sergiotarxz/libvips
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

vips_conva() done and working

Browse Source 
remove im_conva() nest, add tests
This commit is contained in:
John Cupitt 2016-08-03 18:29:50 +01:00
parent cedb904773
commit c658332215
8 changed files with 93 additions and 65 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

6
ChangeLog
View File

@ -27,10 +27,10 @@
- better quality for vips_resize() with linear/cubic kernels
- pyvips8 can create new metadata
- better upsizing with vips_resize()
- added vips_convf(), vips_convi(), vips_convasep() ... im_conv*() functions
rewritten as classes
- added vips_convf(), vips_convi(), vips_convasep(), vips_conva() ...
im_conv*() functions rewritten as classes
- vips_convsep() calls vips_convasep() for the approximate case
- new fixed-point vector path for convi is up to 2x faster
- new fixed-point vector path for convi is up to about 2x faster
- added vips_worley(), generate Worley noise
- added vips_perlin(), generate Perlin noise
- gif loader can write 1, 2, 3, or 4 bands depending on file contents

2
TODO
View File

@ -1,5 +1,7 @@
- tests for convasep ... just check it matches convsep +/- some small amount
- tests for conva
- try this blur.mat
6 1 3896

1
libvips/convolution/Makefile.am
View File

@ -6,6 +6,7 @@ libconvolution_la_SOURCES = \
correlation.c \
correlation.h \
conv.c \
conva.c \
convf.c \
convi.c \
convasep.c \

106
libvips/convolution/conva.c
View File

@ -72,12 +72,16 @@ $ vips im_max abs.v
- are we handling mask offset correctly?
- could we do better with an h and a v cumulativization image? we might
not need the huge intermediate we have now, since any line sum an be
found with simple indexing
*/
/*
*/
#define DEBUG
#define VIPS_DEBUG
*/
#ifdef HAVE_CONFIG_H
#include <config.h>
@ -93,10 +97,14 @@ $ vips im_max abs.v
#include <vips/vips.h>
#include <vips/vector.h>
#include <vips/debug.h>
#include <vips/internal.h>
/* Maximum number of boxes we can break the mask into.
#include "pconvolution.h"
/* Maximum number of boxes we can break the mask into. Don't have this too
* high, it'll make the instance huge, and gobject has a 64kb limit.
*/
#define MAX_LINES (10000)
#define MAX_LINES (1000)
/* The number of edges we consider at once in clustering. Higher values are
* faster, but risk pushing up average error in the result.
@ -167,8 +175,6 @@ typedef struct {
int area;
int rounding;
int offset;
int width;
int height;
/* The horizontal lines we gather. hline[3] writes to band 3 in the
* intermediate image. max_line is the length of the longest hline:
@ -259,7 +265,7 @@ vips_conva_hprint( VipsConva *conva )
conva->velement[y].factor,
conva->hline[b].weight );
for( x = 0; x < 45; x++ ) {
int rx = x * (conva->width + 1) / 45;
int rx = x * (conva->iM->Xsize + 1) / 45;
if( rx >= conva->hline[b].start &&
rx < conva->hline[b].end )
@ -300,6 +306,7 @@ vips_conva_decompose_lines( VipsConva *conva )
{
VipsImage *iM = conva->iM;
const int size = iM->Xsize * iM->Ysize;
double *coeff = VIPS_MATRIX( iM, 0, 0 );
double max;
double min;
@ -355,14 +362,14 @@ vips_conva_decompose_lines( VipsConva *conva )
inside = 0;
for( x = 0; x < iM->Xsize; x++ ) {
double coeff = *VIPS_MATRIX( iM, x, y );
double c = coeff[x + y * iM->Xsize];
/* The vertical line from mask[x, y] to 0 is
* inside. Is our current square (x, y) part
* of that line?
*/
if( (z_positive && coeff >= z_ph) ||
(!z_positive && coeff <= z_ph) ) {
if( (z_positive && c >= z_ph) ||
(!z_positive && c <= z_ph) ) {
if( !inside ) {
vips_conva_hline_start( conva,
x );
@ -382,7 +389,7 @@ vips_conva_decompose_lines( VipsConva *conva )
if( inside &&
vips_conva_hline_end( conva,
mask->xsize, y, z_positive ? 1 : -1 ) )
iM->Xsize, y, z_positive ? 1 : -1 ) )
return( -1 );
}
}
@ -480,7 +487,7 @@ vips_conva_cluster2( VipsConva *conva )
if( conva->hline[j].weight == 0 )
continue;
distance = vips_conva_distance( boxes, i, j );
distance = vips_conva_distance( conva, i, j );
if( distance < worst ) {
conva->edge[worst_i].a = i;
conva->edge[worst_i].b = j;
@ -672,7 +679,7 @@ vips_conva_decompose_boxes( VipsConva *conva )
{
VipsObjectClass *class = VIPS_OBJECT_GET_CLASS( conva );
VipsImage *iM = conva->iM;
double *coeff = VIPS_MATRIX( im, 0, 0 );
double *coeff = VIPS_MATRIX( iM, 0, 0 );
const int size = iM->Xsize * iM->Ysize;
double scale = vips_image_get_scale( iM );
double offset = vips_image_get_offset( iM );
@ -680,18 +687,15 @@ vips_conva_decompose_boxes( VipsConva *conva )
double sum;
int x, y, z;
boxes->n_hline = 0;
boxes->n_velement = 0;
boxes->n_vline = 0;
/* Break into a set of hlines.
*/
if( vips_conva_decompose_lines( boxes ) )
if( vips_conva_decompose_lines( conva ) )
return( -1 );
/* Cluster to find groups of lines.
*/
VIPS_DEBUG_MSG( "boxes_new: clustering with thresh %d ...\n", cluster );
VIPS_DEBUG_MSG( "vips_conva_decompose_boxes: "
"clustering with thresh %d ...\n", conva->cluster );
while( vips_conva_cluster2( conva ) )
;
@ -734,7 +738,7 @@ vips_conva_decompose_boxes( VipsConva *conva )
sum += coeff[z];
conva->area = VIPS_RINT( sum * conva->area / scale );
conva->rounding = (conva->area + 1) / 2 + offset * conva->area;
conva->rounding = (conva->area + 1) / 2;
conva->offset = offset;
#ifdef DEBUG
@ -795,9 +799,7 @@ vips_conva_start( VipsImage *out, void *a, void *b )
VipsConvaSeq *seq;
if( !(seq = VIPS_NEW( out, VipsConvaSeq )) )
return( NULL );
seq = VIPS_NEW( out, VipsConvaSeq );
seq->conva = conva;
seq->ir = vips_region_new( in );
@ -815,11 +817,6 @@ vips_conva_start( VipsImage *out, void *a, void *b )
seq->sum = VIPS_ARRAY( out, conva->n_velement, double );
seq->last_stride = -1;
if( !seq->ir || !seq->start || !seq->end || !seq->sum ) {
vips_conva_stop( seq, in, boxes );
return( NULL );
}
return( seq );
}
@ -863,7 +860,8 @@ G_STMT_START { \
/* Do horizontal masks ... we scan the mask along scanlines.
*/
static int
vips_conva_hgenerate( VipsRegion *or, void *vseq, void *a, void *b )
vips_conva_hgenerate( VipsRegion *or, void *vseq,
void *a, void *b, gboolean *stop )
{
VipsConvaSeq *seq = (VipsConvaSeq *) vseq;
VipsImage *in = (VipsImage *) a;
@ -888,7 +886,7 @@ vips_conva_hgenerate( VipsRegion *or, void *vseq, void *a, void *b )
* than the section of the output image we are producing.
*/
s = *r;
s.width += mask->xsize - 1;
s.width += iM->Xsize - 1;
if( vips_region_prepare( ir, &s ) )
return( -1 );
@ -967,7 +965,7 @@ vips_conva_hgenerate( VipsRegion *or, void *vseq, void *a, void *b )
static int
vips_conva_horizontal( VipsConva *conva, VipsImage *in, VipsImage **out )
{
VipsObjectClass *class = VIPS_OBJECT_GET_CLASS( convasep );
VipsObjectClass *class = VIPS_OBJECT_GET_CLASS( conva );
/* Prepare output. Consider a 7x7 mask and a 7x7 image --- the output
* would be 1x1.
@ -1083,11 +1081,13 @@ G_STMT_START { \
/* Do vertical masks ... we scan the mask down columns of pixels.
*/
static int
vips_conva_vgenerate( VipsRegion *or, void *vseq, void *a, void *b )
vips_conva_vgenerate( VipsRegion *or, void *vseq,
void *a, void *b, gboolean *stop )
{
VipsConvaSeq *seq = (VipsConvaSeq *) vseq;
VipsImage *in = (VipsImage *) a;
VipsConva *conva = (VipsConva *) b;
VipsConvolution *convolution = (VipsConvolution *) conva;
VipsRegion *ir = seq->ir;
const int n_vline = conva->n_vline;
@ -1108,14 +1108,14 @@ vips_conva_vgenerate( VipsRegion *or, void *vseq, void *a, void *b )
* than the section of the output image we are producing.
*/
s = *r;
s.height += mask->ysize - 1;
s.height += iM->Ysize - 1;
if( vips_region_prepare( ir, &s ) )
return( -1 );
istride = VIPS_REGION_LSKIP( ir ) /
VIPS_IMAGE_SIZEOF_ELEMENT( in );
ostride = VIPS_REGION_LSKIP( or ) /
VIPS_IMAGE_SIZEOF_ELEMENT( boxes->out );
VIPS_IMAGE_SIZEOF_ELEMENT( convolution->out );
/* Init offset array.
*/
@ -1130,7 +1130,7 @@ vips_conva_vgenerate( VipsRegion *or, void *vseq, void *a, void *b )
}
}
switch( conva->in->BandFmt ) {
switch( convolution->in->BandFmt ) {
case VIPS_FORMAT_UCHAR:
if( conva->max_line > 256 )
VCONV( unsigned int, \
@ -1143,10 +1143,10 @@ vips_conva_vgenerate( VipsRegion *or, void *vseq, void *a, void *b )
case VIPS_FORMAT_CHAR:
if( conva->max_line > 256 )
VCONV( signed int, \
signed int, signed char, CLIP_UCHAR );
signed int, signed char, CLIP_CHAR );
else
VCONV( signed int, \
signed short, signed char, CLIP_UCHAR );
signed short, signed char, CLIP_CHAR );
break;
case VIPS_FORMAT_USHORT:
@ -1192,7 +1192,8 @@ vips_conva_vgenerate( VipsRegion *or, void *vseq, void *a, void *b )
static int
vips_conva_vertical( VipsConva *conva, VipsImage *in, VipsImage **out )
{
VipsObjectClass *class = VIPS_OBJECT_GET_CLASS( convasep );
VipsObjectClass *class = VIPS_OBJECT_GET_CLASS( conva );
VipsConvolution *convolution = (VipsConvolution *) conva;
/* Prepare output. Consider a 7x7 mask and a 7x7 image --- the output
* would be 1x1.
@ -1208,12 +1209,12 @@ vips_conva_vertical( VipsConva *conva, VipsImage *in, VipsImage **out )
"%s", _( "image too small for mask" ) );
return( -1 );
}
out->Bands = boxes->in->Bands;
out->BandFmt = boxes->in->BandFmt;
(*out)->Bands = convolution->in->Bands;
(*out)->BandFmt = convolution->in->BandFmt;
if( vips_image_generate( out,
if( vips_image_generate( *out,
vips_conva_start, vips_conva_vgenerate, vips_conva_stop,
in, boxes ) )
in, conva ) )
return( -1 );
return( 0 );
@ -1222,7 +1223,6 @@ vips_conva_vertical( VipsConva *conva, VipsImage *in, VipsImage **out )
static int
vips_conva_build( VipsObject *object )
{
VipsObjectClass *class = VIPS_OBJECT_GET_CLASS( object );
VipsConvolution *convolution = (VipsConvolution *) object;
VipsConva *conva = (VipsConva *) object;
VipsImage **t = (VipsImage **) vips_object_local_array( object, 4 );
@ -1237,8 +1237,6 @@ vips_conva_build( VipsObject *object )
if( vips__image_intize( convolution->M, &t[0] ) )
return( -1 );
conva->iM = t[0];
conva->width = conva->iM->Xsize;
conva->height = conva->iM->Ysize;
in = convolution->in;
@ -1248,14 +1246,14 @@ vips_conva_build( VipsObject *object )
g_object_set( conva, "out", vips_image_new(), NULL );
if(
vips_embed( in, &t[1],
conva->width / 2,
conva->height / 2,
in->Xsize + convasep->width - 1,
in->Ysize + convasep->height - 1,
t[0]->Xsize / 2,
t[0]->Ysize / 2,
in->Xsize + t[0]->Xsize - 1,
in->Ysize + t[0]->Ysize - 1,
"extend", VIPS_EXTEND_COPY,
NULL ) ||
vips_conva_horizontal( convasep, t[1], &t[2] ) ||
vips_conva_vertical( convasep, t[2], &t[3] ) ||
vips_conva_horizontal( conva, t[1], &t[2] ) ||
vips_conva_vertical( conva, t[2], &t[3] ) ||
vips_image_write( t[3], convolution->out ) )
return( -1 );
@ -1275,7 +1273,7 @@ vips_conva_class_init( VipsConvaClass *class )
gobject_class->get_property = vips_object_get_property;
object_class->nickname = "conva";
object_class->description = _( "approximate convolution" );
object_class->description = _( "approximate integer convolution" );
object_class->build = vips_conva_build;
VIPS_ARG_INT( class, "layers", 104,
@ -1289,7 +1287,7 @@ vips_conva_class_init( VipsConvaClass *class )
_( "Cluster" ),
_( "Cluster lines closer than this in approximation" ),
VIPS_ARGUMENT_OPTIONAL_INPUT,
G_STRUCT_OFFSET( VipsConv, cluster ),
G_STRUCT_OFFSET( VipsConva, cluster ),
1, 100, 1 );
}
@ -1313,12 +1311,14 @@ vips_conva_init( VipsConva *conva )
* * @layers: %gint, number of layers for approximation
* * @cluster: %gint, cluster lines closer than this distance
*
* Perform an approximate convolution of @in with @mask.
* Perform an approximate integer convolution of @in with @mask.
* This is a low-level operation, see
* vips_conv() for something more convenient.
*
* The output image
* always has the same #VipsBandFormat as the input image.
* Elements of @mask are converted to
* integers before convolution.
*
* Larger values for @layers give more accurate
* results, but are slower. As @layers approaches the mask radius, the

11
libvips/convolution/convasep.c
View File

@ -61,12 +61,16 @@
we could then use orc to write a bit of code to implement this set
of lines
stackoverflow has an algorithm for cumulativization using SIMD and
threads, see that font rasterization with rust piece on medium by
ralph levien
*/
/*
*/
#define DEBUG
#define VIPS_DEBUG
*/
#ifdef HAVE_CONFIG_H
#include <config.h>
@ -883,7 +887,8 @@ vips_convasep_class_init( VipsConvasepClass *class )
gobject_class->get_property = vips_object_get_property;
object_class->nickname = "convasep";
object_class->description = _( "approximate separable convolution" );
object_class->description =
_( "approximate separable integer convolution" );
object_class->build = vips_convasep_build;
VIPS_ARG_INT( class, "layers", 104,
@ -913,7 +918,7 @@ vips_convasep_init( VipsConvasep *convasep )
*
* * @layers: %gint, number of layers for approximation
*
* Approximate separable convolution. This is a low-level operation, see
* Approximate separable integer convolution. This is a low-level operation, see
* vips_convsep() for something more convenient.
*
* The image is convolved twice: once with @mask and then again with @mask

2
libvips/convolution/convolution.c
View File

@ -157,6 +157,7 @@ void
vips_convolution_operation_init( void )
{
extern int vips_conv_get_type( void );
extern int vips_conva_get_type( void );
extern int vips_convf_get_type( void );
extern int vips_convi_get_type( void );
extern int vips_convsep_get_type( void );
@ -168,6 +169,7 @@ vips_convolution_operation_init( void )
extern int vips_gaussblur_get_type( void );
vips_conv_get_type();
vips_conva_get_type();
vips_convf_get_type();
vips_convi_get_type();
vips_compass_get_type();

6
libvips/include/vips/convolution.h
View File

@ -44,15 +44,17 @@ typedef enum {
VIPS_COMBINE_LAST
} VipsCombine;
int vips_conv( VipsImage *in, VipsImage **out, VipsImage *mask, ... )
__attribute__((sentinel));
int vips_convf( VipsImage *in, VipsImage **out, VipsImage *mask, ... )
__attribute__((sentinel));
int vips_convi( VipsImage *in, VipsImage **out, VipsImage *mask, ... )
__attribute__((sentinel));
int vips_convasep( VipsImage *in, VipsImage **out, VipsImage *mask, ... )
int vips_conva( VipsImage *in, VipsImage **out, VipsImage *mask, ... )
__attribute__((sentinel));
int vips_convsep( VipsImage *in, VipsImage **out, VipsImage *mask, ... )
__attribute__((sentinel));
int vips_conv( VipsImage *in, VipsImage **out, VipsImage *mask, ... )
int vips_convasep( VipsImage *in, VipsImage **out, VipsImage *mask, ... )
__attribute__((sentinel));
int vips_compass( VipsImage *in, VipsImage **out, VipsImage *mask, ... )

24
libvips/iofuncs/memory.c
View File

@ -109,7 +109,15 @@ static GMutex *vips_tracked_mutex = NULL;
* @OBJ: allocate memory local to @OBJ, or %NULL for no auto-free
* @T: type of thing to allocate
*
* Returns: A pointer of type @T *, or %NULL on error.
* Allocate memory for a thing of type @T. The memory is not
* cleared.
*
* This macro cannot fail. See vips_tracked_malloc() if you are
* allocating large amounts of memory.
*
* See also: vips_malloc().
*
* Returns: A pointer of type @T *.
*/
/**
@ -118,7 +126,15 @@ static GMutex *vips_tracked_mutex = NULL;
* @N: number of @T 's to allocate
* @T: type of thing to allocate
*
* Returns: A pointer of type @T *, or %NULL on error.
* Allocate memory for an array of objects of type @T. The memory is not
* cleared.
*
* This macro cannot fail. See vips_tracked_malloc() if you are
* allocating large amounts of memory.
*
* See also: vips_malloc().
*
* Returns: A pointer of type @T *.
*/
static void
@ -141,7 +157,7 @@ vips_malloc_cb( VipsObject *object, char *buf )
*
* See also: vips_tracked_malloc().
*
* Returns: (transfer full): a pointer to the allocated memory
* Returns: (transfer full): a pointer to the allocated memory.
*/
void *
vips_malloc( VipsObject *object, size_t size )
@ -166,7 +182,7 @@ vips_malloc( VipsObject *object, size_t size )
*
* g_strdup() a string. When @object is freed, the string will be freed for
* you. If @object is %NULL, you need to
* free the memory explicitly with g_free().
* free the memory yourself with g_free().
*
* This function cannot fail.
*