?
motoguy501 year ago
What a great way to make
a very handy tool for use in the field. Saves big $$$!
over buying a real one.
Can you put the new filter inside the camera where the old one is?
donwblair (author)  cheddarevan1 year ago
Hi!

Great question! I think it's probably too risky to do that, because the filter would need to be right up against the sensor, and it's too easy to scratch / place dust on the sensor. Another reason to keep the filter external is the possibility for changing filters (in case you find another interesting filter to try).

That said, maybe there's a nicer way to attach the filter to the outside -- if you come up with a way (or any other modification), please do feel free to email the DIY camera modding folks at https://groups.google.com/forum/#!forum/publiclaboratory !

Cheers,
Don
thank you for the clear camera disassembly instructions
What a cool project! thank you so much for sharing!