NAME
Image::Leptonica::Func::scalelow
VERSION
version 0.04
scalelow.c
scalelow.c
Color (interpolated) scaling: general case
void scaleColorLILow()
Grayscale (interpolated) scaling: general case
void scaleGrayLILow()
Color (interpolated) scaling: 2x upscaling
void scaleColor2xLILow()
void scaleColor2xLILineLow()
Grayscale (interpolated) scaling: 2x upscaling
void scaleGray2xLILow()
void scaleGray2xLILineLow()
Grayscale (interpolated) scaling: 4x upscaling
void scaleGray4xLILow()
void scaleGray4xLILineLow()
Grayscale and color scaling by closest pixel sampling
l_int32 scaleBySamplingLow()
Color and grayscale downsampling with (antialias) lowpass filter
l_int32 scaleSmoothLow()
void scaleRGBToGray2Low()
Color and grayscale downsampling with (antialias) area mapping
l_int32 scaleColorAreaMapLow()
l_int32 scaleGrayAreaMapLow()
l_int32 scaleAreaMapLow2()
Binary scaling by closest pixel sampling
l_int32 scaleBinaryLow()
Scale-to-gray 2x
void scaleToGray2Low()
l_uint32 *makeSumTabSG2()
l_uint8 *makeValTabSG2()
Scale-to-gray 3x
void scaleToGray3Low()
l_uint32 *makeSumTabSG3()
l_uint8 *makeValTabSG3()
Scale-to-gray 4x
void scaleToGray4Low()
l_uint32 *makeSumTabSG4()
l_uint8 *makeValTabSG4()
Scale-to-gray 6x
void scaleToGray6Low()
l_uint8 *makeValTabSG6()
Scale-to-gray 8x
void scaleToGray8Low()
l_uint8 *makeValTabSG8()
Scale-to-gray 16x
void scaleToGray16Low()
Grayscale mipmap
l_int32 scaleMipmapLow()FUNCTIONS
makeSumTabSG2
l_uint32 * makeSumTabSG2 ( void )
makeSumTabSG2()
Returns a table of 256 l_uint32s, giving the four output
8-bit grayscale sums corresponding to 8 input bits of a binary
image, for a 2x scale-to-gray op. The sums from two
adjacent scanlines are then added and transformed to
output four 8 bpp pixel values, using makeValTabSG2().makeSumTabSG3
l_uint32 * makeSumTabSG3 ( void )
makeSumTabSG3()
Returns a table of 64 l_uint32s, giving the two output
8-bit grayscale sums corresponding to 6 input bits of a binary
image, for a 3x scale-to-gray op. In practice, this would
be used three times (on adjacent scanlines), and the sums would
be added and then transformed to output 8 bpp pixel values,
using makeValTabSG3().makeSumTabSG4
l_uint32 * makeSumTabSG4 ( void )
makeSumTabSG4()
Returns a table of 256 l_uint32s, giving the two output
8-bit grayscale sums corresponding to 8 input bits of a binary
image, for a 4x scale-to-gray op. The sums from four
adjacent scanlines are then added and transformed to
output 8 bpp pixel values, using makeValTabSG4().makeValTabSG2
l_uint8 * makeValTabSG2 ( void )
makeValTabSG2()
Returns an 8 bit value for the sum of ON pixels
in a 2x2 square, according to
val = 255 - (255 * sum)/4
where sum is in set {0,1,2,3,4}makeValTabSG3
l_uint8 * makeValTabSG3 ( void )
makeValTabSG3()
Returns an 8 bit value for the sum of ON pixels
in a 3x3 square, according to
val = 255 - (255 * sum)/9
where sum is in set {0, ... ,9}makeValTabSG4
l_uint8 * makeValTabSG4 ( void )
makeValTabSG4()
Returns an 8 bit value for the sum of ON pixels
in a 4x4 square, according to
val = 255 - (255 * sum)/16
where sum is in set {0, ... ,16}makeValTabSG6
l_uint8 * makeValTabSG6 ( void )
makeValTabSG6()
Returns an 8 bit value for the sum of ON pixels
in a 6x6 square, according to
val = 255 - (255 * sum)/36
where sum is in set {0, ... ,36}makeValTabSG8
l_uint8 * makeValTabSG8 ( void )
makeValTabSG8()
Returns an 8 bit value for the sum of ON pixels
in an 8x8 square, according to
val = 255 - (255 * sum)/64
where sum is in set {0, ... ,64}scaleAreaMapLow2
void scaleAreaMapLow2 ( l_uint32 *datad, l_int32 wd, l_int32 hd, l_int32 wpld, l_uint32 *datas, l_int32 d, l_int32 wpls )
scaleAreaMapLow2()
Note: This function is called with either 8 bpp gray or 32 bpp RGB.
The result is a 2x reduced dest.scaleBinaryLow
l_int32 scaleBinaryLow ( l_uint32 *datad, l_int32 wd, l_int32 hd, l_int32 wpld, l_uint32 *datas, l_int32 ws, l_int32 hs, l_int32 wpls )
scaleBinaryLow()
Notes:
(1) The dest must be cleared prior to this operation,
and we clear it here in the low-level code.
(2) We reuse dest pixels and dest pixel rows whenever
possible for upscaling; downscaling is done by
strict subsampling.scaleBySamplingLow
l_int32 scaleBySamplingLow ( l_uint32 *datad, l_int32 wd, l_int32 hd, l_int32 wpld, l_uint32 *datas, l_int32 ws, l_int32 hs, l_int32 d, l_int32 wpls )
scaleBySamplingLow()
Notes:
(1) The dest must be cleared prior to this operation,
and we clear it here in the low-level code.
(2) We reuse dest pixels and dest pixel rows whenever
possible. This speeds the upscaling; downscaling
is done by strict subsampling and is unaffected.
(3) Because we are sampling and not interpolating, this
routine works directly, without conversion to full
RGB color, for 2, 4 or 8 bpp palette color images.scaleColor2xLILineLow
void scaleColor2xLILineLow ( l_uint32 *lined, l_int32 wpld, l_uint32 *lines, l_int32 ws, l_int32 wpls, l_int32 lastlineflag )
scaleColor2xLILineLow()
Input: lined (ptr to top destline, to be made from current src line)
wpld
lines (ptr to current src line)
ws
wpls
lastlineflag (1 if last src line; 0 otherwise)
Return: void
*** Warning: implicit assumption about RGB component ordering scaleColor2xLILow
void scaleColor2xLILow ( l_uint32 *datad, l_int32 wpld, l_uint32 *datas, l_int32 ws, l_int32 hs, l_int32 wpls )
scaleColor2xLILow()
This is a special case of 2x expansion by linear
interpolation. Each src pixel contains 4 dest pixels.
The 4 dest pixels in src pixel 1 are numbered at
their UL corners. The 4 dest pixels in src pixel 1
are related to that src pixel and its 3 neighboring
src pixels as follows:
1-----2-----|-----|-----|
| | | | |
| | | | |
src 1 --> 3-----4-----| | | <-- src 2
| | | | |
| | | | |
|-----|-----|-----|-----|
| | | | |
| | | | |
src 3 --> | | | | | <-- src 4
| | | | |
| | | | |
|-----|-----|-----|-----|
dest src
---- ---
dp1 = sp1
dp2 = (sp1 + sp2) / 2
dp3 = (sp1 + sp3) / 2
dp4 = (sp1 + sp2 + sp3 + sp4) / 4
We iterate over the src pixels, and unroll the calculation
for each set of 4 dest pixels corresponding to that src
pixel, caching pixels for the next src pixel whenever possible.
The method is exactly analogous to the one we use for
scaleGray2xLILow() and its line version.
P3 speed is about 5 x 10^7 dst pixels/sec/GHzscaleColorAreaMapLow
void scaleColorAreaMapLow ( l_uint32 *datad, l_int32 wd, l_int32 hd, l_int32 wpld, l_uint32 *datas, l_int32 ws, l_int32 hs, l_int32 wpls )
scaleColorAreaMapLow()
This should only be used for downscaling.
We choose to divide each pixel into 16 x 16 sub-pixels.
This is much slower than scaleSmoothLow(), but it gives a
better representation, esp. for downscaling factors between
1.5 and 5. All src pixels are subdivided into 256 sub-pixels,
and are weighted by the number of sub-pixels covered by
the dest pixel. This is about 2x slower than scaleSmoothLow(),
but the results are significantly better on small text.scaleColorLILow
void scaleColorLILow ( l_uint32 *datad, l_int32 wd, l_int32 hd, l_int32 wpld, l_uint32 *datas, l_int32 ws, l_int32 hs, l_int32 wpls )
scaleColorLILow()
We choose to divide each pixel into 16 x 16 sub-pixels.
Linear interpolation is equivalent to finding the
fractional area (i.e., number of sub-pixels divided
by 256) associated with each of the four nearest src pixels,
and weighting each pixel value by this fractional area.
P3 speed is about 7 x 10^6 dst pixels/sec/GHzscaleGray2xLILineLow
void scaleGray2xLILineLow ( l_uint32 *lined, l_int32 wpld, l_uint32 *lines, l_int32 ws, l_int32 wpls, l_int32 lastlineflag )
scaleGray2xLILineLow()
Input: lined (ptr to top destline, to be made from current src line)
wpld
lines (ptr to current src line)
ws
wpls
lastlineflag (1 if last src line; 0 otherwise)
Return: voidscaleGray2xLILow
void scaleGray2xLILow ( l_uint32 *datad, l_int32 wpld, l_uint32 *datas, l_int32 ws, l_int32 hs, l_int32 wpls )
scaleGray2xLILow()
This is a special case of 2x expansion by linear
interpolation. Each src pixel contains 4 dest pixels.
The 4 dest pixels in src pixel 1 are numbered at
their UL corners. The 4 dest pixels in src pixel 1
are related to that src pixel and its 3 neighboring
src pixels as follows:
1-----2-----|-----|-----|
| | | | |
| | | | |
src 1 --> 3-----4-----| | | <-- src 2
| | | | |
| | | | |
|-----|-----|-----|-----|
| | | | |
| | | | |
src 3 --> | | | | | <-- src 4
| | | | |
| | | | |
|-----|-----|-----|-----|
dest src
---- ---
dp1 = sp1
dp2 = (sp1 + sp2) / 2
dp3 = (sp1 + sp3) / 2
dp4 = (sp1 + sp2 + sp3 + sp4) / 4
We iterate over the src pixels, and unroll the calculation
for each set of 4 dest pixels corresponding to that src
pixel, caching pixels for the next src pixel whenever possible.scaleGray4xLILineLow
void scaleGray4xLILineLow ( l_uint32 *lined, l_int32 wpld, l_uint32 *lines, l_int32 ws, l_int32 wpls, l_int32 lastlineflag )
scaleGray4xLILineLow()
Input: lined (ptr to top destline, to be made from current src line)
wpld
lines (ptr to current src line)
ws
wpls
lastlineflag (1 if last src line; 0 otherwise)
Return: voidscaleGray4xLILow
void scaleGray4xLILow ( l_uint32 *datad, l_int32 wpld, l_uint32 *datas, l_int32 ws, l_int32 hs, l_int32 wpls )
scaleGray4xLILow()
This is a special case of 4x expansion by linear
interpolation. Each src pixel contains 16 dest pixels.
The 16 dest pixels in src pixel 1 are numbered at
their UL corners. The 16 dest pixels in src pixel 1
are related to that src pixel and its 3 neighboring
src pixels as follows:
1---2---3---4---|---|---|---|---|
| | | | | | | | |
5---6---7---8---|---|---|---|---|
| | | | | | | | |
src 1 --> 9---a---b---c---|---|---|---|---| <-- src 2
| | | | | | | | |
d---e---f---g---|---|---|---|---|
| | | | | | | | |
|===|===|===|===|===|===|===|===|
| | | | | | | | |
|---|---|---|---|---|---|---|---|
| | | | | | | | |
src 3 --> |---|---|---|---|---|---|---|---| <-- src 4
| | | | | | | | |
|---|---|---|---|---|---|---|---|
| | | | | | | | |
|---|---|---|---|---|---|---|---|
dest src
---- ---
dp1 = sp1
dp2 = (3 * sp1 + sp2) / 4
dp3 = (sp1 + sp2) / 2
dp4 = (sp1 + 3 * sp2) / 4
dp5 = (3 * sp1 + sp3) / 4
dp6 = (9 * sp1 + 3 * sp2 + 3 * sp3 + sp4) / 16
dp7 = (3 * sp1 + 3 * sp2 + sp3 + sp4) / 8
dp8 = (3 * sp1 + 9 * sp2 + 1 * sp3 + 3 * sp4) / 16
dp9 = (sp1 + sp3) / 2
dp10 = (3 * sp1 + sp2 + 3 * sp3 + sp4) / 8
dp11 = (sp1 + sp2 + sp3 + sp4) / 4
dp12 = (sp1 + 3 * sp2 + sp3 + 3 * sp4) / 8
dp13 = (sp1 + 3 * sp3) / 4
dp14 = (3 * sp1 + sp2 + 9 * sp3 + 3 * sp4) / 16
dp15 = (sp1 + sp2 + 3 * sp3 + 3 * sp4) / 8
dp16 = (sp1 + 3 * sp2 + 3 * sp3 + 9 * sp4) / 16
We iterate over the src pixels, and unroll the calculation
for each set of 16 dest pixels corresponding to that src
pixel, caching pixels for the next src pixel whenever possible.scaleGrayAreaMapLow
void scaleGrayAreaMapLow ( l_uint32 *datad, l_int32 wd, l_int32 hd, l_int32 wpld, l_uint32 *datas, l_int32 ws, l_int32 hs, l_int32 wpls )
scaleGrayAreaMapLow()
This should only be used for downscaling.
We choose to divide each pixel into 16 x 16 sub-pixels.
This is about 2x slower than scaleSmoothLow(), but the results
are significantly better on small text, esp. for downscaling
factors between 1.5 and 5. All src pixels are subdivided
into 256 sub-pixels, and are weighted by the number of
sub-pixels covered by the dest pixel.scaleGrayLILow
void scaleGrayLILow ( l_uint32 *datad, l_int32 wd, l_int32 hd, l_int32 wpld, l_uint32 *datas, l_int32 ws, l_int32 hs, l_int32 wpls )
scaleGrayLILow()
We choose to divide each pixel into 16 x 16 sub-pixels.
Linear interpolation is equivalent to finding the
fractional area (i.e., number of sub-pixels divided
by 256) associated with each of the four nearest src pixels,
and weighting each pixel value by this fractional area.scaleMipmapLow
l_int32 scaleMipmapLow ( l_uint32 *datad, l_int32 wd, l_int32 hd, l_int32 wpld, l_uint32 *datas1, l_int32 wpls1, l_uint32 *datas2, l_int32 wpls2, l_float32 red )
scaleMipmapLow()
See notes in scale.c for pixScaleToGrayMipmap(). This function
is here for pedagogical reasons. It gives poor results on document
images because of aliasing.scaleRGBToGray2Low
void scaleRGBToGray2Low ( l_uint32 *datad, l_int32 wd, l_int32 hd, l_int32 wpld, l_uint32 *datas, l_int32 wpls, l_float32 rwt, l_float32 gwt, l_float32 bwt )
scaleRGBToGray2Low()
Notes:
(1) This function is called with 32 bpp RGB src and 8 bpp,
half-resolution dest. The weights should add to 1.0.scaleSmoothLow
l_int32 scaleSmoothLow ( l_uint32 *datad, l_int32 wd, l_int32 hd, l_int32 wpld, l_uint32 *datas, l_int32 ws, l_int32 hs, l_int32 d, l_int32 wpls, l_int32 size )
scaleSmoothLow()
Notes:
(1) This function is called on 8 or 32 bpp src and dest images.
(2) size is the full width of the lowpass smoothing filter.
It is correlated with the reduction ratio, being the
nearest integer such that size is approximately equal to hs / hd.scaleToGray16Low
void scaleToGray16Low ( l_uint32 *datad, l_int32 wd, l_int32 hd, l_int32 wpld, l_uint32 *datas, l_int32 wpls, l_int32 *tab8 )
scaleToGray16Low()
Input: usual image variables
tab8 (made from makePixelSumTab8())
Return: 0 if OK; 1 on error.
The output is processed one dest byte at a time, corresponding
to 16 rows consisting each of 2 src bytes in the input image.
This uses one lookup table, tab8, which gives the sum of
ON pixels in a byte. After summing for all ON pixels in the
32 src bytes, which is between 0 and 256, this is converted
to an 8 bpp grayscale value between 0 (for 255 or 256 bits ON)
and 255 (for 0 bits ON).scaleToGray2Low
void scaleToGray2Low ( l_uint32 *datad, l_int32 wd, l_int32 hd, l_int32 wpld, l_uint32 *datas, l_int32 wpls, l_uint32 *sumtab, l_uint8 *valtab )
scaleToGray2Low()
Input: usual image variables
sumtab (made from makeSumTabSG2())
valtab (made from makeValTabSG2())
Return: 0 if OK; 1 on error.
The output is processed in sets of 4 output bytes on a row,
corresponding to 4 2x2 bit-blocks in the input image.
Two lookup tables are used. The first, sumtab, gets the
sum of ON pixels in 4 sets of two adjacent bits,
storing the result in 4 adjacent bytes. After sums from
two rows have been added, the second table, valtab,
converts from the sum of ON pixels in the 2x2 block to
an 8 bpp grayscale value between 0 (for 4 bits ON)
and 255 (for 0 bits ON).scaleToGray3Low
void scaleToGray3Low ( l_uint32 *datad, l_int32 wd, l_int32 hd, l_int32 wpld, l_uint32 *datas, l_int32 wpls, l_uint32 *sumtab, l_uint8 *valtab )
scaleToGray3Low()
Input: usual image variables
sumtab (made from makeSumTabSG3())
valtab (made from makeValTabSG3())
Return: 0 if OK; 1 on error
Each set of 8 3x3 bit-blocks in the source image, which
consist of 72 pixels arranged 24 pixels wide by 3 scanlines,
is converted to a row of 8 8-bit pixels in the dest image.
These 72 pixels of the input image are runs of 24 pixels
in three adjacent scanlines. Each run of 24 pixels is
stored in the 24 LSbits of a 32-bit word. We use 2 LUTs.
The first, sumtab, takes 6 of these bits and stores
sum, taken 3 bits at a time, in two bytes. (See
makeSumTabSG3). This is done for each of the 3 scanlines,
and the results are added. We now have the sum of ON pixels
in the first two 3x3 blocks in two bytes. The valtab LUT
then converts these values (which go from 0 to 9) to
grayscale values between between 255 and 0. (See makeValTabSG3).
This process is repeated for each of the other 3 sets of
6x3 input pixels, giving 8 output pixels in total.
Note: because the input image is processed in groups of
24 x 3 pixels, the process clips the input height to
(h - h % 3) and the input width to (w - w % 24).scaleToGray4Low
void scaleToGray4Low ( l_uint32 *datad, l_int32 wd, l_int32 hd, l_int32 wpld, l_uint32 *datas, l_int32 wpls, l_uint32 *sumtab, l_uint8 *valtab )
scaleToGray4Low()
Input: usual image variables
sumtab (made from makeSumTabSG4())
valtab (made from makeValTabSG4())
Return: 0 if OK; 1 on error.
The output is processed in sets of 2 output bytes on a row,
corresponding to 2 4x4 bit-blocks in the input image.
Two lookup tables are used. The first, sumtab, gets the
sum of ON pixels in two sets of four adjacent bits,
storing the result in 2 adjacent bytes. After sums from
four rows have been added, the second table, valtab,
converts from the sum of ON pixels in the 4x4 block to
an 8 bpp grayscale value between 0 (for 16 bits ON)
and 255 (for 0 bits ON).scaleToGray6Low
void scaleToGray6Low ( l_uint32 *datad, l_int32 wd, l_int32 hd, l_int32 wpld, l_uint32 *datas, l_int32 wpls, l_int32 *tab8, l_uint8 *valtab )
scaleToGray6Low()
Input: usual image variables
tab8 (made from makePixelSumTab8())
valtab (made from makeValTabSG6())
Return: 0 if OK; 1 on error
Each set of 4 6x6 bit-blocks in the source image, which
consist of 144 pixels arranged 24 pixels wide by 6 scanlines,
is converted to a row of 4 8-bit pixels in the dest image.
These 144 pixels of the input image are runs of 24 pixels
in six adjacent scanlines. Each run of 24 pixels is
stored in the 24 LSbits of a 32-bit word. We use 2 LUTs.
The first, tab8, takes 6 of these bits and stores
sum in one byte. This is done for each of the 6 scanlines,
and the results are added.
We now have the sum of ON pixels in the first 6x6 block. The
valtab LUT then converts these values (which go from 0 to 36) to
grayscale values between between 255 and 0. (See makeValTabSG6).
This process is repeated for each of the other 3 sets of
6x6 input pixels, giving 4 output pixels in total.
Note: because the input image is processed in groups of
24 x 6 pixels, the process clips the input height to
(h - h % 6) and the input width to (w - w % 24).scaleToGray8Low
void scaleToGray8Low ( l_uint32 *datad, l_int32 wd, l_int32 hd, l_int32 wpld, l_uint32 *datas, l_int32 wpls, l_int32 *tab8, l_uint8 *valtab )
scaleToGray8Low()
Input: usual image variables
tab8 (made from makePixelSumTab8())
valtab (made from makeValTabSG8())
Return: 0 if OK; 1 on error.
The output is processed one dest byte at a time,
corresponding to 8 rows of src bytes in the input image.
Two lookup tables are used. The first, tab8, gets the
sum of ON pixels in a byte. After sums from 8 rows have
been added, the second table, valtab, converts from this
value (which is between 0 and 64) to an 8 bpp grayscale
value between 0 (for all 64 bits ON) and 255 (for 0 bits ON).AUTHOR
Zakariyya Mughal <zmughal@cpan.org>
COPYRIGHT AND LICENSE
This software is copyright (c) 2014 by Zakariyya Mughal.
This is free software; you can redistribute it and/or modify it under the same terms as the Perl 5 programming language system itself.