Within PrecisionMapper we have the ability to process multiband data from the Parrot Sequoia and Micasense RedEdge sensors. Since these sensors function very differently from your standard visual sensor some processing steps and output options differ for this particular sensor group.
Electromagnetic Spectrum- Multispectral Bands
Multispectral data comes in two different forms in the hardware that we provide- Modified and Multiband. To understand the difference between these two types of Multispectral data you must first understand Electromagnetic Spectrum and how these signatures function in a sensor.
A standard sensor collects Red Green and Blue wavelengths of light. Outside of the visual range of colors there are non visible wavelengths that you cannot see with the human eye but is possible to sense with the right hardware.
Above is the full chart of electromagnetic spectrum. As you can see only a small portion of the spectrum is actually visible. Multispectral sensors for the purposes of agriculture typically use some combination of Red, Green, Blue, and Near Infrared (NIR) bands wavelengths of the chart above. The NIR band is key in this Multispectral collection. This additional band allows us to make more accurate calculations around vegetation.
When you look at these bands and how they interact with Vegetation it becomes apparent why the NIR spectrum is so valuable. The Blue and Red bands of light are predominantly absorbed by leaves while the Green bands of the visual spectrum reflect back (Giving living plants their green color). Healthy vegetation will absorb more Blue and Red light to fuel photosynthesis and create chlorophyll in the plant. A plant with more chlorophyll will reflect more NIR energy than an unhealthy plant. Thus, by comparing the absorption of this blue and red light to the reflectance of the NIR wavelength we are able to provide valid information about a plants health.
Multiband Multispectral Imagery
A multi band multispectral sensor collects many different bands separately instead of a single combination of 3 bands collected by the Modified Multispectral sensor (See Multispectral Sensors article). There are only two Multi band sensors that are currently compatible in the PrecisionMapper system- The Parrot Sequoia and the Micasense RedEdge. Each of these sensors has a slightly different variation in the bands that it collects.
Parrot Sequoia- Collects 4 bands of data- Red, Green, Red Edge, and NIR. Separately sensor captures a full color visual image as well.
Micasense Red Edge- Collects 5 bands of data- Red, Green, Blue, Red Edge, and NIR
With these multiband sensors it collects a gray scale depiction of each spectrum that it collects. Since they are all provided separately this give you the ability to mix different combinations of 3 to meet your needs.With 5 band Micasense Red Edge sensor you can create RGB (Visual Imagery), BGNIR, or RGNIR without doing 3 separate flights.
Uploading Your Data
The steps to uploading your Micasense and/or Sequoia data follow nearly the same process as a standard upload mentioned in our Uploading Your Data section of the knowledge base. However there are a few things to check for specifically when uploading Multi band data.
- You should always use the PrecisionViewer desktop program for these uploads. There are typically more images in a survey from a multiband sensor like the Sequoia or Micansense. Larger uploads will perform best in the PrecisionViewer program.
- As you are uploading your data all bands from a capture must be present in the source data.
- Example: If you collected with a Micasense Rededge sensor you should have 5 bands of data present for each individual capture. If one of the 5 bands is left out PrecisionMapper will not accept the upload.
Here the software is looking for bands 1-5 captured by the Micasense sensor. But since the band #5 is missing you see an error message on the remaining 3 bands.
Since the Sequoia sensor itself outputs 4 individual bands and 1 composite visual image the upload of this data can be done a few different ways to yield either Visual or RGNIR data types in your final orthomosaic.
- Uploading the 4 individual bands of the capture will process a Red, Green, Near Infrared (RGNIR) orthomosaic
- Uploading the Visual image separate from your 4 bands will be process a Visual (RGB) orthomosaic product
- If you try to upload the 4 Bands + Visual image together the system will leave out the RGB composite from the process
This is important in order to understand the Applicable Algorithms to this particular sensor. It will only produce either Visual or RGNIR data types.
Through processing you are able to select all the same output options as your standard processing (Mentioned in PrecisionMapper Outputs article).
Since the Micasense sensor outputs 5 seprate bands there are more data types that can be created with this data.
- BGNIR- Blue, Green, Near Infrared
- RGNIR- Red, Green, Near Infrared
- RGB- Red, Green, Blue
This is important in order to understand the Applicable Algorithms to this particular sensor.