Introduction to ARCGIS Pro and Multispectral Imagery

Figure 1: ArcGIS Pro Opening Screen

Figure 2: ArcGIS logo

Introduction:

Taking imagery with unmanned aerial vehicles has many applications. Different sensors, platforms, and software's all impact how you are able to collect, process, and send out data. When taking certain types of imagery, specialized software is needed in order to properly utilize the collected data, such is the case with the following exercise. 

Disclaimer:

This is the second time I have done this exercise, but for some reason there were some discrepancies between the two times I completed this exercise. As such, I will be including examples of steps done where images show up differently the two times. I am not sure where the error occurred, but it is helpful to show the differences as it shows how knowing the software well allows a user to properly process data.

Doakburn Exercise: 

Metadata

An important first step when conducting projects like this is to have what is called a single source of truth. A single source of truth allows anyone working on the data to find information quickly and efficiently, in this case we used a metadata text file, the content of which is shown below.

Vehicle: Bramor PPX

Sensor: Altum

Flight Number: 2

Takeoff Time: 12:18 PM

Landing Time: 12:35 PM

Altitude (m): 121

Sensor Angle: NADIR

Having this quickly available allows accountability and consistency in your work, as well as functioning well with other tools such as Sky Vector or Landsat.

Wavelength Bands

Figure 3: Microsense RedEdge Peak Band Reflectance

Knowing which band display which reflectance values is important. Depending on the band used different wavelengths will be displayed. The values of the band 1-5 are displayed below.

Blue:           475 nm

Green:         560 nm

Red:            675 nm

Red Edge:    775 nm

NIR:            825-850 nm

Imagery

                                Figure 4: Doakburn field pre-burn

                                Figure 5: Doakburn field post-burn                                 Figure 6: symbology menu

The images about show the field without any alteration of bands displayed. You can see which areas were burned, but what is there in the data that we cannot see? Altering which bands are displayed lets a user see what the naked eye cannot. Changes which bands are displayed is done through the symbology window as shown to the right of figure 5.

Figure 7: Pre burn image with bands 5,3,2

Figure 8: Post burn image with bands 5,3,2

Figure 9: Post burn image with bands 4,3,2

Figure 10: Post burn image with bands 1,4,5

As you can readily see, changes the bands vastly changes the appearance of the image. Figures 7 and 8 show what is known as false color infrared. Using multispectral images such as these can show the health of vegetation, while showing areas of the field that are lacking in vegetation that you cannot see in the original image.

NDVI

Another option to available in ArcGIS pro is NDVI. This function works as a sort of color gradient mode. where the user can choose color scheme in order to display contrasting changes. For some reason on the second run of this exercise, I was unable to get color images of NDVI. As such the NDVI knowledge and imagery I will mention is

Figure 11: Postburn NDVI B/W

Figure 12: Preburn NDVI B/W

Like with the Symbology window to change bands, there is a symbology window for NDVI images that allows the user to change the color scheme of the image, either showing continuous data or discrete data. If the user chooses a gradient color scheme, the image will show continuous data. If the user chooses a non-gradient color scheme, the image will show discrete data. However, some images will not work with a discrete data scheme, such as the ones used in this lab, this is illustrated below:

Figure 13: NDVI color non-gradient

Figure 14: NDVI color gradient

Discussion:

Using software such as ArcGIS pro has many applications with multi-spectral imagery. In the images above with changed bands, or the NDVI gradient images, it is much easier to see which areas of the field lack vegetation post burn. Areas that you wouldn't normally notice, such as the footpaths become more obvious to the viewer when they know what to look for. To an uninformed viewer it may be a little confusing at first with the different colors, but after an explanation it becomes clear how this is an effective tool to gather data.

Conclusion:

Knowing which type of image processing one can use when handling data is important. Obviously some of the images used in this lab, such as the non-gradient ones, do not have very practical applications outside of experimenting with this software. However, when an individual knows how to properly use software such as ArcGIS pro to process imagery to find things in their data not easily seen, they will be able to extrapolate data gathered much more efficiently.