source: prdatasets/texturel.m @ 101

Last change on this file since 101 was 81, checked in by bduin, 11 years ago
File size: 953 bytes
RevLine 
[80]1%TEXTUREL 128 x 640 objects (pixels) with 7 features in 5 classes
2%
3%       A = TEXTUREL
4%
5% Load the dataset in A. These are 5 texture classes each given by 7
6% features which are 7 different filtered versions of 5 different images.
7% The objects are the pixels. Filter number 7 is the unity filter, so the
8% corresponding featurevalues are the original pixels. The 5 classes are
9% stored as 5 horizontally concatenated images. This may be visualized by
10% SHOW(A), which shows 7 128*640 images (5*128=640) and the labels.
11%
12% The dataset TEXTURET deals with a similar texture image given by the same
13% texture features but as a composite 256*256 image.
14%
15% See also DATASETS, PRDATASETS, TEXTURET
16
17% Copyright: R.P.W. Duin, r.p.w.duin@prtools.org
18% Faculty EWI, Delft University of Technology
19% P.O. Box 5031, 2600 GA Delft, The Netherlands
20
21function a = texturel
22
23prdatasets(mfilename,4);
[81]24a = pr_dataset('texturel');
[80]25a = setname(a,'Texture Learning Set');
Note: See TracBrowser for help on using the repository browser.