This is a testing site only. See the live Public Lab site here »

Multispectral imaging

The Public Lab near infrared imaging project is an open source community effort to modify consumer cameras to capture near infrared imagery for a range of purposes, including plant health. All open or accessible near infrared imaging hardware and software efforts are welcome here! **Join in by:** * Reading about goals and asking great questions * Converting a camera using one of our starter kits * Trying (and critiquing) our community-made how-to guides * Posting your own how-to guides and mods * Building on others’ work; hack and remix the kits to refine and expand them * (proposed?) Submit your improvements for inclusion in an upcoming starter kit release or add-on * (proposed?) Serve on a Research Review Group for a 3 month period ## Activities This is a list of community-generated guides for specific applications using your near-infrared imaging setup (either a camera you converted yourself with a filter pack, a ready-made near-infrared camera, or double camera setup). Some may be more reproduced -- or reproducible -- than others. Try them out to build your skills, and help improve them by leaving comments. Together, we can repeat and refine the activities into experiments. [activities:multispectral-imaging] Add your guide here Request a guide _Guides should include a materials list and a step-by-step construction guide with photo documentation. See an example._ **** ## Hardware Mods Have you added to your starter kit, improved it, or redesigned it? Show others how to take it to the next level by posting a build guide here: [upgrades:multispectral-imaging] Add your hardware modification here Request a hardware modification _Upgrades should include a parts list and a step-by-step construction guide with photo documentation. See an example._ **** ## Builds There’s a lot going on in open source near-infrared imaging -- if you’ve developed another open source design you’d like to show others how to construct, post it here! * Rasberry Pi NoIR * X * XX **** ## Choosing a tool / Starter Kits The question to start with is whether you can capture all the channels you need for your research question with a single converted camera or whether you should use a dual camera rig with one converted camera and one unconverted. That choice plays out in terms of what filter (blue or red) to use for converting your camera. Public Lab’s Kits initiative offers several starter kits, one with the basic components and instructions for converting your own digital camera to capture near-infrared imagery and a second option -- a readymade lightweight near-infrared camera. The point of the kits is to lower the barrier to capturing your own near-infrared imagery. * **The Infragram filter pack** is our least expensive way to get started with near-infrared imaging, but it does require an afternoon and some technical work to perform the camera conversion yourself. ... etc etc etc **[Visit the Infragram DIY filter pack]()** * **The Infragram point and shoot** is a Mobius camera that we worked with a factory to modify. ... etc etc etc **[Visit the Infragram Point&Shoot page](/wiki/infragram-point-shoot)** **** ## Processing near-infrared imagery Once you take a multispectral photograph with a modified camera, you must post-process it, compositing the infrared and visible data to generate a new image which (if it works) displays healthy, photosynthetically active areas as bright regions. In-depth articles on the technique by Chris Fastie can be found here: * * **** ## Software How to process your images: we're working on an easy process to generate composite, infrared + visible images that will reveal new details of plant health and photosynthesis. There are several approaches: * The **easiest way** is to process your images online at the free, open source []( * [Ned Horning's](/profile/nedhorning) [PhotoMonitoring plugin](/wiki/photo-monitoring-plugin) * Manual processing * [in Photoshop](/notes/warren/10-25-2011/video-tutorial-creating-infrared-composites-aerial-wetlands-imagery) * [or GIMP](/notes/warren/10-27-2011/video-tutorial-creating-false-color-ndvi-aerial-wetlands-imagery) * Command-line processing of single images and rendering of movies using a Python script: Source code is [here]( and here: [here]( * For those who use the webcam and have a Python interpreter, there are some image processing codes available at [Python Webcam Codes](/wiki/python-webcam-codes) * Using (deprecated) **** ## Comparison to standard tools Infrared imagery for agricultural and ecological assessment is usually captured from satellites and planes, and the information is used mainly by large farms, vineyards, and academic research projects. For example, see this illustrated [PDF, page 210]( from a commercial imagery provider who has been studying the usefulness of infrared imagery and has quotes from farmers who make use of it. There are public sources of infrared photography for the US available through the Department of Agriculture -- [NAIP]( and [Vegscape]( -- but this imagery is not collected when, as often, or at useable scale for individuals who are managing small plots. **** ## Frequently Asked Questions Ask a question about multispectral-imaging Get notified of new questions and help out [notes:question:multispectral-imaging] ...

Author Comment Last activity Moderation
warren "ah, good point. I guess both are relevant. We could distribute them in a dark envelope or folder, perhaps.  Yes, i also wondered about this -- bot..." | Read more » over 4 years ago
nedhorning "Hi Jeff - I wasn't sure if you mean varying over time or over the surface. Over time both will vary but from my experience they seem fairly stable ..." | Read more » over 4 years ago
warren "Hi, @nedhorning! I wanted to ask if you thought tar paper that people purchase in a store would have a consistent enough reflectance that you think..." | Read more » over 4 years ago
warren "I posted an activity for doing NDVI-like work using a camera like this -- with a coefficient (on the output of NDVI, so like "boosted" NDVI) here: ..." | Read more » almost 5 years ago
cheneyshreve ":smiley: " | Read more » almost 5 years ago
warren "@warren awards a barnstar to Claytonb for their awesome contribution! " | Read more » almost 5 years ago
warren "Just a note to self - as we plan out auto-registration features for this, could be good to look at #image-registration ! " | Read more » almost 5 years ago
warren "Great idea @amirberagain. @maggpi if you have a chance to record video of a live insect's face that would be so cool... :ant: " | Read more » almost 5 years ago
amirberAgain "This is interesting, did you try to re-focus for the NIR LEDs? It's very likely that the lens wasn't designed for these wavelengths and re-focusing..." | Read more » almost 5 years ago
Bronwen "Wow, this is amazing! I was wondering what this might look like... I'm excited to try to replicate this one! " | Read more » almost 5 years ago
warren ":grinning: " | Read more » almost 5 years ago
warren "This is so cool!!!! " | Read more » almost 5 years ago
warren "Cool! We've tested blue filters with Raspberry Pi cameras, and it works well although the separate camera technique does result in cleaner separati..." | Read more » almost 5 years ago
xose "During the leptos development we bought this board. We'are now playing with a point and shoot NDVI with raspi. I'll give the "IVMECH" board a chanc..." | Read more » almost 5 years ago
tech4gt "This is so cool @warren " | Read more » about 5 years ago
warren "@imvec @xose too! " | Read more » about 5 years ago
warren "@tech4gt @ccpandhare check it out! :-) " | Read more » about 5 years ago
tech4gt "@jywarren this is awesome!!! :D " | Read more » about 5 years ago
warren "Probably! Can you point me at a MODIS source with bands youre interested in? " | Read more » about 5 years ago
eustatic "hunh, can you do this with MODIS? " | Read more » about 5 years ago
ccpandhare "@warren wow this looks very promising!! :-D " | Read more » about 5 years ago
jbreen "@warren I think you may have stumped me. Have you looked at GloVis? " | Read more » about 5 years ago
warren "@wward1400 @adam-griffith @MicheleTobias @lprashad @eustatic @nedhorning @cfastie you might like this little experiment too! " | Read more » about 5 years ago
warren "@ccpandhare @tech4gt exciting! @jbreen - you'd know this... are there any tiles available for LANDSAT's near-infrared bands? Especially like NGB t..." | Read more » about 5 years ago