LEDs and NDVI

[Colin Taylor]

Hello again

What effect if any would IR/UV/other leds have on the NDVI results.

For example; use an (IR?) Led to shine on foliage in cameras field of view. The foliage pictured then recieves the same light as all other pictures.

Would it provide an absolute value considering all pictures recieve same light. Would light need to be brighter than ambient to serve that purpose?

Main question is what leds to use? Can get very bright LEDs of a wide range of wavelenghts

Using a handheld MAPiR with Blue filter

Thanks in advance

C

[Chris Fastie]

Is the plan to do this in a darkened room, so that only light from the LEDs illuminates the plants? If so you can learn a lot about how healthy foliage reflects most NIR but absorbs most visible light. The results will not necessarily lead directly to NDVI which assumes that the light illuminating the plants has the proportion of NIR:VIS in sunlight. If you get the proportion close to that of sunlight, computing NDVI could be meaningful especially for comparisons among your results. 

Your camera with a blue filter uses blue light for visible, so the visible LEDs you use should be blue. Depending on the filter, the NIR captured by the camera will be between 750-1000nm. An advantage of precisely controlling the wavelength of light illuminating the plants is that you can avoid some cross contamination. The red channel in the camera will be used to capture NIR, but it also captures red. If you illuminate the plants with light that contains no red, you have solved a big problem because there can be no red light contaminating the red channel. The blue channel will be used to capture blue, but it also captures NIR. If you know or can determine which NIR wavelengths the blue channel is most sensitive to, you can choose an NIR LED that avoids those wavelengths. Then there will be less NIR contaminating the blue channel. 

Chris