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

NAME

6       r.resamp.filter   -  Resamples raster map layers using an analytic ker‐
7       nel.
8

KEYWORDS

10       raster, resample, kernel filter, filter
11

SYNOPSIS

13       r.resamp.filter
14       r.resamp.filter --help
15       r.resamp.filter [-n] input=name output=name  filter=string[,string,...]
16       [radius=float[,float,...]]                 [x_radius=float[,float,...]]
17       [y_radius=float[,float,...]]    [--overwrite]   [--help]    [--verbose]
18       [--quiet]  [--ui]
19
20   Flags:
21       -n
22           Propagate NULLs
23
24       --overwrite
25           Allow output files to overwrite existing files
26
27       --help
28           Print usage summary
29
30       --verbose
31           Verbose module output
32
33       --quiet
34           Quiet module output
35
36       --ui
37           Force launching GUI dialog
38
39   Parameters:
40       input=name [required]
41           Name of input raster map
42
43       output=name [required]
44           Name for output raster map
45
46       filter=string[,string,...] [required]
47           Filter kernel(s)
48           Options:  box,  bartlett,  gauss,  normal, hermite, sinc, lanczos1,
49           lanczos2, lanczos3, hann, hamming, blackman
50
51       radius=float[,float,...]
52           Filter radius
53
54       x_radius=float[,float,...]
55           Filter radius (horizontal)
56
57       y_radius=float[,float,...]
58           Filter radius (vertical)
59

DESCRIPTION

61       r.resamp.filter resamples an input raster, filtering the input with  an
62       analytic  kernel. Each output cell is typically calculated based upon a
63       small subset of the input cells, not the entire input.  r.resamp.filter
64       performs  convolution  (i.e.  a  weighted  sum  is calculated for every
65       raster cell).
66
67       The module maps the input range to the width of the window function, so
68       wider windows will be "sharper" (have a higher cut-off frequency), e.g.
69       lanczos3 will be sharper than lanczos2.
70
71       r.resamp.filter implements FIR (finite impulse response) filtering. All
72       of  the  functions  are  low-pass  filters,  as they are symmetric. See
73       Wikipedia: Window function for examples of common window functions  and
74       their frequency responses.
75
76       A  piecewise-continuous function defined by sampled data can be consid‐
77       ered a mixture (sum) of the underlying signal and  quantisation  noise.
78       The  intent  of  a low pass filter is to discard the quantisation noise
79       while retaining the signal.  The cut-off frequency is  normally  chosen
80       according to the sampling frequency, as the quantisation noise is domi‐
81       nated by the sampling frequency and  its  harmonics.  In  general,  the
82       cut-off frequency is inversely proportional to the width of the central
83       "lobe" of the window function.
84
85       When using r.resamp.filter with a specific radius, a  specific  cut-off
86       frequency  regardless of the method is chosen. So while lanczos3 uses 3
87       times as large a window as lanczos1, the cut-off frequency remains  the
88       same. Effectively, the radius is "normalised".
89
90       All  of  the  kernels  specified by the filter parameter are multiplied
91       together. Typical usage will use either a single kernel or an  infinite
92       kernel along with a finite window.
93

NOTES

95       Resampling   modules   (r.resample,   r.resamp.stats,  r.resamp.interp,
96       r.resamp.rst, r.resamp.filter) resample the map to  match  the  current
97       region settings.
98
99       When using a kernel which can have negative values (sinc, Lanczos), the
100       -n flag should be used. Otherwise, extreme values can arise due to  the
101       total weight being close (or even equal) to zero.
102
103       Kernels  with  infinite  extent  (Gauss,  normal,  sinc, Hann, Hamming,
104       Blackman) must be used in conjunction with a finite windowing  function
105       (box, Bartlett, Hermite, Lanczos).
106
107       The way that Lanczos filters are defined, the number of samples is sup‐
108       posed to be proportional to the  order  ("a"  parameter),  so  lanczos3
109       should  use  3  times  as many samples (at the same sampling frequency,
110       i.e.  cover 3 times as large a time interval) as lanczos1 in  order  to
111       get  a  similar  frequency response (higher-order filters will fall off
112       faster, but the frequency at which the fall-off starts  should  be  the
113       same).  See  Wikipedia: Lanczos-kernel.svg for an illustration. If both
114       graphs were drawn on the same axes, they would have  roughly  the  same
115       shape, but the a=3 window would have a longer tail. By scaling the axes
116       to the same width, the a=3 window has a narrower central lobe.
117
118       For longitude-latitude locations, the interpolation algorithm is  based
119       on  degree  fractions,  not on the absolute distances between cell cen‐
120       ters.  Any attempt to implement the latter would violate the  integrity
121       of the interpolation method.
122

SEE ALSO

124          g.region,   r.mfilter,  r.resample,  r.resamp.interp,  r.resamp.rst,
125       r.resamp.stats
126
127       Overview: Interpolation and Resampling in GRASS GIS
128

AUTHOR

130       Glynn Clements
131
132       Last changed: $Date: 2016-09-19 12:29:41 +0200 (Mon, 19 Sep 2016) $
133

SOURCE CODE

135       Available at: r.resamp.filter source code (history)
136
137       Main index | Raster index | Topics index | Keywords index  |  Graphical
138       index | Full index
139
140       © 2003-2019 GRASS Development Team, GRASS GIS 7.6.0 Reference Manual
141
142
143
144GRASS 7.6.0                                                 r.resamp.filter(1)
Impressum