r.texture(1)

1r.texture(1)                  Grass User's Manual                 r.texture(1)
2
3
4

NAME

6       r.texture  - Generate images with textural features from a raster map.
7

KEYWORDS

9       raster
10

SYNOPSIS

12       r.texture
13       r.texture help
14       r.texture  [-qackviswxedpmno]  input=name  prefix=string   [size=value]
15       [distance=value]   [--overwrite]  [--verbose]  [--quiet]
16
17   Flags:
18       -q
19           Quiet
20
21       -a
22           Angular Second Moment
23
24       -c
25           Contrast
26
27       -k
28           Correlation
29
30       -v
31           Variance
32
33       -i
34           Inverse Diff Moment
35
36       -s
37           Sum Average
38
39       -w
40           Sum Variance
41
42       -x
43           Sum Entropy
44
45       -e
46           Entropy
47
48       -d
49           Difference Variance
50
51       -p
52           Difference Entropy
53
54       -m
55           Measure of Correlation-1
56
57       -n
58           Measure of Correlation-2
59
60       -o
61           Max Correlation Coeff
62
63       --overwrite
64           Allow output files to overwrite existing files
65
66       --verbose
67           Verbose module output
68
69       --quiet
70           Quiet module output
71
72   Parameters:
73       input=name
74           Name of input raster map
75
76       prefix=string
77           Prefix for ouput raster map(s)
78
79       size=value
80           The size of sliding window (odd and >= 3)
81           Default: 3
82
83       distance=value
84           The distance between two samples (>= 1)
85           Default: 1
86

DESCRIPTION

88       r.texture - Creates map raster with textural features  for  user-speci‐
89       fied raster map layer. The module calculates textural features based on
90       spatial dependence matrices at 0, 45, 90, and 135 degrees  for  a  dis‐
91       tance (default = 1).
92
93       r.texture  reads  a  GRASS  raster map as input and calculates textural
94       features based on spatial dependence matrices  for  north-south,  east-
95       west,  northwest,  and southwest directions using a side by side neigh‐
96       borhood (i.e., a distance of 1). Be sure to carefully set your  resolu‐
97       tion  (using  g.region)  before running this program, or else your com‐
98       puter could run out of memory.  Also, make sure that  your  raster  map
99       has  no more than 255 categories.  The output consists into four images
100       for each textural feature, one for every direction.
101
102       A commonly used texture model is based on the so-called grey level  co-
103       occurrence  matrix.  This matrix is a two-dimensional histogram of grey
104       levels for a pair of pixels which are  separated  by  a  fixed  spatial
105       relationship.   The matrix approximates the joint probability distribu‐
106       tion of a pair of pixels.  Several texture measures are  directly  com‐
107       puted from the grey level co-occurrence matrix.
108
109       The following are brief explanations of texture measures:
110
111                      Angular  Second Moment: This is a measure of local homo‐
112                     geneity and the opposite of Entropy.  It is high when the
113                     local window a few pixels with high values; low, when the
114                     pixels are almost equal.
115
116                      Contrast: This measure considers  the  amount  of  local
117                     variation  and  is the opposite of Homogeneity (when high
118                     pixel values concentrate along the diagonal).
119
120                      Correlation: This measure  analyses  the  linear  depen‐
121                     dency  of  grey  levels  of neighboring pixels. Typically
122                     high, when the scale of local texture is larger than  the
123                     distance.
124
125                      Entropy:  This  measure  is  high when the values of the
126                     local window have similar values.  It  is  low  when  the
127                     values  are  close to either 0 or 1 (i.e. when the pixels
128                     in the local window are uniform).
129

NOTES

131       Algorithm taken from:
132       Haralick, R.M., K. Shanmugam, and I. Dinstein. 1973. Textural  features
133       for image classification. IEEE Transactions on Systems, Man, and Cyber‐
134       netics, SMC-3(6):610-621.
135
136       The code was taken by  permission  from  pgmtexture,  part  of  PBMPLUS
137       (Copyright  1991,  Jef Poskanser and Texas Agricultural Experiment Sta‐
138       tion, employer for hire of James Darrell McCauley).
139       Man page of pgmtexture
140

BUGS

142       - The program can run incredibly slow for large raster maps.
143
144       - The method for finding the  maximal  correlation  coefficient,  which
145       requires  finding the second largest eigenvalue of a matrix Q, does not
146       always converge.
147

REFERENCES

149       Haralick, R.M., K. Shanmugam, and I. Dinstein (1973). Textural features
150       for image classification. IEEE Transactions on Systems, Man, and Cyber‐
151       netics, SMC-3(6):610-621.
152
153       Bouman C. A., Shapiro M.,(March 1994).A Multiscale Random  Field  Model
154       for  Bayesian Image Segmentation, IEEE Trans. on Image Processing, vol.
155       3, no.2.
156
157       Haralick R., (May 1979). Statistical and structural approaches to  tex‐
158       ture, Proceedings of the IEEE, vol. 67, No.5, pp. 786-804
159

AUTHOR

164       G.  Antoniol  -  RCOST  (Research Centre on Software Technology - Viale
165       Traiano - 82100 Benevento)
166       C. Basco -  RCOST (Research Centre on Software Technology - Viale  Tra‐
167       iano - 82100 Benevento)
168       M. Ceccarelli - Facolta di Scienze, Universita del Sannio, Benevento
169
170       Last changed: $Date: 2007-07-13 16:18:35 +0200 (Fri, 13 Jul 2007) $
171
172       Full index
173
174       © 2003-2008 GRASS Development Team
175
176
177
178GRASS 6.3.0                                                       r.texture(1)