9315 Rate this article:
4.8

"Pop!" Goes the Feature

Rebecca Lasica

n the surface (no pun intended), temporal data might look like a straight forward time-enabled data set. But thinking about it outside the box, there are many different ways that time series tools lend themselves to looking at raster data, even if it is a single file taken at a single time! 

Note: All data used in these videos courtesy of Digital Globe: Video 1: QuickBird Video 2: WorldView -2 

For example, take a look at the first video. This is a raster series of iso data classification results for the same file as input. A different number of classes were selected for each iteration, resulting in a series that is interesting because various features are clearly delineated at different class thresholds. Watch how the water bodies really stand out in the end. Also watch how the healthy fields are separated from the other vegetation when the number of classes increase.

Another way I thought I’d look at a raster series was to take a multispectral file and run all the appropriate spectral indices to see if a certain index highlights my feature of interest more clearly than another. For this particular example, I used a WV-2 data set courtesy of Digital Globe. The data were first calibrated, corrected, and pan sharpened using the Radiometric Calibration tool, QUAC atmospheric correction,and NN Diffuse pan sharpening respectively. Then I wrote a small ENVI extension to identify and run all the appropriate spectral indices based on the input wavelengths = 39 out of 64 spectral indices were calculated! I imported a few of those results into the Raster Series tool to generate the movie you see here. I added some color to make the features “pop”.  See which indices you think are best for delineating water, vegetation, or impervious surfaces in this scene:

Included here are:

  1. Soil Adjusted Vegetation Index
  2. WorldView Improved Vegetation Index
  3. WorldView New Iron Oxide Index
  4. WorldView Built-upIndex
  5. Optimized Soil Adjusted Vegetation Index
  6. Modified Simple Raito
  7. Green Difference Vegetation Index

Please login or register to post comments.