This discussion provides a brief introduction to plotting spectra in n-D space. To bypass this and to begin using the n-D Visualizer, skip to Starting the n-D Visualizer.

Think of spectra as points in an n-D scatter plot, where n is the number of bands. The coordinates of the points in n-D space consist of n spectral radiance or reflectance values in each band for a given pixel. You can use the distribution of these points in n-D space to estimate the number of spectral endmembers and their pure spectral signatures.

Endmembers are pure spectrally unique materials that occur in a scene. Using a linear unmixing model, you can reconstruct every spectrum in the image as some combination of image endmember spectra.

The n-D Visualizer was designed to receive region of interest (ROI) input containing the spectrally purest pixels in a scene, and to allow you to segregate these pure pixels into their respective endmembers. ROIs are used because interactively rotating all of the data in a full image is too computationally intensive. A pure pixel is one that is the closest to containing only one spectrally unique endmember material. Defining an ROI that contains the purest pixels in an image is easy when you use the Pixel Purity Index. You can also import an ROI to the n-D Visualizer that is not based on the purest pixels in the image. However, doing so means that some endmembers may not be represented in the resulting n-D scatter plot.