pca2
|
Contents: Description, Arguments, Usage, Examples, Images, Sub-Functions, Related Functions, Source Supported Under Version: >1.00
Description This function computes principal component images from an input image and a set of eigenvectors. This function is different from pca in that it allows for application to various multiband images in multiple data formats. It also take the output from eigen2, which calculates the eigenvectors the slow way, which greatly increases the accuracy of higher order eigenvectors and eigenvalues.
Arguments and Return Values Arguments: Input image cube, output structure from eigen2. Return Value: Image cube with the same x and y dimensions as the input and the same z dimension as the number of input eigenvectors and aigenvalues.
Usage Syntax: pca2(image,ev) image: surface pressure (m x n x k array) ev: output structure from eigen2 containing .evecs (k x j x 1 array), .evals (j x 1 x 1 array), .avg (k x 1 x 1 array), and .C (i x j x 1 array) Arrays are arbitrary in size, but 'k' must be the same in image and ev.evecs.
Examples dv> a
272x242x8 array of float, bsq format [2,106,368 bytes]
dv> ev = eigen2(a[1:272:5,1:242:5]*5000,8)
EV# Iterat. Eigenvalue
---+--------+-----------
1 200 771.579984
2 21 2.266776
3 49 0.748638
4 31 0.491434
5 38 0.257143
6 142 0.168165
7 53 0.150262
8 4 0.114633
struct, 4 elements
evecs: 8x8x1 array of double, bsq format [512 bytes]
evals: 8x1x1 array of float, bsq format [32 bytes]
avg: 8x1x1 array of double, bsq format [64 bytes]
C: 2695x8x1 array of float, bsq format [86,240 bytes]
dv> apc = pca2(a,ev)
272x242x8 array of double, bip format [4,212,736 bytes]
dv>
Images Individual principal component images from the example above. PC1 is the upper left image and PC8 is the lower right image.
|
DavinciWiki Mini-Nav Bar Contents
Contact Developers
All other topics
Recent Library Changes Created On: 01-01-2008 |
