Interact

S p e c t r a l I n f o r m a t i o n

One way to use aerial images is through a "pixel by pixel" classification process. Pixels of similar spectral information (which we can simplify as meaning "color") are grouped together into different classes (or colors) representing different tree species. The example below shows what's called the "maximum likelihood classification technique" applied to an aerial image of the tropical rainforest.

Fine spatial resolution imagery of tropical tree canopy:
Digital images of the forest canopy from above can be obtained using the new generation of digital cameras, or by scanning traditional photographs/slides into computer memory. This example shows a digitized photograph of tropical forest canopy from a height of approximately 100 meters.

Image enhancement can improve understanding of the exact content of the image:
Various image enhancement techniques may be applied to digital imagery for better visualization of the image content. This example shows the first digital image after a "decorrelation stretch." Note how the intensity of the various colors increased, making the individual crowns of different tree species more distinguishable from each other.

A "classified" (or sorted) image, showing biodiversity of trees within the image:
By utilizing spectral information (e.g., the color of objects within an image) we can readily group together similar colors and produce a classified image. Primary colors were arbitrarily assigned to crowns and portions of the canopy with similar spectral reflectance, indicating more clearly the presence of similar and different tree species.

Remote Sensing: Seeing the Forest and the Trees

  Spatial Resolution
  Spectral Information
  Morphometric Information
  Tree Phenology
  Jungle Tower

	Fine spatial resolution imagery of tropical tree canopy: Digital images of the forest canopy from above can be obtained using the new generation of digital cameras, or by scanning traditional photographs/slides into computer memory. This example shows a digitized photograph of tropical forest canopy from a height of approximately 100 meters.
	Image enhancement can improve understanding of the exact content of the image: Various image enhancement techniques may be applied to digital imagery for better visualization of the image content. This example shows the first digital image after a "decorrelation stretch." Note how the intensity of the various colors increased, making the individual crowns of different tree species more distinguishable from each other.
	A "classified" (or sorted) image, showing biodiversity of trees within the image: By utilizing spectral information (e.g., the color of objects within an image) we can readily group together similar colors and produce a classified image. Primary colors were arbitrarily assigned to crowns and portions of the canopy with similar spectral reflectance, indicating more clearly the presence of similar and different tree species.