Four alternative models for an infrared plant health camera.

1. 1) One camera with a filter switcher which switches between IR block and IR pass filters
2. 2) Two cameras, one unmodified and one with an IR pass filter
3. 3) One camera with a superblue (Schott BG3) filter.
4. 4) Two cameras, one unmodified and one with a superblue filter.

|Feature | ..1..|..2..|..3..|..4..| |:----------------------------------------------------------------|:--------------------:|:--------------------:|:-------------------:|:-----------------: |number of cameras|1|2|1|2 |filter switcher|1|0|0|0 |normal RGB color output|Y|Y|N|Y |optics modified for RGB image?|Y|N|na|N |NRG output|Y|Y|N|Y |NDVI output|Y|Y|Y|Y |shutter sync for NDVI|N|Y|N|N |shutter sync for NRG|N|Y|na|Y |rectification for NDVI|Y|Y|N|N |rectification for NRG|Y|Y|na|Y |parallax for NDVI|N|Y|N|N |parallax for NRG|N|Y|na|Y |RGB/NIR synchrony possible?|N|Y|Y|Y |buy superblue filter?|N|N|Y|Y .

Notes: 1. Shutter sync: Is it necessary to synchronize the shutters of two cameras to make multispectral images when the cameras or subject are moving? 2. Rectification: Is it necessary to rectify (align) data from two different photos to make multispectral images? 3. Parallax: Is rectifying two images subject to parallax errors when subjects in the scene are both near and far from the camera? 4.RGB/NIR synchrony: Is it possible to capture all the data needed for multispectral images simultaneously (which is needed when the camera[s] or subject are not stationary)?