Add modified Zeng's method to palette sorting.

Also add palette sorting to crunch configurations.

Change-Id: I010a8bf8f1921279db6e9c7209307d8d19a4d105
This commit is contained in:
Vincent Rabaud 2021-04-27 11:34:36 +02:00
parent 88c90c4528
commit b1674240f9
2 changed files with 276 additions and 69 deletions

View File

@ -65,25 +65,22 @@ static WEBP_INLINE void SwapColor(uint32_t* const col1, uint32_t* const col2) {
*col2 = tmp;
}
static void GreedyMinimizeDeltas(uint32_t palette[], int num_colors) {
// Find greedily always the closest color of the predicted color to minimize
// deltas in the palette. This reduces storage needs since the
// palette is stored with delta encoding.
uint32_t predict = 0x00000000;
int i, k;
for (i = 0; i < num_colors; ++i) {
int best_ix = i;
uint32_t best_score = ~0U;
for (k = i; k < num_colors; ++k) {
const uint32_t cur_score = PaletteColorDistance(palette[k], predict);
if (best_score > cur_score) {
best_score = cur_score;
best_ix = k;
}
static WEBP_INLINE int SearchColorNoIdx(const uint32_t sorted[], uint32_t color,
int num_colors) {
int low = 0, hi = num_colors;
if (sorted[low] == color) return low; // loop invariant: sorted[low] != color
while (1) {
const int mid = (low + hi) >> 1;
if (sorted[mid] == color) {
return mid;
} else if (sorted[mid] < color) {
low = mid;
} else {
hi = mid;
}
SwapColor(&palette[best_ix], &palette[i]);
predict = palette[i];
}
assert(0);
return 0;
}
// The palette has been sorted by alpha. This function checks if the other
@ -92,7 +89,8 @@ static void GreedyMinimizeDeltas(uint32_t palette[], int num_colors) {
// no benefit to re-organize them greedily. A monotonic development
// would be spotted in green-only situations (like lossy alpha) or gray-scale
// images.
static int PaletteHasNonMonotonousDeltas(uint32_t palette[], int num_colors) {
static int PaletteHasNonMonotonousDeltas(const uint32_t* const palette,
int num_colors) {
uint32_t predict = 0x000000;
int i;
uint8_t sign_found = 0x00;
@ -115,28 +113,215 @@ static int PaletteHasNonMonotonousDeltas(uint32_t palette[], int num_colors) {
return (sign_found & (sign_found << 1)) != 0; // two consequent signs.
}
static void PaletteSortMinimizeDeltas(const uint32_t* const palette_sorted,
int num_colors, uint32_t* const palette) {
uint32_t predict = 0x00000000;
int i, k;
memcpy(palette, palette_sorted, num_colors * sizeof(*palette));
if (!PaletteHasNonMonotonousDeltas(palette_sorted, num_colors)) return;
// Find greedily always the closest color of the predicted color to minimize
// deltas in the palette. This reduces storage needs since the
// palette is stored with delta encoding.
for (i = 0; i < num_colors; ++i) {
int best_ix = i;
uint32_t best_score = ~0U;
for (k = i; k < num_colors; ++k) {
const uint32_t cur_score = PaletteColorDistance(palette[k], predict);
if (best_score > cur_score) {
best_score = cur_score;
best_ix = k;
}
}
SwapColor(&palette[best_ix], &palette[i]);
predict = palette[i];
}
}
// Sort palette in increasing order and prepare an inverse mapping array.
static void PrepareMapToPalette(const uint32_t palette[], uint32_t num_colors,
uint32_t sorted[], uint32_t idx_map[]) {
uint32_t i;
memcpy(sorted, palette, num_colors * sizeof(*sorted));
qsort(sorted, num_colors, sizeof(*sorted), PaletteCompareColorsForQsort);
for (i = 0; i < num_colors; ++i) {
idx_map[SearchColorNoIdx(sorted, palette[i], num_colors)] = i;
}
}
// -----------------------------------------------------------------------------
// Modified Zeng method from "A Survey on Palette Reordering
// Methods for Improving the Compression of Color-Indexed Images" by Armando J.
// Pinho and Antonio J. R. Neves.
// Finds the biggest cooccurrence in the matrix.
static void CoOccurrenceFindMax(const uint32_t* const cooccurrence,
uint32_t num_colors, uint8_t* const c1,
uint8_t* const c2) {
// Find the index that is most frequently located adjacent to other
// (different) indexes.
uint32_t best_sum = 0u;
uint32_t i, j, best_cooccurrence;
*c1 = 0u;
for (i = 0; i < num_colors; ++i) {
uint32_t sum = 0;
for (j = 0; j < num_colors; ++j) sum += cooccurrence[i * num_colors + j];
if (sum > best_sum) {
best_sum = sum;
*c1 = i;
}
}
// Find the index that is most frequently found adjacent to *c1.
*c2 = 0u;
best_cooccurrence = 0u;
for (i = 0; i < num_colors; ++i) {
if (cooccurrence[*c1 * num_colors + i] > best_cooccurrence) {
best_cooccurrence = cooccurrence[*c1 * num_colors + i];
*c2 = i;
}
}
assert(*c1 != *c2);
}
// Builds the cooccurrence matrix
static WebPEncodingError CoOccurrenceBuild(const WebPPicture* const pic,
const uint32_t* const palette,
uint32_t num_colors,
uint32_t* cooccurrence) {
uint32_t *lines, *line_top, *line_current, *line_tmp;
int x, y;
const uint32_t* src = pic->argb;
uint32_t prev_pix = ~src[0];
uint32_t prev_idx = 0u;
uint32_t idx_map[MAX_PALETTE_SIZE] = {0};
uint32_t palette_sorted[MAX_PALETTE_SIZE];
lines = (uint32_t*)WebPSafeMalloc(2 * pic->width, sizeof(*lines));
if (lines == NULL) return VP8_ENC_ERROR_OUT_OF_MEMORY;
line_top = &lines[0];
line_current = &lines[pic->width];
PrepareMapToPalette(palette, num_colors, palette_sorted, idx_map);
for (y = 0; y < pic->height; ++y) {
for (x = 0; x < pic->width; ++x) {
const uint32_t pix = src[x];
if (pix != prev_pix) {
prev_idx = idx_map[SearchColorNoIdx(palette_sorted, pix, num_colors)];
prev_pix = pix;
}
line_current[x] = prev_idx;
// 4-connectivity is what works best as mentioned in "On the relation
// between Memon's and the modified Zeng's palette reordering methods".
if (x > 0 && prev_idx != line_current[x - 1]) {
const uint32_t left_idx = line_current[x - 1];
++cooccurrence[prev_idx * num_colors + left_idx];
++cooccurrence[left_idx * num_colors + prev_idx];
}
if (y > 0 && prev_idx != line_top[x]) {
const uint32_t top_idx = line_top[x];
++cooccurrence[prev_idx * num_colors + top_idx];
++cooccurrence[top_idx * num_colors + prev_idx];
}
}
line_tmp = line_top;
line_top = line_current;
line_current = line_tmp;
src += pic->argb_stride;
}
WebPSafeFree(lines);
return VP8_ENC_OK;
}
struct Sum {
uint16_t index;
uint32_t sum;
};
// Implements the modified Zeng method from "A Survey on Palette Reordering
// Methods for Improving the Compression of Color-Indexed Images" by Armando J.
// Pinho and Antonio J. R. Neves.
static WebPEncodingError PaletteSortModifiedZeng(
const WebPPicture* const pic, const uint32_t* const palette_sorted,
uint32_t num_colors, uint32_t* const palette) {
uint32_t i, j, ind;
uint8_t remapping[MAX_PALETTE_SIZE];
uint32_t* cooccurrence;
struct Sum sums[MAX_PALETTE_SIZE];
uint32_t first, last;
uint32_t num_sums;
// TODO(vrabaud) check whether one color images should use palette or not.
if (num_colors <= 1) return VP8_ENC_OK;
// Build the co-occurrence matrix.
cooccurrence =
(uint32_t*)WebPSafeCalloc(num_colors * num_colors, sizeof(*cooccurrence));
if (cooccurrence == NULL) return VP8_ENC_ERROR_OUT_OF_MEMORY;
if (CoOccurrenceBuild(pic, palette_sorted, num_colors, cooccurrence) !=
VP8_ENC_OK) {
WebPSafeFree(cooccurrence);
return VP8_ENC_ERROR_OUT_OF_MEMORY;
}
// Initialize the mapping list with the two best indices.
CoOccurrenceFindMax(cooccurrence, num_colors, &remapping[0], &remapping[1]);
// We need to append and prepend to the list of remapping. To this end, we
// actually define the next start/end of the list as indices in a vector (with
// a wrap around when the end is reached).
first = 0;
last = 1;
num_sums = num_colors - 2; // -2 because we know the first two values
if (num_sums > 0) {
// Initialize the sums with the first two remappings and find the best one
struct Sum* best_sum = &sums[0];
best_sum->index = 0u;
best_sum->sum = 0u;
for (i = 0, j = 0; i < num_colors; ++i) {
if (i == remapping[0] || i == remapping[1]) continue;
sums[j].index = i;
sums[j].sum = cooccurrence[i * num_colors + remapping[0]] +
cooccurrence[i * num_colors + remapping[1]];
if (sums[j].sum > best_sum->sum) best_sum = &sums[j];
++j;
}
while (num_sums > 0) {
const uint16_t best_index = best_sum->index;
// Compute delta to know if we need to prepend or append the best index.
int32_t delta = 0;
const int32_t n = num_colors - num_sums;
for (ind = first, j = 0; (ind + j) % num_colors != last + 1; ++j) {
const uint16_t l_j = remapping[(ind + j) % num_colors];
delta += (n - 1 - 2 * (int32_t)j) *
(int32_t)cooccurrence[best_index * num_colors + l_j];
}
if (delta > 0) {
first = (first == 0) ? num_colors - 1 : first - 1;
remapping[first] = best_index;
} else {
++last;
remapping[last] = best_index;
}
// Remove best_sum from sums.
*best_sum = sums[num_sums - 1];
--num_sums;
// Update all the sums and find the best one.
best_sum = &sums[0];
for (i = 0; i < num_sums; ++i) {
sums[i].sum += cooccurrence[best_index * num_colors + sums[i].index];
if (sums[i].sum > best_sum->sum) best_sum = &sums[i];
}
}
}
assert((last + 1) % num_colors == first);
WebPSafeFree(cooccurrence);
// Re-map the palette.
for (i = 0; i < num_colors; ++i) {
palette[i] = palette_sorted[remapping[(first + i) % num_colors]];
}
return VP8_ENC_OK;
}
// -----------------------------------------------------------------------------
// Palette
// If number of colors in the image is less than or equal to MAX_PALETTE_SIZE,
// creates a palette and returns true, else returns false.
static int AnalyzeAndCreatePalette(const WebPPicture* const pic,
int low_effort,
uint32_t palette[MAX_PALETTE_SIZE],
int* const palette_size) {
const int num_colors = WebPGetColorPalette(pic, palette);
if (num_colors > MAX_PALETTE_SIZE) {
*palette_size = 0;
return 0;
}
*palette_size = num_colors;
qsort(palette, num_colors, sizeof(*palette), PaletteCompareColorsForQsort);
if (!low_effort && PaletteHasNonMonotonousDeltas(palette, num_colors)) {
GreedyMinimizeDeltas(palette, num_colors);
}
return 1;
}
// These five modes are evaluated and their respective entropy is computed.
typedef enum {
kDirect = 0,
@ -148,6 +333,13 @@ typedef enum {
kNumEntropyIx = 6
} EntropyIx;
typedef enum {
kSortedDefault = 0,
kMinimizeDelta = 1,
kModifiedZeng = 2,
kUnusedPalette = 3,
} PaletteSorting;
typedef enum {
kHistoAlpha = 0,
kHistoAlphaPred,
@ -362,11 +554,14 @@ typedef struct {
} CrunchSubConfig;
typedef struct {
int entropy_idx_;
PaletteSorting palette_sorting_type_;
CrunchSubConfig sub_configs_[CRUNCH_SUBCONFIGS_MAX];
int sub_configs_size_;
} CrunchConfig;
#define CRUNCH_CONFIGS_MAX kNumEntropyIx
// +2 because we add a palette sorting configuration for kPalette and
// kPaletteAndSpatial.
#define CRUNCH_CONFIGS_MAX (kNumEntropyIx + 2)
static int EncoderAnalyze(VP8LEncoder* const enc,
CrunchConfig crunch_configs[CRUNCH_CONFIGS_MAX],
@ -386,9 +581,15 @@ static int EncoderAnalyze(VP8LEncoder* const enc,
int do_no_cache = 0;
assert(pic != NULL && pic->argb != NULL);
use_palette =
AnalyzeAndCreatePalette(pic, low_effort,
enc->palette_, &enc->palette_size_);
// Check whether a palette is possible.
enc->palette_size_ = WebPGetColorPalette(pic, enc->palette_sorted_);
use_palette = (enc->palette_size_ <= MAX_PALETTE_SIZE);
if (!use_palette) {
enc->palette_size_ = 0;
} else {
qsort(enc->palette_sorted_, enc->palette_size_,
sizeof(*enc->palette_sorted_), PaletteCompareColorsForQsort);
}
// Empirical bit sizes.
enc->histo_bits_ = GetHistoBits(method, use_palette,
@ -398,6 +599,8 @@ static int EncoderAnalyze(VP8LEncoder* const enc,
if (low_effort) {
// AnalyzeEntropy is somewhat slow.
crunch_configs[0].entropy_idx_ = use_palette ? kPalette : kSpatialSubGreen;
crunch_configs[0].palette_sorting_type_ =
use_palette ? kSortedDefault : kUnusedPalette;
n_lz77s = 1;
*crunch_configs_size = 1;
} else {
@ -418,13 +621,28 @@ static int EncoderAnalyze(VP8LEncoder* const enc,
// a palette.
if ((i != kPalette && i != kPaletteAndSpatial) || use_palette) {
assert(*crunch_configs_size < CRUNCH_CONFIGS_MAX);
crunch_configs[(*crunch_configs_size)++].entropy_idx_ = i;
crunch_configs[(*crunch_configs_size)].entropy_idx_ = i;
if (use_palette && (i == kPalette || i == kPaletteAndSpatial)) {
crunch_configs[(*crunch_configs_size)].palette_sorting_type_ =
kMinimizeDelta;
++*crunch_configs_size;
// Also add modified Zeng's method.
crunch_configs[(*crunch_configs_size)].entropy_idx_ = i;
crunch_configs[(*crunch_configs_size)].palette_sorting_type_ =
kModifiedZeng;
} else {
crunch_configs[(*crunch_configs_size)].palette_sorting_type_ =
kUnusedPalette;
}
++*crunch_configs_size;
}
}
} else {
// Only choose the guessed best transform.
*crunch_configs_size = 1;
crunch_configs[0].entropy_idx_ = min_entropy_ix;
crunch_configs[0].palette_sorting_type_ =
use_palette ? kMinimizeDelta : kUnusedPalette;
if (config->quality >= 75 && method == 5) {
// Test with and without color cache.
do_no_cache = 1;
@ -432,6 +650,7 @@ static int EncoderAnalyze(VP8LEncoder* const enc,
if (min_entropy_ix == kPalette) {
*crunch_configs_size = 2;
crunch_configs[1].entropy_idx_ = kPaletteAndSpatial;
crunch_configs[1].palette_sorting_type_ = kMinimizeDelta;
}
}
}
@ -1283,22 +1502,6 @@ static WebPEncodingError MakeInputImageCopy(VP8LEncoder* const enc) {
// -----------------------------------------------------------------------------
static WEBP_INLINE int SearchColorNoIdx(const uint32_t sorted[], uint32_t color,
int hi) {
int low = 0;
if (sorted[low] == color) return low; // loop invariant: sorted[low] != color
while (1) {
const int mid = (low + hi) >> 1;
if (sorted[mid] == color) {
return mid;
} else if (sorted[mid] < color) {
low = mid;
} else {
hi = mid;
}
}
}
#define APPLY_PALETTE_GREEDY_MAX 4
static WEBP_INLINE uint32_t SearchColorGreedy(const uint32_t palette[],
@ -1333,17 +1536,6 @@ static WEBP_INLINE uint32_t ApplyPaletteHash2(uint32_t color) {
(32 - PALETTE_INV_SIZE_BITS);
}
// Sort palette in increasing order and prepare an inverse mapping array.
static void PrepareMapToPalette(const uint32_t palette[], int num_colors,
uint32_t sorted[], uint32_t idx_map[]) {
int i;
memcpy(sorted, palette, num_colors * sizeof(*sorted));
qsort(sorted, num_colors, sizeof(*sorted), PaletteCompareColorsForQsort);
for (i = 0; i < num_colors; ++i) {
idx_map[SearchColorNoIdx(sorted, palette[i], num_colors)] = i;
}
}
// Use 1 pixel cache for ARGB pixels.
#define APPLY_PALETTE_FOR(COLOR_INDEX) do { \
uint32_t prev_pix = palette[0]; \
@ -1604,6 +1796,19 @@ static int EncodeStreamHook(void* input, void* data2) {
// Encode palette
if (enc->use_palette_) {
if (crunch_configs[idx].palette_sorting_type_ == kSortedDefault) {
// Nothing to do, we have already sorted the palette.
memcpy(enc->palette_, enc->palette_sorted_,
enc->palette_size_ * sizeof(*enc->palette_));
} else if (crunch_configs[idx].palette_sorting_type_ == kMinimizeDelta) {
PaletteSortMinimizeDeltas(enc->palette_sorted_, enc->palette_size_,
enc->palette_);
} else {
assert(crunch_configs[idx].palette_sorting_type_ == kModifiedZeng);
err = PaletteSortModifiedZeng(enc->pic_, enc->palette_sorted_,
enc->palette_size_, enc->palette_);
if (err != VP8_ENC_OK) goto Error;
}
err = EncodePalette(bw, low_effort, enc);
if (err != VP8_ENC_OK) goto Error;
err = MapImageFromPalette(enc, use_delta_palette);

View File

@ -69,6 +69,8 @@ typedef struct {
int use_palette_;
int palette_size_;
uint32_t palette_[MAX_PALETTE_SIZE];
// Sorted version of palette_ for cache purposes.
uint32_t palette_sorted_[MAX_PALETTE_SIZE];
// Some 'scratch' (potentially large) objects.
struct VP8LBackwardRefs refs_[4]; // Backward Refs array for temporaries.