code.ivysaur.me/imagequant
1
0
Fork 0
Code Issues Pull Requests Releases 2 Wiki Activity

update bundled pngquant to v2.12.2

Browse Source
This commit is contained in:
mappu 2018-12-31 16:35:00 +13:00
parent 4d88cab911
commit 7c7d9a661c
9 changed files with 424 additions and 311 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

22
kmeans.c
View File

@ -1,20 +1,6 @@
/*
© 2011-2016 by Kornel Lesiński.
This file is part of libimagequant.
libimagequant is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation, either version 3 of the License, or
(at your option) any later version.
libimagequant is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License
along with libimagequant. If not, see <http://www.gnu.org/licenses/>.
** © 2011-2016 by Kornel Lesiński.
** See COPYRIGHT file for license.
*/
#include "libimagequant.h"
@ -80,14 +66,14 @@ LIQ_PRIVATE void kmeans_finalize(colormap *map, const unsigned int max_threads,
LIQ_PRIVATE double kmeans_do_iteration(histogram *hist, colormap *const map, kmeans_callback callback)
{
const unsigned int max_threads = omp_get_max_threads();
kmeans_state average_color[(KMEANS_CACHE_LINE_GAP+map->colors) * max_threads];
LIQ_ARRAY(kmeans_state, average_color, (KMEANS_CACHE_LINE_GAP+map->colors) * max_threads);
kmeans_init(map, max_threads, average_color);
struct nearest_map *const n = nearest_init(map);
hist_item *const achv = hist->achv;
const int hist_size = hist->size;
double total_diff=0;
#pragma omp parallel for if (hist_size > 3000) \
#pragma omp parallel for if (hist_size > 2000) \
schedule(static) default(none) shared(average_color,callback) reduction(+:total_diff)
for(int j=0; j < hist_size; j++) {
float diff;

393
libimagequant.c
View File

@ -1,31 +1,9 @@
/*
** © 2009-2016 by Kornel Lesiński.
** © 2009-2018 by Kornel Lesiński.
** © 1989, 1991 by Jef Poskanzer.
** © 1997, 2000, 2002 by Greg Roelofs; based on an idea by Stefan Schneider.
**
** This file is part of libimagequant.
**
** libimagequant is free software: you can redistribute it and/or modify
** it under the terms of the GNU General Public License as published by
** the Free Software Foundation, either version 3 of the License, or
** (at your option) any later version.
**
** libimagequant is distributed in the hope that it will be useful,
** but WITHOUT ANY WARRANTY; without even the implied warranty of
** MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
** GNU General Public License for more details.
**
** You should have received a copy of the GNU General Public License
** along with libimagequant. If not, see <http://www.gnu.org/licenses/>.
*/
/* Copyright (C) 1989, 1991 by Jef Poskanzer.
** Copyright (C) 1997, 2000, 2002 by Greg Roelofs; based on an idea by
** Stefan Schneider.
**
** Permission to use, copy, modify, and distribute this software and its
** documentation for any purpose and without fee is hereby granted, provided
** that the above copyright notice appear in all copies and that both that
** copyright notice and this permission notice appear in supporting
** documentation. This software is provided "as is" without express or
** implied warranty.
** See COPYRIGHT file for license.
*/
#include <stdio.h>
@ -43,7 +21,9 @@
#ifdef _OPENMP
#include <omp.h>
#define LIQ_TEMP_ROW_WIDTH(img_width) (((img_width) | 15) + 1) /* keep alignment & leave space between rows to avoid cache line contention */
#else
#define LIQ_TEMP_ROW_WIDTH(img_width) (img_width)
#define omp_get_max_threads() 1
#define omp_get_thread_num() 0
#endif
@ -78,7 +58,8 @@ struct liq_attr {
unsigned int max_colors, max_histogram_entries;
unsigned int min_posterization_output /* user setting */, min_posterization_input /* speed setting */;
unsigned int kmeans_iterations, feedback_loop_trials;
bool last_index_transparent, use_contrast_maps, use_dither_map;
bool last_index_transparent, use_contrast_maps;
unsigned char use_dither_map;
unsigned char speed;
unsigned char progress_stage1, progress_stage2, progress_stage3;
@ -100,11 +81,12 @@ struct liq_image {
rgba_pixel **rows;
double gamma;
unsigned int width, height;
unsigned char *noise, *edges, *dither_map;
unsigned char *importance_map, *edges, *dither_map;
rgba_pixel *pixels, *temp_row;
f_pixel *temp_f_row;
liq_image_get_rgba_row_callback *row_callback;
void *row_callback_user_info;
liq_image *background;
float min_opaque_val;
f_pixel fixed_colors[256];
unsigned short fixed_colors_count;
@ -124,7 +106,8 @@ typedef struct liq_remapping_result {
liq_palette int_palette;
double gamma, palette_error;
float dither_level;
bool use_dither_map; unsigned char progress_stage1;
unsigned char use_dither_map;
unsigned char progress_stage1;
} liq_remapping_result;
struct liq_result {
@ -141,7 +124,7 @@ struct liq_result {
float dither_level;
double gamma, palette_error;
int min_posterization_output;
bool use_dither_map;
unsigned char use_dither_map;
};
struct liq_histogram {
@ -161,6 +144,7 @@ static void modify_alpha(liq_image *input_image, rgba_pixel *const row_pixels) L
static void contrast_maps(liq_image *image) LIQ_NONNULL;
static liq_error finalize_histogram(liq_histogram *input_hist, liq_attr *options, histogram **hist_output) LIQ_NONNULL;
static const rgba_pixel *liq_image_get_row_rgba(liq_image *input_image, unsigned int row) LIQ_NONNULL;
static bool liq_image_get_row_f_init(liq_image *img) LIQ_NONNULL;
static const f_pixel *liq_image_get_row_f(liq_image *input_image, unsigned int row) LIQ_NONNULL;
static void liq_remapping_result_destroy(liq_remapping_result *result) LIQ_NONNULL;
static liq_error pngquant_quantize(histogram *hist, const liq_attr *options, const int fixed_colors_count, const f_pixel fixed_colors[], const double gamma, bool fixed_result_colors, liq_result **) LIQ_NONNULL;
@ -174,7 +158,7 @@ LIQ_NONNULL static void liq_verbose_printf(const liq_attr *context, const char *
int required_space = vsnprintf(NULL, 0, fmt, va)+1; // +\0
va_end(va);
char buf[required_space];
LIQ_ARRAY(char, buf, required_space);
va_start(va, fmt);
vsnprintf(buf, required_space, fmt, va);
va_end(va);
@ -210,8 +194,13 @@ LIQ_NONNULL static bool liq_remap_progress(const liq_remapping_result *quant, co
#if USE_SSE
inline static bool is_sse_available()
{
#if (defined(__x86_64__) || defined(__amd64))
#if (defined(__x86_64__) || defined(__amd64) || defined(_WIN64))
return true;
#elif _MSC_VER
int info[4];
__cpuid(info, 1);
/* bool is implemented as a built-in type of size 1 in MSVC */
return info[3] & (1<<26) ? true : false;
#else
int a,b,c,d;
cpuid(1, a, b, c, d);
@ -348,7 +337,8 @@ LIQ_EXPORT LIQ_NONNULL liq_error liq_set_speed(liq_attr* attr, int speed)
if (!CHECK_STRUCT_TYPE(attr, liq_attr)) return LIQ_INVALID_POINTER;
if (speed < 1 || speed > 10) return LIQ_VALUE_OUT_OF_RANGE;
unsigned int iterations = MAX(8-speed, 0); iterations += iterations * iterations/2;
unsigned int iterations = MAX(8-speed, 0);
iterations += iterations * iterations/2;
attr->kmeans_iterations = iterations;
attr->kmeans_iteration_limit = 1.0/(double)(1<<(23-speed));
attr->feedback_loop_trials = MAX(56-9*speed, 0);
@ -356,11 +346,16 @@ LIQ_EXPORT LIQ_NONNULL liq_error liq_set_speed(liq_attr* attr, int speed)
attr->max_histogram_entries = (1<<17) + (1<<18)*(10-speed);
attr->min_posterization_input = (speed >= 8) ? 1 : 0;
attr->use_dither_map = (speed <= (omp_get_max_threads() > 1 ? 7 : 5)); // parallelized dither map might speed up floyd remapping
if (attr->use_dither_map && speed < 3) {
attr->use_dither_map = 2; // always
}
attr->use_contrast_maps = (speed <= 7) || attr->use_dither_map;
attr->speed = speed;
attr->progress_stage1 = attr->use_contrast_maps ? 20 : 8;
if (attr->feedback_loop_trials < 2) attr->progress_stage1 += 30;
if (attr->feedback_loop_trials < 2) {
attr->progress_stage1 += 30;
}
attr->progress_stage3 = 50 / (1+speed);
attr->progress_stage2 = 100 - attr->progress_stage1 - attr->progress_stage3;
return LIQ_OK;
@ -400,7 +395,7 @@ LIQ_EXPORT LIQ_NONNULL int liq_get_min_opacity(const liq_attr *attr)
{
if (!CHECK_STRUCT_TYPE(attr, liq_attr)) return -1;
return MIN(255, 256.0 * attr->min_opaque_val);
return MIN(255.f, 256.f * attr->min_opaque_val);
}
LIQ_EXPORT LIQ_NONNULL void liq_set_last_index_transparent(liq_attr* attr, int is_last)
@ -520,7 +515,7 @@ LIQ_EXPORT liq_attr* liq_attr_create_with_allocator(void* (*custom_malloc)(size_
.target_mse = 0,
.max_mse = MAX_DIFF,
};
liq_set_speed(attr, 3);
liq_set_speed(attr, 4);
return attr;
}
@ -540,7 +535,7 @@ LIQ_EXPORT LIQ_NONNULL liq_error liq_image_add_fixed_color(liq_image *img, liq_c
return LIQ_OK;
}
LIQ_NONNULL static liq_error liq_histogram_add_fixed_color_internal(liq_histogram *hist, f_pixel color)
LIQ_NONNULL static liq_error liq_histogram_add_fixed_color_f(liq_histogram *hist, f_pixel color)
{
if (hist->fixed_colors_count > 255) return LIQ_UNSUPPORTED;
@ -548,9 +543,24 @@ LIQ_NONNULL static liq_error liq_histogram_add_fixed_color_internal(liq_histogra
return LIQ_OK;
}
LIQ_EXPORT LIQ_NONNULL liq_error liq_histogram_add_fixed_color(liq_histogram *hist, liq_color color, double gamma)
{
if (!CHECK_STRUCT_TYPE(hist, liq_histogram)) return LIQ_INVALID_POINTER;
float gamma_lut[256];
to_f_set_gamma(gamma_lut, gamma ? gamma : 0.45455);
const f_pixel px = rgba_to_f(gamma_lut, (rgba_pixel){
.r = color.r,
.g = color.g,
.b = color.b,
.a = color.a,
});
return liq_histogram_add_fixed_color_f(hist, px);
}
LIQ_NONNULL static bool liq_image_use_low_memory(liq_image *img)
{
img->temp_f_row = img->malloc(sizeof(img->f_pixels[0]) * img->width * omp_get_max_threads());
img->temp_f_row = img->malloc(sizeof(img->f_pixels[0]) * LIQ_TEMP_ROW_WIDTH(img->width) * omp_get_max_threads());
return img->temp_f_row != NULL;
}
@ -586,7 +596,7 @@ static liq_image *liq_image_create_internal(const liq_attr *attr, rgba_pixel* ro
};
if (!rows || attr->min_opaque_val < 1.f) {
img->temp_row = attr->malloc(sizeof(img->temp_row[0]) * width * omp_get_max_threads());
img->temp_row = attr->malloc(sizeof(img->temp_row[0]) * LIQ_TEMP_ROW_WIDTH(width) * omp_get_max_threads());
if (!img->temp_row) return NULL;
}
@ -630,6 +640,57 @@ LIQ_EXPORT LIQ_NONNULL liq_error liq_image_set_memory_ownership(liq_image *img,
return LIQ_OK;
}
LIQ_NONNULL static void liq_image_free_maps(liq_image *input_image);
LIQ_NONNULL static void liq_image_free_importance_map(liq_image *input_image);
LIQ_EXPORT LIQ_NONNULL liq_error liq_image_set_importance_map(liq_image *img, unsigned char importance_map[], size_t buffer_size, enum liq_ownership ownership) {
if (!CHECK_STRUCT_TYPE(img, liq_image)) return LIQ_INVALID_POINTER;
if (!CHECK_USER_POINTER(importance_map)) return LIQ_INVALID_POINTER;
const size_t required_size = img->width * img->height;
if (buffer_size < required_size) {
return LIQ_BUFFER_TOO_SMALL;
}
if (ownership == LIQ_COPY_PIXELS) {
unsigned char *tmp = img->malloc(required_size);
if (!tmp) {
return LIQ_OUT_OF_MEMORY;
}
memcpy(tmp, importance_map, required_size);
importance_map = tmp;
} else if (ownership != LIQ_OWN_PIXELS) {
return LIQ_UNSUPPORTED;
}
liq_image_free_importance_map(img);
img->importance_map = importance_map;
return LIQ_OK;
}
LIQ_EXPORT LIQ_NONNULL liq_error liq_image_set_background(liq_image *img, liq_image *background)
{
if (!CHECK_STRUCT_TYPE(img, liq_image)) return LIQ_INVALID_POINTER;
if (!CHECK_STRUCT_TYPE(background, liq_image)) return LIQ_INVALID_POINTER;
if (background->background) {
return LIQ_UNSUPPORTED;
}
if (img->width != background->width || img->height != background->height) {
return LIQ_BUFFER_TOO_SMALL;
}
if (img->background) {
liq_image_destroy(img->background);
}
img->background = background;
liq_image_free_maps(img); // Force them to be re-analyzed with the background
return LIQ_OK;
}
LIQ_NONNULL static bool check_image_size(const liq_attr *attr, const int width, const int height)
{
if (!CHECK_STRUCT_TYPE(attr, liq_attr)) {
@ -730,7 +791,7 @@ LIQ_NONNULL static const rgba_pixel *liq_image_get_row_rgba(liq_image *img, unsi
}
assert(img->temp_row);
rgba_pixel *temp_row = img->temp_row + img->width * omp_get_thread_num();
rgba_pixel *temp_row = img->temp_row + LIQ_TEMP_ROW_WIDTH(img->width) * omp_get_thread_num();
if (img->rows) {
memcpy(temp_row, img->rows[row], img->width * sizeof(temp_row[0]));
} else {
@ -753,31 +814,40 @@ LIQ_NONNULL static void convert_row_to_f(liq_image *img, f_pixel *row_f_pixels,
}
}
LIQ_NONNULL static bool liq_image_get_row_f_init(liq_image *img)
{
assert(omp_get_thread_num() == 0);
if (img->f_pixels) {
return true;
}
if (!liq_image_should_use_low_memory(img, false)) {
img->f_pixels = img->malloc(sizeof(img->f_pixels[0]) * img->width * img->height);
}
if (!img->f_pixels) {
return liq_image_use_low_memory(img);
}
if (!liq_image_has_rgba_pixels(img)) {
return false;
}
float gamma_lut[256];
to_f_set_gamma(gamma_lut, img->gamma);
for(unsigned int i=0; i < img->height; i++) {
convert_row_to_f(img, &img->f_pixels[i*img->width], i, gamma_lut);
}
return true;
}
LIQ_NONNULL static const f_pixel *liq_image_get_row_f(liq_image *img, unsigned int row)
{
if (!img->f_pixels) {
if (img->temp_f_row) {
float gamma_lut[256];
to_f_set_gamma(gamma_lut, img->gamma);
f_pixel *row_for_thread = img->temp_f_row + img->width * omp_get_thread_num();
convert_row_to_f(img, row_for_thread, row, gamma_lut);
return row_for_thread;
}
assert(omp_get_thread_num() == 0);
if (!liq_image_should_use_low_memory(img, false)) {
img->f_pixels = img->malloc(sizeof(img->f_pixels[0]) * img->width * img->height);
}
if (!img->f_pixels) {
if (!liq_image_use_low_memory(img)) return NULL;
return liq_image_get_row_f(img, row);
}
assert(img->temp_f_row); // init should have done that
float gamma_lut[256];
to_f_set_gamma(gamma_lut, img->gamma);
for(unsigned int i=0; i < img->height; i++) {
convert_row_to_f(img, &img->f_pixels[i*img->width], i, gamma_lut);
}
f_pixel *row_for_thread = img->temp_f_row + LIQ_TEMP_ROW_WIDTH(img->width) * omp_get_thread_num();
convert_row_to_f(img, row_for_thread, row, gamma_lut);
return row_for_thread;
}
return img->f_pixels + img->width * row;
}
@ -818,23 +888,34 @@ LIQ_NONNULL static void liq_image_free_rgba_source(liq_image *input_image)
}
}
LIQ_NONNULL static void liq_image_free_importance_map(liq_image *input_image) {
if (input_image->importance_map) {
input_image->free(input_image->importance_map);
input_image->importance_map = NULL;
}
}
LIQ_NONNULL static void liq_image_free_maps(liq_image *input_image) {
liq_image_free_importance_map(input_image);
if (input_image->edges) {
input_image->free(input_image->edges);
input_image->edges = NULL;
}
if (input_image->dither_map) {
input_image->free(input_image->dither_map);
input_image->dither_map = NULL;
}
}
LIQ_EXPORT LIQ_NONNULL void liq_image_destroy(liq_image *input_image)
{
if (!CHECK_STRUCT_TYPE(input_image, liq_image)) return;
liq_image_free_rgba_source(input_image);
if (input_image->noise) {
input_image->free(input_image->noise);
}
if (input_image->edges) {
input_image->free(input_image->edges);
}
if (input_image->dither_map) {
input_image->free(input_image->dither_map);
}
liq_image_free_maps(input_image);
if (input_image->f_pixels) {
input_image->free(input_image->f_pixels);
@ -848,6 +929,10 @@ LIQ_EXPORT LIQ_NONNULL void liq_image_destroy(liq_image *input_image)
input_image->free(input_image->temp_f_row);
}
if (input_image->background) {
liq_image_destroy(input_image->background);
}
input_image->magic_header = liq_freed_magic;
input_image->free(input_image);
}
@ -1077,7 +1162,7 @@ LIQ_NONNULL static void sort_palette(colormap *map, const liq_attr *options)
*/
if (options->last_index_transparent) {
for(unsigned int i=0; i < map->colors; i++) {
if (map->palette[i].acolor.a < 1.0/256.0) {
if (map->palette[i].acolor.a < 1.f/256.f) {
const unsigned int old = i, transparent_dest = map->colors-1;
SWAP_PALETTE(map, transparent_dest, old);
@ -1100,7 +1185,7 @@ LIQ_NONNULL static void sort_palette(colormap *map, const liq_attr *options)
/* move transparent colors to the beginning to shrink trns chunk */
unsigned int num_transparent = 0;
for(unsigned int i = 0; i < non_fixed_colors; i++) {
if (map->palette[i].acolor.a < 255.0/256.0) {
if (map->palette[i].acolor.a < 255.f/256.f) {
// current transparent color is swapped with earlier opaque one
if (i != num_transparent) {
SWAP_PALETTE(map, num_transparent, i);
@ -1174,24 +1259,37 @@ LIQ_NONNULL static float remap_to_palette(liq_image *const input_image, unsigned
const unsigned int cols = input_image->width;
double remapping_error=0;
if (!liq_image_get_row_f(input_image, 0)) { // trigger lazy conversion
if (!liq_image_get_row_f_init(input_image)) {
return -1;
}
if (input_image->background && !liq_image_get_row_f_init(input_image->background)) {
return -1;
}
const colormap_item *acolormap = map->palette;
struct nearest_map *const n = nearest_init(map);
const int transparent_index = input_image->background ? nearest_search(n, &(f_pixel){0,0,0,0}, 0, NULL) : 0;
const unsigned int max_threads = omp_get_max_threads();
kmeans_state average_color[(KMEANS_CACHE_LINE_GAP+map->colors) * max_threads];
LIQ_ARRAY(kmeans_state, average_color, (KMEANS_CACHE_LINE_GAP+map->colors) * max_threads);
kmeans_init(map, max_threads, average_color);
#pragma omp parallel for if (rows*cols > 3000) \
schedule(static) default(none) shared(average_color) reduction(+:remapping_error)
schedule(static) default(none) shared(acolormap) shared(average_color) reduction(+:remapping_error)
for(int row = 0; row < rows; ++row) {
const f_pixel *const row_pixels = liq_image_get_row_f(input_image, row);
const f_pixel *const bg_pixels = input_image->background && acolormap[transparent_index].acolor.a < 1.f/256.f ? liq_image_get_row_f(input_image->background, row) : NULL;
unsigned int last_match=0;
for(unsigned int col = 0; col < cols; ++col) {
float diff;
output_pixels[row][col] = last_match = nearest_search(n, &row_pixels[col], last_match, &diff);
last_match = nearest_search(n, &row_pixels[col], last_match, &diff);
if (bg_pixels && colordifference(bg_pixels[col], acolormap[last_match].acolor) <= diff) {
last_match = transparent_index;
}
output_pixels[row][col] = last_match;
remapping_error += diff;
kmeans_update_color(row_pixels[col], 1.0, map, last_match, omp_get_thread_num(), average_color);
@ -1226,25 +1324,25 @@ inline static f_pixel get_dithered_pixel(const float dither_level, const float m
else { if (px.b + sb < max_underflow) ratio = MIN(ratio, (max_underflow-px.b)/sb); }
float a = px.a + sa;
if (a > 1.0) { a = 1.0; }
if (a > 1.f) { a = 1.f; }
else if (a < 0) { a = 0; }
// If dithering error is crazy high, don't propagate it that much
// This prevents crazy geen pixels popping out of the blue (or red or black! ;)
const float dither_error = sr*sr + sg*sg + sb*sb + sa*sa;
if (dither_error > max_dither_error) {
ratio *= 0.8;
ratio *= 0.8f;
} else if (dither_error < 2.f/256.f/256.f) {
// don't dither areas that don't have noticeable error — makes file smaller
return px;
}
}
return (f_pixel){
.r=px.r + sr * ratio,
.g=px.g + sg * ratio,
.b=px.b + sb * ratio,
.a=a,
};
return (f_pixel) {
.r=px.r + sr * ratio,
.g=px.g + sg * ratio,
.b=px.b + sb * ratio,
.a=a,
};
}
/**
@ -1260,25 +1358,32 @@ LIQ_NONNULL static bool remap_to_palette_floyd(liq_image *input_image, unsigned
const colormap *map = quant->palette;
const colormap_item *acolormap = map->palette;
if (!liq_image_get_row_f_init(input_image)) {
return false;
}
if (input_image->background && !liq_image_get_row_f_init(input_image->background)) {
return false;
}
/* Initialize Floyd-Steinberg error vectors. */
f_pixel *restrict thiserr, *restrict nexterr;
const size_t errsize = (cols + 2) * sizeof(*thiserr) * 2;
thiserr = input_image->malloc(errsize); // +2 saves from checking out of bounds access
const size_t errwidth = cols+2;
f_pixel *restrict thiserr = input_image->malloc(errwidth * sizeof(thiserr[0]) * 2); // +2 saves from checking out of bounds access
if (!thiserr) return false;
memset(thiserr, 0, errsize);
nexterr = thiserr + (cols + 2);
f_pixel *restrict nexterr = thiserr + errwidth;
memset(thiserr, 0, errwidth * sizeof(thiserr[0]));
bool ok = true;
struct nearest_map *const n = nearest_init(map);
const int transparent_index = input_image->background ? nearest_search(n, &(f_pixel){0,0,0,0}, 0, NULL) : 0;
// response to this value is non-linear and without it any value < 0.8 would give almost no dithering
float base_dithering_level = quant->dither_level;
base_dithering_level = 1.0 - (1.0-base_dithering_level)*(1.0-base_dithering_level);
base_dithering_level = 1.f - (1.f-base_dithering_level)*(1.f-base_dithering_level);
if (dither_map) {
base_dithering_level *= 1.0/255.0; // convert byte to float
base_dithering_level *= 1.f/255.f; // convert byte to float
}
base_dithering_level *= 15.0/16.0; // prevent small errors from accumulating
base_dithering_level *= 15.f/16.f; // prevent small errors from accumulating
int fs_direction = 1;
unsigned int last_match=0;
@ -1288,10 +1393,11 @@ LIQ_NONNULL static bool remap_to_palette_floyd(liq_image *input_image, unsigned
break;
}
memset(nexterr, 0, (cols + 2) * sizeof(*nexterr));
memset(nexterr, 0, errwidth * sizeof(nexterr[0]));
int col = (fs_direction > 0) ? 0 : (cols - 1);
const f_pixel *const row_pixels = liq_image_get_row_f(input_image, row);
const f_pixel *const bg_pixels = input_image->background && acolormap[transparent_index].acolor.a < 1.f/256.f ? liq_image_get_row_f(input_image->background, row) : NULL;
do {
float dither_level = base_dithering_level;
@ -1302,9 +1408,16 @@ LIQ_NONNULL static bool remap_to_palette_floyd(liq_image *input_image, unsigned
const f_pixel spx = get_dithered_pixel(dither_level, max_dither_error, thiserr[col + 1], row_pixels[col]);
const unsigned int guessed_match = output_image_is_remapped ? output_pixels[row][col] : last_match;
output_pixels[row][col] = last_match = nearest_search(n, &spx, guessed_match, NULL);
float diff;
last_match = nearest_search(n, &spx, guessed_match, &diff);
f_pixel output_px = acolormap[last_match].acolor;
if (bg_pixels && colordifference(bg_pixels[col], output_px) <= diff) {
output_px = bg_pixels[col];
output_pixels[row][col] = transparent_index;
} else {
output_pixels[row][col] = last_match;
}
const f_pixel output_px = acolormap[last_match].acolor;
f_pixel err = {
.r = (spx.r - output_px.r),
.g = (spx.g - output_px.g),
@ -1315,10 +1428,10 @@ LIQ_NONNULL static bool remap_to_palette_floyd(liq_image *input_image, unsigned
// If dithering error is crazy high, don't propagate it that much
// This prevents crazy geen pixels popping out of the blue (or red or black! ;)
if (err.r*err.r + err.g*err.g + err.b*err.b + err.a*err.a > max_dither_error) {
err.r *= 0.75;
err.g *= 0.75;
err.b *= 0.75;
err.a *= 0.75;
err.r *= 0.75f;
err.g *= 0.75f;
err.b *= 0.75f;
err.a *= 0.75f;
}
/* Propagate Floyd-Steinberg error terms. */
@ -1370,7 +1483,7 @@ LIQ_NONNULL static bool remap_to_palette_floyd(liq_image *input_image, unsigned
if (fs_direction > 0) {
if (col >= cols) break;
} else {
if (col <= 0) break;
if (col < 0) break;
}
} while(1);
@ -1389,7 +1502,7 @@ LIQ_NONNULL static bool remap_to_palette_floyd(liq_image *input_image, unsigned
/* fixed colors are always included in the palette, so it would be wasteful to duplicate them in palette from histogram */
LIQ_NONNULL static void remove_fixed_colors_from_histogram(histogram *hist, const int fixed_colors_count, const f_pixel fixed_colors[], const float target_mse)
{
const float max_difference = MAX(target_mse/2.0, 2.0/256.0/256.0);
const float max_difference = MAX(target_mse/2.f, 2.f/256.f/256.f);
if (fixed_colors_count) {
for(int j=0; j < hist->size; j++) {
for(unsigned int i=0; i < fixed_colors_count; i++) {
@ -1461,26 +1574,28 @@ LIQ_EXPORT LIQ_NONNULL liq_error liq_histogram_add_image(liq_histogram *input_hi
const unsigned int cols = input_image->width, rows = input_image->height;
if (!input_image->noise && options->use_contrast_maps) {
if (!input_image->importance_map && options->use_contrast_maps) {
contrast_maps(input_image);
}
input_hist->gamma = input_image->gamma;
for(int i = 0; i < input_image->fixed_colors_count; i++) {
liq_error res = liq_histogram_add_fixed_color_internal(input_hist, input_image->fixed_colors[i]);
liq_error res = liq_histogram_add_fixed_color_f(input_hist, input_image->fixed_colors[i]);
if (res != LIQ_OK) {
return res;
}
}
/*
** Step 2: attempt to make a histogram of the colors, unclustered.
** If at first we don't succeed, increase ignorebits to increase color
** coherence and try again.
*/
/*
** Step 2: attempt to make a histogram of the colors, unclustered.
** If at first we don't succeed, increase ignorebits to increase color
** coherence and try again.
*/
if (liq_progress(options, options->progress_stage1 * 0.4f)) return LIQ_ABORTED;
if (liq_progress(options, options->progress_stage1 * 0.4f)) {
return LIQ_ABORTED;
}
const bool all_rows_at_once = liq_image_can_use_rgba_rows(input_image);
@ -1498,11 +1613,11 @@ LIQ_EXPORT LIQ_NONNULL liq_error liq_histogram_add_image(liq_histogram *input_hi
for(unsigned int row=0; row < rows; row++) {
bool added_ok;
if (all_rows_at_once) {
added_ok = pam_computeacolorhash(input_hist->acht, (const rgba_pixel *const *)input_image->rows, cols, rows, input_image->noise);
added_ok = pam_computeacolorhash(input_hist->acht, (const rgba_pixel *const *)input_image->rows, cols, rows, input_image->importance_map);
if (added_ok) break;
} else {
const rgba_pixel* rows_p[1] = { liq_image_get_row_rgba(input_image, row) };
added_ok = pam_computeacolorhash(input_hist->acht, rows_p, cols, 1, input_image->noise ? &input_image->noise[row * cols] : NULL);
added_ok = pam_computeacolorhash(input_hist->acht, rows_p, cols, 1, input_image->importance_map ? &input_image->importance_map[row * cols] : NULL);
}
if (!added_ok) {
input_hist->ignorebits++;
@ -1517,10 +1632,7 @@ LIQ_EXPORT LIQ_NONNULL liq_error liq_histogram_add_image(liq_histogram *input_hi
input_hist->had_image_added = true;
if (input_image->noise) {
input_image->free(input_image->noise);
input_image->noise = NULL;
}
liq_image_free_importance_map(input_image);
if (input_image->free_pixels && input_image->f_pixels) {
liq_image_free_rgba_source(input_image); // bow can free the RGBA source if copy has been made in f_pixels
@ -1578,7 +1690,7 @@ LIQ_NONNULL static void modify_alpha(liq_image *input_image, rgba_pixel *const r
/**
Builds two maps:
noise - approximation of areas with high-frequency noise, except straight edges. 1=flat, 0=noisy.
importance_map - approximation of areas with high-frequency noise, except straight edges. 1=flat, 0=noisy.
edges - noise map including all edges
*/
LIQ_NONNULL static void contrast_maps(liq_image *image)
@ -1588,14 +1700,14 @@ LIQ_NONNULL static void contrast_maps(liq_image *image)
return;
}
unsigned char *restrict noise = image->noise ? image->noise : image->malloc(cols*rows);
image->noise = NULL;
unsigned char *restrict noise = image->importance_map ? image->importance_map : image->malloc(cols*rows);
image->importance_map = NULL;
unsigned char *restrict edges = image->edges ? image->edges : image->malloc(cols*rows);
image->edges = NULL;
unsigned char *restrict tmp = image->malloc(cols*rows);
if (!noise || !edges || !tmp) {
if (!noise || !edges || !tmp || !liq_image_get_row_f_init(image)) {
image->free(noise);
image->free(edges);
image->free(tmp);
@ -1637,11 +1749,11 @@ LIQ_NONNULL static void contrast_maps(liq_image *image)
z = 1.f - MAX(z,MIN(horiz,vert));
z *= z; // noise is amplified
z *= z;
z *= 256.f;
noise[j*cols+i] = z < 256 ? z : 255;
z = (1.f-edge)*256.f;
edges[j*cols+i] = z > 0 ? (z < 256 ? z : 255) : 0;
// 85 is about 1/3rd of weight (not 0, because noisy pixels still need to be included, just not as precisely).
const unsigned int z_int = 85 + (unsigned int)(z * 171.f);
noise[j*cols+i] = MIN(z_int, 255);
const int e_int = 255 - (int)(edge * 256.f);
edges[j*cols+i] = e_int > 0 ? MIN(e_int, 255) : 0;
}
}
@ -1663,7 +1775,7 @@ LIQ_NONNULL static void contrast_maps(liq_image *image)
image->free(tmp);
image->noise = noise;
image->importance_map = noise;
image->edges = edges;
}
@ -1674,7 +1786,7 @@ LIQ_NONNULL static void contrast_maps(liq_image *image)
* and peeks 1 pixel above/below. Full 2d algorithm doesn't improve it significantly.
* Correct flood fill doesn't have visually good properties.
*/
LIQ_NONNULL static void update_dither_map(unsigned char *const *const row_pointers, liq_image *input_image)
LIQ_NONNULL static void update_dither_map(liq_image *input_image, unsigned char *const *const row_pointers, colormap *map)
{
const unsigned int width = input_image->width;
const unsigned int height = input_image->height;
@ -1686,6 +1798,10 @@ LIQ_NONNULL static void update_dither_map(unsigned char *const *const row_pointe
for(unsigned int col=1; col < width; col++) {
const unsigned char px = row_pointers[row][col];
if (input_image->background && map->palette[px].acolor.a < 1.f/256.f) {
// Transparency may or may not create an edge. When there's an explicit background set, assume no edge.
continue;
}
if (px != lastpixel || col == width-1) {
int neighbor_count = 10 * (col-lastcol);
@ -1758,6 +1874,10 @@ static colormap *find_best_palette(histogram *hist, const liq_attr *options, con
// at this point actual gamma is not set, so very conservative posterization estimate is used
const double target_mse = MIN(max_mse, MAX(options->target_mse, pow((1<<options->min_posterization_output)/1024.0, 2)));
int feedback_loop_trials = options->feedback_loop_trials;
if (hist->size > 5000) {feedback_loop_trials = (feedback_loop_trials*3 + 3)/4;}
if (hist->size > 25000) {feedback_loop_trials = (feedback_loop_trials*3 + 3)/4;}
if (hist->size > 50000) {feedback_loop_trials = (feedback_loop_trials*3 + 3)/4;}
if (hist->size > 100000) {feedback_loop_trials = (feedback_loop_trials*3 + 3)/4;}
colormap *acolormap = NULL;
double least_error = MAX_DIFF;
double target_mse_overshoot = feedback_loop_trials>0 ? 1.05 : 1.0;
@ -1767,7 +1887,7 @@ static colormap *find_best_palette(histogram *hist, const liq_attr *options, con
colormap *newmap;
if (hist->size && fixed_colors_count < max_colors) {
newmap = mediancut(hist, max_colors-fixed_colors_count, target_mse * target_mse_overshoot, MAX(MAX(45.0/65536.0, target_mse), least_error)*1.2,
options->malloc, options->free);
options->malloc, options->free);
} else {
feedback_loop_trials = 0;
newmap = NULL;
@ -1864,7 +1984,7 @@ LIQ_NONNULL static liq_error pngquant_quantize(histogram *hist, const liq_attr *
}
// K-Means iteration approaches local minimum for the palette
const double iteration_limit = options->kmeans_iteration_limit;
double iteration_limit = options->kmeans_iteration_limit;
unsigned int iterations = options->kmeans_iterations;
if (!iterations && palette_error < 0 && max_mse < MAX_DIFF) iterations = 1; // otherwise total error is never calculated and MSE limit won't work
@ -1877,6 +1997,11 @@ LIQ_NONNULL static liq_error pngquant_quantize(histogram *hist, const liq_attr *
}
}
if (hist->size > 5000) {iterations = (iterations*3 + 3)/4;}
if (hist->size > 25000) {iterations = (iterations*3 + 3)/4;}
if (hist->size > 50000) {iterations = (iterations*3 + 3)/4;}
if (hist->size > 100000) {iterations = (iterations*3 + 3)/4; iteration_limit *= 2;}
verbose_print(options, " moving colormap towards local minimum");
double previous_palette_error = MAX_DIFF;
@ -1958,7 +2083,7 @@ LIQ_EXPORT LIQ_NONNULL liq_error liq_write_remapped_image(liq_result *result, li
return LIQ_BUFFER_TOO_SMALL;
}
unsigned char *rows[input_image->height];
LIQ_ARRAY(unsigned char *, rows, input_image->height);
unsigned char *buffer_bytes = buffer;
for(unsigned int i=0; i < input_image->height; i++) {
rows[i] = &buffer_bytes[input_image->width * i];
@ -1998,11 +2123,13 @@ LIQ_EXPORT LIQ_NONNULL liq_error liq_write_remapped_image_rows(liq_result *quant
set_rounded_palette(&result->int_palette, result->palette, result->gamma, quant->min_posterization_output);
remapping_error = remap_to_palette(input_image, row_pointers, result->palette);
} else {
const bool generate_dither_map = result->use_dither_map && (input_image->edges && !input_image->dither_map);
const bool is_image_huge = (input_image->width * input_image->height) > 2000 * 2000;
const bool allow_dither_map = result->use_dither_map == 2 || (!is_image_huge && result->use_dither_map);
const bool generate_dither_map = allow_dither_map && (input_image->edges && !input_image->dither_map);
if (generate_dither_map) {
// If dithering (with dither map) is required, this image is used to find areas that require dithering
remapping_error = remap_to_palette(input_image, row_pointers, result->palette);
update_dither_map(row_pointers, input_image);
update_dither_map(input_image, row_pointers, result->palette);
}
if (liq_remap_progress(result, result->progress_stage1 * 0.5f)) {

13
libimagequant.h
View File

@ -13,8 +13,8 @@
#define LIQ_EXPORT extern
#endif
#define LIQ_VERSION 20900
#define LIQ_VERSION_STRING "2.9.0"
#define LIQ_VERSION 21200
#define LIQ_VERSION_STRING "2.12.2"
#ifndef LIQ_PRIVATE
#if defined(__GNUC__) || defined (__llvm__)
@ -60,7 +60,11 @@ typedef enum liq_error {
LIQ_UNSUPPORTED,
} liq_error;
enum liq_ownership {LIQ_OWN_ROWS=4, LIQ_OWN_PIXELS=8};
enum liq_ownership {
LIQ_OWN_ROWS=4,
LIQ_OWN_PIXELS=8,
LIQ_COPY_PIXELS=16,
};
typedef struct liq_histogram_entry {
liq_color color;
@ -75,6 +79,7 @@ LIQ_EXPORT void liq_attr_destroy(liq_attr *attr) LIQ_NONNULL;
LIQ_EXPORT LIQ_USERESULT liq_histogram* liq_histogram_create(const liq_attr* attr);
LIQ_EXPORT liq_error liq_histogram_add_image(liq_histogram *hist, const liq_attr *attr, liq_image* image) LIQ_NONNULL;
LIQ_EXPORT liq_error liq_histogram_add_colors(liq_histogram *hist, const liq_attr *attr, const liq_histogram_entry entries[], int num_entries, double gamma) LIQ_NONNULL;
LIQ_EXPORT liq_error liq_histogram_add_fixed_color(liq_histogram *hist, liq_color color, double gamma) LIQ_NONNULL;
LIQ_EXPORT void liq_histogram_destroy(liq_histogram *hist) LIQ_NONNULL;
LIQ_EXPORT liq_error liq_set_max_colors(liq_attr* attr, int colors) LIQ_NONNULL;
@ -107,6 +112,8 @@ typedef void liq_image_get_rgba_row_callback(liq_color row_out[], int row, int w
LIQ_EXPORT LIQ_USERESULT liq_image *liq_image_create_custom(const liq_attr *attr, liq_image_get_rgba_row_callback *row_callback, void* user_info, int width, int height, double gamma);
LIQ_EXPORT liq_error liq_image_set_memory_ownership(liq_image *image, int ownership_flags) LIQ_NONNULL;
LIQ_EXPORT liq_error liq_image_set_background(liq_image *img, liq_image *background_image) LIQ_NONNULL;
LIQ_EXPORT liq_error liq_image_set_importance_map(liq_image *img, unsigned char buffer[], size_t buffer_size, enum liq_ownership memory_handling) LIQ_NONNULL;
LIQ_EXPORT liq_error liq_image_add_fixed_color(liq_image *img, liq_color color) LIQ_NONNULL;
LIQ_EXPORT LIQ_USERESULT int liq_image_get_width(const liq_image *img) LIQ_NONNULL;
LIQ_EXPORT LIQ_USERESULT int liq_image_get_height(const liq_image *img) LIQ_NONNULL;

161
mediancut.c
View File

@ -1,32 +1,9 @@
/*
** © 2009-2015 by Kornel Lesiński.
** © 2009-2018 by Kornel Lesiński.
** © 1989, 1991 by Jef Poskanzer.
** © 1997, 2000, 2002 by Greg Roelofs; based on an idea by Stefan Schneider.
**
** This file is part of libimagequant.
**
** libimagequant is free software: you can redistribute it and/or modify
** it under the terms of the GNU General Public License as published by
** the Free Software Foundation, either version 3 of the License, or
** (at your option) any later version.
**
** libimagequant is distributed in the hope that it will be useful,
** but WITHOUT ANY WARRANTY; without even the implied warranty of
** MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
** GNU General Public License for more details.
**
** You should have received a copy of the GNU General Public License
** along with libimagequant. If not, see <http://www.gnu.org/licenses/>.
*/
/*
** Copyright (C) 1989, 1991 by Jef Poskanzer.
** Copyright (C) 1997, 2000, 2002 by Greg Roelofs; based on an idea by
** Stefan Schneider.
**
** Permission to use, copy, modify, and distribute this software and its
** documentation for any purpose and without fee is hereby granted, provided
** that the above copyright notice appear in all copies and that both that
** copyright notice and this permission notice appear in supporting
** documentation. This software is provided "as is" without express or
** implied warranty.
** See COPYRIGHT file for license.
*/
#include <stdlib.h>
@ -141,7 +118,7 @@ inline static unsigned int qsort_partition(hist_item *const base, const unsigned
}
/** quick select algorithm */
static void hist_item_sort_range(hist_item *base, unsigned int len, unsigned int sort_start)
static void hist_item_sort_range(hist_item base[], unsigned int len, unsigned int sort_start)
{
for(;;) {
const unsigned int l = qsort_partition(base, len), r = l+1;
@ -157,7 +134,7 @@ static void hist_item_sort_range(hist_item *base, unsigned int len, unsigned int
}
/** sorts array to make sum of weights lower than halfvar one side, returns edge between <halfvar and >halfvar parts of the set */
static hist_item *hist_item_sort_halfvar(hist_item *base, unsigned int len, double *const lowervar, const double halfvar)
static hist_item *hist_item_sort_halfvar(hist_item base[], unsigned int len, double *const lowervar, const double halfvar)
{
do {
const unsigned int l = qsort_partition(base, len), r = l+1;
@ -198,7 +175,7 @@ typedef struct {
static int comparevariance(const void *ch1, const void *ch2)
{
return ((const channelvariance*)ch1)->variance > ((const channelvariance*)ch2)->variance ? -1 :
(((const channelvariance*)ch1)->variance < ((const channelvariance*)ch2)->variance ? 1 : 0);
(((const channelvariance*)ch1)->variance < ((const channelvariance*)ch2)->variance ? 1 : 0);
}
/** Finds which channels need to be sorted first and preproceses achv for fast sort */
@ -216,12 +193,16 @@ static double prepare_sort(struct box *b, hist_item achv[])
qsort(channels, 4, sizeof(channels[0]), comparevariance);
for(unsigned int i=0; i < b->colors; i++) {
const float *chans = (const float *)&achv[b->ind + i].acolor;
const unsigned int ind1 = b->ind;
const unsigned int colors = b->colors;
#pragma omp parallel for if (colors > 25000) \
schedule(static) default(none) shared(achv, channels)
for(unsigned int i=0; i < colors; i++) {
const float *chans = (const float *)&achv[ind1 + i].acolor;
// Only the first channel really matters. When trying median cut many times
// with different histogram weights, I don't want sort randomness to influence outcome.
achv[b->ind + i].tmp.sort_value = ((unsigned int)(chans[channels[0].chan]*65535.0)<<16) |
(unsigned int)((chans[channels[2].chan] + chans[channels[1].chan]/2.0 + chans[channels[3].chan]/4.0)*65535.0);
achv[ind1 + i].tmp.sort_value = ((unsigned int)(chans[channels[0].chan]*65535.0)<<16) |
(unsigned int)((chans[channels[2].chan] + chans[channels[1].chan]/2.0 + chans[channels[3].chan]/4.0)*65535.0);
}
const f_pixel median = get_median(b, achv);
@ -229,6 +210,8 @@ static double prepare_sort(struct box *b, hist_item achv[])
// box will be split to make color_weight of each side even
const unsigned int ind = b->ind, end = ind+b->colors;
double totalvar = 0;
#pragma omp parallel for if (end - ind > 15000) \
schedule(static) default(shared) reduction(+:totalvar)
for(unsigned int j=ind; j < end; j++) totalvar += (achv[j].color_weight = color_weight(median, achv[j]));
return totalvar / 2.0;
}
@ -251,7 +234,7 @@ static f_pixel get_median(const struct box *b, hist_item achv[])
/*
** Find the best splittable box. -1 if no boxes are splittable.
*/
static int best_splittable_box(struct box* bv, unsigned int boxes, const double max_mse)
static int best_splittable_box(struct box bv[], unsigned int boxes, const double max_mse)
{
int bi=-1; double maxsum=0;
for(unsigned int i=0; i < boxes; i++) {
@ -281,8 +264,8 @@ inline static double color_weight(f_pixel median, hist_item h)
return sqrt(diff) * (sqrt(1.0+h.adjusted_weight)-1.0);
}
static void set_colormap_from_boxes(colormap *map, struct box* bv, unsigned int boxes, hist_item *achv);
static void adjust_histogram(hist_item *achv, const struct box* bv, unsigned int boxes);
static void set_colormap_from_boxes(colormap *map, struct box bv[], unsigned int boxes, hist_item *achv);
static void adjust_histogram(hist_item *achv, const struct box bv[], unsigned int boxes);
static double box_error(const struct box *box, const hist_item achv[])
{
@ -326,8 +309,11 @@ static void box_init(struct box *box, const hist_item *achv, const unsigned int
box->colors = colors;
box->sum = sum;
box->total_error = -1;
box->color = averagepixels(colors, &achv[ind]);
#pragma omp task if (colors > 5000)
box->variance = box_variance(achv, box);
#pragma omp task if (colors > 8000)
box->max_error = box_max_error(achv, box);
}
@ -339,67 +325,78 @@ static void box_init(struct box *box, const hist_item *achv, const unsigned int
LIQ_PRIVATE colormap *mediancut(histogram *hist, unsigned int newcolors, const double target_mse, const double max_mse, void* (*malloc)(size_t), void (*free)(void*))
{
hist_item *achv = hist->achv;
struct box bv[newcolors];
LIQ_ARRAY(struct box, bv, newcolors);
unsigned int boxes = 1;
/*
** Set up the initial box.
*/
#pragma omp parallel
#pragma omp single
{
double sum = 0;
for(unsigned int i=0; i < hist->size; i++) {
for(unsigned int i=0; i < hist->size; i++) {
sum += achv[i].adjusted_weight;
}
box_init(&bv[0], achv, 0, hist->size, sum);
#pragma omp taskgroup
{
box_init(&bv[0], achv, 0, hist->size, sum);
}
unsigned int boxes = 1;
/*
** Main loop: split boxes until we have enough.
*/
while (boxes < newcolors) {
// first splits boxes that exceed quality limit (to have colors for things like odd green pixel),
// later raises the limit to allow large smooth areas/gradients get colors.
const double current_max_mse = max_mse + (boxes/(double)newcolors)*16.0*max_mse;
const int bi = best_splittable_box(bv, boxes, current_max_mse);
if (bi < 0)
break; /* ran out of colors! */
unsigned int indx = bv[bi].ind;
unsigned int clrs = bv[bi].colors;
/*
Classic implementation tries to get even number of colors or pixels in each subdivision.
Here, instead of popularity I use (sqrt(popularity)*variance) metric.
Each subdivision balances number of pixels (popular colors) and low variance -
boxes can be large if they have similar colors. Later boxes with high variance
will be more likely to be split.
Median used as expected value gives much better results than mean.
** Main loop: split boxes until we have enough.
*/
while (boxes < newcolors) {
const double halfvar = prepare_sort(&bv[bi], achv);
double lowervar=0;
// first splits boxes that exceed quality limit (to have colors for things like odd green pixel),
// later raises the limit to allow large smooth areas/gradients get colors.
const double current_max_mse = max_mse + (boxes/(double)newcolors)*16.0*max_mse;
const int bi = best_splittable_box(bv, boxes, current_max_mse);
if (bi < 0) {
break; /* ran out of colors! */
}
// hist_item_sort_halfvar sorts and sums lowervar at the same time
// returns item to break at …minus one, which does smell like an off-by-one error.
hist_item *break_p = hist_item_sort_halfvar(&achv[indx], clrs, &lowervar, halfvar);
unsigned int break_at = MIN(clrs-1, break_p - &achv[indx] + 1);
unsigned int indx = bv[bi].ind;
unsigned int clrs = bv[bi].colors;
/*
** Split the box.
*/
double sm = bv[bi].sum;
double lowersum = 0;
for(unsigned int i=0; i < break_at; i++) lowersum += achv[indx + i].adjusted_weight;
/*
Classic implementation tries to get even number of colors or pixels in each subdivision.
box_init(&bv[bi], achv, bv[bi].ind, break_at, lowersum);
box_init(&bv[boxes], achv, indx + break_at, clrs - break_at, sm - lowersum);
Here, instead of popularity I use (sqrt(popularity)*variance) metric.
Each subdivision balances number of pixels (popular colors) and low variance -
boxes can be large if they have similar colors. Later boxes with high variance
will be more likely to be split.
++boxes;
Median used as expected value gives much better results than mean.
*/
if (total_box_error_below_target(target_mse, bv, boxes, hist)) {
break;
const double halfvar = prepare_sort(&bv[bi], achv);
double lowervar=0;
// hist_item_sort_halfvar sorts and sums lowervar at the same time
// returns item to break at …minus one, which does smell like an off-by-one error.
hist_item *break_p = hist_item_sort_halfvar(&achv[indx], clrs, &lowervar, halfvar);
unsigned int break_at = MIN(clrs-1, break_p - &achv[indx] + 1);
/*
** Split the box.
*/
double sm = bv[bi].sum;
double lowersum = 0;
for(unsigned int i=0; i < break_at; i++) lowersum += achv[indx + i].adjusted_weight;
#pragma omp taskgroup
{
box_init(&bv[bi], achv, indx, break_at, lowersum);
box_init(&bv[boxes], achv, indx + break_at, clrs - break_at, sm - lowersum);
}
++boxes;
if (total_box_error_below_target(target_mse, bv, boxes, hist)) {
break;
}
}
}
@ -446,6 +443,8 @@ static f_pixel averagepixels(unsigned int clrs, const hist_item achv[])
{
double r = 0, g = 0, b = 0, a = 0, sum = 0;
#pragma omp parallel for if (clrs > 25000) \
schedule(static) default(shared) reduction(+:a) reduction(+:r) reduction(+:g) reduction(+:b) reduction(+:sum)
for(unsigned int i = 0; i < clrs; i++) {
const f_pixel px = achv[i].acolor;
const double weight = achv[i].adjusted_weight;

31
mempool.c
View File

@ -1,20 +1,9 @@
/*
© 2011-2016 by Kornel Lesiński.
This file is part of libimagequant.
libimagequant is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation, either version 3 of the License, or
(at your option) any later version.
libimagequant is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License
along with libimagequant. If not, see <http://www.gnu.org/licenses/>.
** © 2009-2017 by Kornel Lesiński.
** © 1989, 1991 by Jef Poskanzer.
** © 1997, 2000, 2002 by Greg Roelofs; based on an idea by Stefan Schneider.
**
** See COPYRIGHT file for license.
*/
#include "libimagequant.h"
@ -32,7 +21,7 @@ struct mempool {
void (*free)(void*);
struct mempool *next;
};
LIQ_PRIVATE void* mempool_create(mempool *mptr, const unsigned int size, unsigned int max_size, void* (*malloc)(size_t), void (*free)(void*))
LIQ_PRIVATE void* mempool_create(mempoolptr *mptr, const unsigned int size, unsigned int max_size, void* (*malloc)(size_t), void (*free)(void*))
{
if (*mptr && ((*mptr)->used+size) <= (*mptr)->size) {
unsigned int prevused = (*mptr)->used;
@ -40,7 +29,7 @@ LIQ_PRIVATE void* mempool_create(mempool *mptr, const unsigned int size, unsigne
return ((char*)(*mptr)) + prevused;
}
mempool old = *mptr;
mempoolptr old = *mptr;
if (!max_size) max_size = (1<<17);
max_size = size+ALIGN_MASK > max_size ? size+ALIGN_MASK : max_size;
@ -60,7 +49,7 @@ LIQ_PRIVATE void* mempool_create(mempool *mptr, const unsigned int size, unsigne
return mempool_alloc(mptr, size, size);
}
LIQ_PRIVATE void* mempool_alloc(mempool *mptr, const unsigned int size, const unsigned int max_size)
LIQ_PRIVATE void* mempool_alloc(mempoolptr *mptr, const unsigned int size, const unsigned int max_size)
{
if (((*mptr)->used+size) <= (*mptr)->size) {
unsigned int prevused = (*mptr)->used;
@ -71,10 +60,10 @@ LIQ_PRIVATE void* mempool_alloc(mempool *mptr, const unsigned int size, const un
return mempool_create(mptr, size, max_size, (*mptr)->malloc, (*mptr)->free);
}
LIQ_PRIVATE void mempool_destroy(mempool m)
LIQ_PRIVATE void mempool_destroy(mempoolptr m)
{
while (m) {
mempool next = m->next;
mempoolptr next = m->next;
m->free(m);
m = next;
}

8
mempool.h
View File

@ -4,10 +4,10 @@
#include <stddef.h>
struct mempool;
typedef struct mempool *mempool;
typedef struct mempool *mempoolptr;
LIQ_PRIVATE void* mempool_create(mempool *mptr, const unsigned int size, unsigned int capacity, void* (*malloc)(size_t), void (*free)(void*));
LIQ_PRIVATE void* mempool_alloc(mempool *mptr, const unsigned int size, const unsigned int capacity);
LIQ_PRIVATE void mempool_destroy(mempool m);
LIQ_PRIVATE void* mempool_create(mempoolptr *mptr, const unsigned int size, unsigned int capacity, void* (*malloc)(size_t), void (*free)(void*));
LIQ_PRIVATE void* mempool_alloc(mempoolptr *mptr, const unsigned int size, const unsigned int capacity);
LIQ_PRIVATE void mempool_destroy(mempoolptr m);
#endif

33
nearest.c
View File

@ -1,20 +1,9 @@
/*
© 2011-2015 by Kornel Lesiński.
This file is part of libimagequant.
libimagequant is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation, either version 3 of the License, or
(at your option) any later version.
libimagequant is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License
along with libimagequant. If not, see <http://www.gnu.org/licenses/>.
** © 2009-2015 by Kornel Lesiński.
** © 1989, 1991 by Jef Poskanzer.
** © 1997, 2000, 2002 by Greg Roelofs; based on an idea by Stefan Schneider.
**
** See COPYRIGHT file for license.
*/
#include "libimagequant.h"
@ -45,7 +34,7 @@ struct nearest_map {
vp_node *root;
const colormap_item *palette;
float nearest_other_color_dist[256];
mempool mempool;
mempoolptr mempool;
};
static void vp_search_node(const vp_node *node, const f_pixel *const needle, vp_search_tmp *const best_candidate);
@ -56,7 +45,7 @@ static int vp_compare_distance(const void *ap, const void *bp) {
return a > b ? 1 : -1;
}
static void vp_sort_indexes_by_distance(const f_pixel vantage_point, vp_sort_tmp *indexes, int num_indexes, const colormap_item items[]) {
static void vp_sort_indexes_by_distance(const f_pixel vantage_point, vp_sort_tmp indexes[], int num_indexes, const colormap_item items[]) {
for(int i=0; i < num_indexes; i++) {
indexes[i].distance_squared = colordifference(vantage_point, items[indexes[i].idx].acolor);
}
@ -66,7 +55,7 @@ static void vp_sort_indexes_by_distance(const f_pixel vantage_point, vp_sort_tmp
/*
* Usually it should pick farthest point, but picking most popular point seems to make search quicker anyway
*/
static int vp_find_best_vantage_point_index(vp_sort_tmp *indexes, int num_indexes, const colormap_item items[]) {
static int vp_find_best_vantage_point_index(vp_sort_tmp indexes[], int num_indexes, const colormap_item items[]) {
int best = 0;
float best_popularity = items[indexes[0].idx].popularity;
for(int i = 1; i < num_indexes; i++) {
@ -78,7 +67,7 @@ static int vp_find_best_vantage_point_index(vp_sort_tmp *indexes, int num_indexe
return best;
}
static vp_node *vp_create_node(mempool *m, vp_sort_tmp *indexes, int num_indexes, const colormap_item items[]) {
static vp_node *vp_create_node(mempoolptr *m, vp_sort_tmp indexes[], int num_indexes, const colormap_item items[]) {
if (num_indexes <= 0) {
return NULL;
}
@ -118,10 +107,10 @@ static vp_node *vp_create_node(mempool *m, vp_sort_tmp *indexes, int num_indexes
}
LIQ_PRIVATE struct nearest_map *nearest_init(const colormap *map) {
mempool m = NULL;
mempoolptr m = NULL;
struct nearest_map *handle = mempool_create(&m, sizeof(handle[0]), sizeof(handle[0]) + sizeof(vp_node)*map->colors+16, map->malloc, map->free);
vp_sort_tmp indexes[map->colors];
LIQ_ARRAY(vp_sort_tmp, indexes, map->colors);
for(unsigned int i=0; i < map->colors; i++) {
indexes[i].idx = i;

62
pam.c
View File

@ -1,16 +1,10 @@
/* pam.c - pam (portable alpha map) utility library
**
** Copyright (C) 1989, 1991 by Jef Poskanzer.
** Copyright (C) 1997, 2000, 2002 by Greg Roelofs; based on an idea by
** Stefan Schneider.
** © 2009-2016 by Kornel Lesinski.
** © 2009-2017 by Kornel Lesiński.
** © 1989, 1991 by Jef Poskanzer.
** © 1997, 2000, 2002 by Greg Roelofs; based on an idea by Stefan Schneider.
**
** Permission to use, copy, modify, and distribute this software and its
** documentation for any purpose and without fee is hereby granted, provided
** that the above copyright notice appear in all copies and that both that
** copyright notice and this permission notice appear in supporting
** documentation. This software is provided "as is" without express or
** implied warranty.
** See COPYRIGHT file for license.
*/
#include <stdlib.h>
@ -33,11 +27,8 @@ LIQ_PRIVATE bool pam_computeacolorhash(struct acolorhash_table *acht, const rgba
/* Go through the entire image, building a hash table of colors. */
for(unsigned int row = 0; row < rows; ++row) {
float boost=1.0;
for(unsigned int col = 0; col < cols; ++col) {
if (importance_map) {
boost = 0.5f+ (double)*importance_map++/255.f;
}
unsigned int boost;
// RGBA color is casted to long for easier hasing/comparisons
union rgba_as_int px = {pixels[row][col]};
@ -45,12 +36,22 @@ LIQ_PRIVATE bool pam_computeacolorhash(struct acolorhash_table *acht, const rgba
if (!px.rgba.a) {
// "dirty alpha" has different RGBA values that end up being the same fully transparent color
px.l=0; hash=0;
boost = 10;
boost = 2000;
if (importance_map) {
importance_map++;
}
} else {
// mask posterizes all 4 channels in one go
px.l = (px.l & posterize_mask) | ((px.l & posterize_high_mask) >> (8-ignorebits));
// fancier hashing algorithms didn't improve much
hash = px.l % hash_size;
if (importance_map) {
boost = *importance_map++;
} else {
boost = 255;
}
}
if (!pam_add_to_hash(acht, hash, boost, px, row, rows)) {
@ -64,7 +65,7 @@ LIQ_PRIVATE bool pam_computeacolorhash(struct acolorhash_table *acht, const rgba
return true;
}
LIQ_PRIVATE bool pam_add_to_hash(struct acolorhash_table *acht, unsigned int hash, float boost, union rgba_as_int px, unsigned int row, unsigned int rows)
LIQ_PRIVATE bool pam_add_to_hash(struct acolorhash_table *acht, unsigned int hash, unsigned int boost, union rgba_as_int px, unsigned int row, unsigned int rows)
{
/* head of the hash function stores first 2 colors inline (achl->used = 1..2),
to reduce number of allocations of achl->other_items.
@ -160,7 +161,7 @@ LIQ_PRIVATE struct acolorhash_table *pam_allocacolorhash(unsigned int maxcolors,
const size_t estimated_colors = MIN(maxcolors, surface/(ignorebits + (surface > 512*512 ? 6 : 5)));
const size_t hash_size = estimated_colors < 66000 ? 6673 : (estimated_colors < 200000 ? 12011 : 24019);
mempool m = NULL;
mempoolptr m = NULL;
const size_t buckets_size = hash_size * sizeof(struct acolorhist_arr_head);
const size_t mempool_size = sizeof(struct acolorhash_table) + buckets_size + estimated_colors * sizeof(struct acolorhist_arr_item);
struct acolorhash_table *t = mempool_create(&m, sizeof(*t) + buckets_size, mempool_size, malloc, free);
@ -175,11 +176,15 @@ LIQ_PRIVATE struct acolorhash_table *pam_allocacolorhash(unsigned int maxcolors,
return t;
}
ALWAYS_INLINE static float pam_add_to_hist(const float *gamma_lut, hist_item *achv, unsigned int j, const struct acolorhist_arr_item *entry, const float max_perceptual_weight)
ALWAYS_INLINE static float pam_add_to_hist(const float *gamma_lut, hist_item *achv, unsigned int *j, const struct acolorhist_arr_item *entry, const float max_perceptual_weight)
{
achv[j].acolor = rgba_to_f(gamma_lut, entry->color.rgba);
const float w = MIN(entry->perceptual_weight, max_perceptual_weight);
achv[j].adjusted_weight = achv[j].perceptual_weight = w;
if (entry->perceptual_weight == 0) {
return 0;
}
const float w = MIN(entry->perceptual_weight/128.f, max_perceptual_weight);
achv[*j].adjusted_weight = achv[*j].perceptual_weight = w;
achv[*j].acolor = rgba_to_f(gamma_lut, entry->color.rgba);
*j += 1;
return w;
}
@ -203,22 +208,27 @@ LIQ_PRIVATE histogram *pam_acolorhashtoacolorhist(const struct acolorhash_table
float max_perceptual_weight = 0.1f * acht->cols * acht->rows;
double total_weight = 0;
for(unsigned int j=0, i=0; i < acht->hash_size; ++i) {
unsigned int j=0;
for(unsigned int i=0; i < acht->hash_size; ++i) {
const struct acolorhist_arr_head *const achl = &acht->buckets[i];
if (achl->used) {
total_weight += pam_add_to_hist(gamma_lut, hist->achv, j++, &achl->inline1, max_perceptual_weight);
total_weight += pam_add_to_hist(gamma_lut, hist->achv, &j, &achl->inline1, max_perceptual_weight);
if (achl->used > 1) {
total_weight += pam_add_to_hist(gamma_lut, hist->achv, j++, &achl->inline2, max_perceptual_weight);
total_weight += pam_add_to_hist(gamma_lut, hist->achv, &j, &achl->inline2, max_perceptual_weight);
for(unsigned int k=0; k < achl->used-2; k++) {
total_weight += pam_add_to_hist(gamma_lut, hist->achv, j++, &achl->other_items[k], max_perceptual_weight);
total_weight += pam_add_to_hist(gamma_lut, hist->achv, &j, &achl->other_items[k], max_perceptual_weight);
}
}
}
}
hist->size = j;
hist->total_perceptual_weight = total_weight;
if (!j) {
pam_freeacolorhist(hist);
return NULL;
}
return hist;
}

12
pam.h
View File

@ -62,6 +62,12 @@
# define SSE_ALIGN
#endif
#ifndef _MSC_VER
#define LIQ_ARRAY(type, var, count) type var[count]
#else
#define LIQ_ARRAY(type, var, count) type* var = (type*)_alloca(sizeof(type)*(count))
#endif
#if defined(__GNUC__) || defined (__llvm__)
#define ALWAYS_INLINE __attribute__((always_inline)) inline
#define NEVER_INLINE __attribute__ ((noinline))
@ -85,7 +91,7 @@ typedef struct {
float a, r, g, b;
} SSE_ALIGN f_pixel;
static const double internal_gamma = 0.5499;
static const float internal_gamma = 0.5499f;
LIQ_PRIVATE void to_f_set_gamma(float gamma_lut[], const double gamma);
@ -238,7 +244,7 @@ typedef struct colormap {
struct acolorhist_arr_item {
union rgba_as_int color;
float perceptual_weight;
unsigned int perceptual_weight;
};
struct acolorhist_arr_head {
@ -260,7 +266,7 @@ LIQ_PRIVATE void pam_freeacolorhash(struct acolorhash_table *acht);
LIQ_PRIVATE struct acolorhash_table *pam_allocacolorhash(unsigned int maxcolors, unsigned int surface, unsigned int ignorebits, void* (*malloc)(size_t), void (*free)(void*));
LIQ_PRIVATE histogram *pam_acolorhashtoacolorhist(const struct acolorhash_table *acht, const double gamma, void* (*malloc)(size_t), void (*free)(void*));
LIQ_PRIVATE bool pam_computeacolorhash(struct acolorhash_table *acht, const rgba_pixel *const pixels[], unsigned int cols, unsigned int rows, const unsigned char *importance_map);
LIQ_PRIVATE bool pam_add_to_hash(struct acolorhash_table *acht, unsigned int hash, float boost, union rgba_as_int px, unsigned int row, unsigned int rows);
LIQ_PRIVATE bool pam_add_to_hash(struct acolorhash_table *acht, unsigned int hash, unsigned int boost, union rgba_as_int px, unsigned int row, unsigned int rows);
LIQ_PRIVATE void pam_freeacolorhist(histogram *h);