Hello,
I want to make infragram camera using i-ball web cam to measure NDVI INDEX.So,I had purchase 20-Mega pixel i-ball usb web camera.To replicate infragram camera,i had replace the IR filter with rosco #19 fire filter(Red) as per suggestion of public-lab 's below link
http://store.publiclab.org/products/infragram-diy-filter-pack
To test this camera result,I had capture images in presence or absence of sunlight. 1) In presence of sunlight,I get below attached image(Outdoor_image.jpg).
2) In absence of sunlight( or low sun light ),I get below attached image(indoor_image.jpg).
I was getting totally different images in absence or present of sun light.
So,requested to address my below quire
1) what are the reasons for getting these kind of image difference in presence or absent of sunlight ?
Any help in this case is appreciated.
Please let me know if you need more information form my side.
thanks, Parth
2 Comments
Hi Parth,
The primary difference between your two photos is exposure. The bright one is overexposed. If there is software that allows you to control the exposure while you take photos you should be able to get better photos.
To get meaningful NDVI images directly from your photos the camera will have to allow custom white balancing. Without an artificial adjustment to the relative brightness of the blue and red channels in each photo, NDVI computed from the photos will not be useful. Most webcams do not allow custom white balance.
Another possibility is to adjust the values in the blue and red channels after the photos are taken. To know how to adjust the values, targets of known reflectance in the red and NIR can be placed in each photo and the procedure outlined here can be followed: https://publiclab.org/notes/nedhorning/07-22-2015/introducing-the-calibration-plugin-for-imagej-fiji
Chris
Reply to this comment...
Log in to comment
Hi Parth,
I agree with Chris, also it seems to me to see that the camera has a problem focusing. If you do not get images in focus, it could be that their thick filter Rosco, is significantly different from the original IR.
Reply to this comment...
Log in to comment
Login to comment.