logitech c310 conversion

[nicorem]

i just converted my logitech c310 (#74 filter)
I am not convinced by the result after testting on infragram.org
i attached one picture
some body can tell me if it works or not ?

thanks

[nicorem]

i attach others pictures after infragram app process
as you will see there is a UFO in the sky, a green UFO

[Jeffrey Warren]

Hello Nicorem - the last 3 images look better than the others. Try taking a photo of a lawn, and also a regular photo for comparison. 

I think your UFO is just the sun -- probably it's so bright it's overpowering the sensors entirely in all channels, so that green patch is just close to a 0 NDVI value.

Thanks for sharing!

--
You received this message because you are subscribed to the Google Groups "plots-infrared" group.
To unsubscribe from this group and stop receiving emails from it, send an email to plots-infrare...@googlegroups.com.
To post to this group, send email to plots-i...@googlegroups.com.
Visit this group at http://groups.google.com/group/plots-infrared.

[Nicolas GRAVELLE]

hello

my house and the lawn ,
i have doubts, the wall has a ndvi index upper 0

thanks for your comments

[Chris Fastie]

Hi Nicolas,

It looks like the color balance system in that webcam is producing photos with barely enough information to make good NDVI images. Using Ned's plugin for Fiji I stretched the blue and NIR histograms (parameter=30) and then added 40 to the NDVI value in each pixel:

That generally put the plant pixels in the 0.1 to 0.7 range and the non-plants in the range below 0.2. So that is a reasonable NDVI image.  I don't know how to replicate this at Infragram.org or the Infragram sandbox. Maybe somebody can tell us if that is possible.  The histogram stretching trick is really useful even with Powershot Infragram cameras, and may be critical for webcam Infragram cameras. There is probably a way to simulate a histogram stretch in the sandbox with an algebraic statement, but maybe an explicit function would be worthwhile.

[Nicolas GRAVELLE]

hi Chris

thank you for the time you spent looking for a solution
my project is to place a camera in a rapeseed field
in fact the camera is already in the field
here is a picture, it is also a logitech c 310
I use a raspberry pi to take a picture a day and send it on the internet.
I wanted to measure the NDVI
but I'm pretty disappointed with the quality of the images and the result for this type of camera.
I also use the camera of the Pi (Pi board camera) to take another picture (a trap for insects in the same field)
so i can't use two pi cam on the same pi,

[Jeffrey Warren]

Perhaps you can use the webcam for your trap photos -- not every camera works very well for infragram photography, but we know the raspberry pi cam does, once its filter is removed (hard) or if you get one of the new NoIR cameras: http://publiclab.org/notes/warren/10-29-2013/raspberry-pi-infrared-camera-ships-with-public-lab-s-infragram-filter

--

[nicorem]

Yes, I just ordered the NoIR on Radiospare
But I also need a good resolution to identify insects
Anyway , I take to have experts on the subject
In fact I'm looking for what  can bring me the spectral camera on crop management .
In my farm, I am already using remote sensing tools to manage particular crops to modulate nitrogen inputs , using satellite photos.
But what can i learn from a photo that is not an aerial photography ?.
I searched a lot on internet and from colleagues agronomists
But the relationship between an index ndvi and prescription practice field seems very complex.

I would like to establish a relationship between the spectral image and crop growth.
Do you think it is possible with these images ( none aerial) to make an estimate of the height of the plant?
How measuring plants growth, plants height by comparing pictures taken at the same place evrey day. (i’m a programmer, i can programm it, but i need to know what data i have to analyze !)

In fact I need to measure the height of crops, each day,  which would allow me to adjust the spreading of fungicides including the regulator, whose purpose is to limit the growth of plants.
Thank you for your opinion on the subject , hoping that my english is understandable

Nicolas

[Ned Horning]

Hi Nicolas,

I agree with your observation that the relationship between NDVI and prescription practice is complex. For the foreseeable future I don't see much value in using NDVI on the average farm. I expect relying on your mind will help more than relying on an index. I'm not saying it can't be done but that it's probably not worth it. I see tremendous value in visually interpreting aerial photos using an unmodified or even a camera with a NIR filter but creating meaningful products from those images is tricky and I have yet to imagine the benefits of doing so.

Why do you want to create a relationship between spectral information and crop growth? There are a few ways to get height information using imagery but I do not think using spectral information is one of the better approaches. One indirect use of the spectral information is to relate the crop canopy texture to height. As a canopy grows it tends to cast more shadows and those can be "measured" by running texture algorithms on the image. Another very direct way to do it is to use a camera to photograph poles in the field with measurement increments on them (similar to� a stream gauge) and then take a picture of the pole to see how the crops are growing. If you need a lot of sample points that might not be practical. A more interesting approach from my perspective is to use an aerial stereo camera system and use that to measure crop height. This is the same technique people have used to measure tree heights for nearly 100 years. There is open source software out there that should allow you to automate the height retrieval. It would take a good bit of work to get a working prototype but (assuming that it is precise enough) then it would be easy for anyone to replicate the process - unlike any spectral index or measure. Some other folks on this list might promote using a sensor netowork to measure light intensity going through the plant canopy and relate that to crop height.

I'm not sure if this helps but hopefully it gives you more to think about. If you're interested in pursuing any of these ideas I'm happy to provide feedback.

All the best,

Ned

On 11/05/2013 12:43 PM, nicorem wrote:

Yes, I just ordered the NoIR on Radiospare

But I also need a good resolution to identify insects
Anyway , I take to have experts on the subject

In fact I'm looking for what �can bring me the spectral camera on crop management .

In my farm, I am already using remote sensing tools to manage particular crops to modulate nitrogen inputs , using satellite photos.
But what can i learn from a photo that is not an aerial photography ?.
I searched a lot on internet and from colleagues agronomists
But the relationship between an index ndvi and prescription practice field seems very complex.

I would like to establish a relationship between the spectral image and crop growth.
Do you think it is possible with these images ( none aerial) to make an estimate of the height of the plant?

How measuring plants growth, plants height by comparing pictures taken at the same place evrey day. (i�m a programmer, i can programm it, but i need to know what data i have to analyze !)

In fact I need to measure the height of crops, each day, �which would allow me to adjust the spreading of fungicides including the regulator, whose purpose is to limit the growth of plants.

Thank you for your opinion on the subject , hoping that my english is understandable

Nicolas

[nicorem]

thank you ned, many ways to explore

I also thought from a photo taken every day, calculate the number of pixels with NDVI greater than 0,
and see how this number changes
but I'm not sure we can establish a relation between this number and plant growth

can we say that if there is more pixel with NDVI greater than 0 on a picture,  then there is more vegetation in this picture ?
why pictures and why spectral information, ? NDVI  should be more representative than calculate pixel green color on the picture.

Analyze one picture per day is easier than flying a drone each day
and  what is interesting is to see, to compare changes every day

I also thought of an ultra sound sensor but it measure only one point and it is not really representative.

[Ned Horning]

Nicolas,

If you count the number of pixels in an image with an NDVI value greater than a threshold you will get an estimate of how much of your area is covered with plants. The NDVI threshold will almost certainly be higher than 0 but it depends on the equipment and methods you are using.

If you used a stereo camera setup you could still take one shot per day without needing a drone. You would just need two cameras mounted a few inches appart or a system that moves one camera a few inches so the two photos can be taken from slightly different positions.

A sonar solution probably wouldn't work well since the viewing area (footprint) of a sonar is quite wide and a plant canopy is a very irregular surface. If you want to use an active sensor you could consider lidar. A scanning lidar system is quite expensive even for something very simple. There are some very interesting DIY 3D scanner designs on the Internet but most of these work best indoors or at night since sunlight tends to interfere with the laser signal. I"m not certain how well these work work over a plant canopy but it would be worth testing. Maybe you could build a scanning laser system to use at night? That would be fun and quite useful.

Ned

[Jeffrey Warren]

It would take some work to line up the 3d models, but Chris and Ned have done a good bit of 3d mapping of crop growth at the iFarm event in New Hampshire, on Dorn Cox's farm: http://publiclab.org/notes/cfastie/6-3-2012/nrg-images-cover-crop-trials

[fur...@gmail.com]

Hey... I removed the IR filter on the C310 but when I put it back I get a black screen. How did you get IR data like that?
I uninstalled the drivers + software and let windows install the default drivers. I also played around with the white balance and exposure settings.... but I can't get anything other than black.

[bubai.sup...@gmail.com]

Hi there, its the first time I am trying to remove the IR filter. I followed several DIYs on the internet to remove it but actually I am not able to locate the IR filter on the C310 webcam. Can you please exactly tell me where the IR filter is is located in the webcam? Further if you have images showing the location of the filter in the C310 webcam, please post them along.
Any help would be gracefully received.

[jom.e...@gmail.com]

How do you convert your c310 to an infrared camera
How do you remove the IR filter ?

[Jeffrey Warren]

Hi! There are a lot of conversion activities listed on this page -- hope this helps! Many cameras are relatively similar:

https://publiclab.org/wiki/infragram-convertible-cameras

On Mon, Oct 22, 2018 at 10:34 AM <jom.e...@gmail.com> wrote:

How do you convert your c310 to an infrared camera
How do you remove the IR filter ?

--
Post to this group at plots-i...@googlegroups.com

Public Lab mailing lists (http://publiclab.org/lists) are great for discussion, but to get attribution, open source your work, and make it easy for others to find and cite your contributions, please publish your work at http://publiclab.org
---

[Jeffrey Warren]

Hi! It can be risky with some cameras, but is more likely to result in a scratched lens (or sometimes even dusty sensor) than a completely broken camera. But if you do it carefully, slowly, in a low-dust environment, many cameras are convertible! I'm not familiar with the C310, however. Can you share a photo of the camera? 

Best,

Jeff

On Mon, Oct 22, 2018 at 10:45 AM Jomar Cristian Erandio <jom.e...@gmail.com> wrote:

Thank you sir, I'm just not sure whether I will remove the IR Filter fix to it sensor, I have a C310 camera, will it make my camera dead?

[Jeffrey Warren]

Responding to Jomar: Hi! It can be risky with some cameras, but is more likely to result in a scratched lens (or sometimes even dusty sensor) than a completely broken camera. But if you do it carefully, slowly, in a low-dust environment, many cameras are convertible! I'm not familiar with the C310, however. Can you share a photo of the camera? 

Best,

Jeff