Multispectral Image Indices

Introduction:
Multispectral images allow the original image to be put through multiple spectrums in order to gain clarity or even other information that wasn't inherently present. In this lab, a field that was burned in controlled areas was photographed before and after. The images were then put through multiple spectrums in order to see how the areas of burn, the vegetation, and the non-burn areas were changed.

Method:
Using ArcGIS Pro, the images were loaded into the program to be viewed/edited. The images were first viewed in the bands red-1, green-2, and blue-3 (Figure.1 and Figure.2). This showed the fields as they were taken and the burned areas are shown in a dark brown/purple color while the other areas are shown in green. Next, the bands of the same images were changed to red-5, green-3, and blue-2 (Figure.3 and Figure.4). This essentially flipped the colors, making the burned areas a dark green and the non-burned areas a light purple. After this, the NDVI images were loaded in order to see the vegetation as well in different spectrums. The color scheme of these post and pre-burn images were changed to "stretch." This introduced a multitude of different colors into the spectrum. (Figure.5 and Figure.6) Lastly, the color scheme of these images was changed to "cyan to purple." (Figure.7 and Figure.8).

Figure.1 Bands 3-2-1 Post Burn

Figure.2 Bands 3-2-1 Pre Burn

Figure.3 Bands 5-3-2 Post Burn

Figure.4 Bands 5-3-2 Pre Burn

Figure.5 NDVI Stretch Post Burn

Figure.6 NDVI Stretch Pre Burn

Figure.7 NDVI Cyan Post Burn

Figure.8 NDVI Cyan Pre Burn

Discussion:
Multispectral indices allow for images to be viewed in different colors and spectrums that differ from the original. In this lab, it is shown that changing the color spectrum gives the post and pre burn imaging very different information. For instance, when the colors were drastically changed in Figures 7 and 8, the viewer can easily distinguish the post and pre burn areas from each other. However, a large downside to this is if the editor does not state exactly what each area/color represents. This could cause some issues down the line if communication and information are not relayed properly with multispectral imaging.

Conclusion:
Multispectral imaging is a great way to edit images in order to pinpoint the exact information the viewer is looking for. However, it should be used with caution to ensure all parties involved can easily view and understand the images