Adds a wiener filter based restoration scheme in loop which can be optionally selected instead of the bilateral filter. The LMMSE filter generated per frame is a separable symmetric 7 tap filter. Three parameters for each of horizontal and vertical filters are transmitted in the bitstream. The fourth parameter is obtained assuming the sum is normalized to 1. Also integerizes the bilateral filters, along with other refactoring necessary in order to support the new switchable restoration type framework. derflr: -0.75% BDRATE [A lot of videos still prefer bilateral, however since many frames now use the simpler separable filter, the decoding speed is much better]. Further experiments to follow, related to replacing the bilateral. Change-Id: I6b1879983d50aab7ec5647340b6aef6b22299636
203 lines
6.8 KiB
C
203 lines
6.8 KiB
C
/*
|
|
* Copyright (c) 2010 The WebM project authors. All Rights Reserved.
|
|
*
|
|
* Use of this source code is governed by a BSD-style license
|
|
* that can be found in the LICENSE file in the root of the source
|
|
* tree. An additional intellectual property rights grant can be found
|
|
* in the file PATENTS. All contributing project authors may
|
|
* be found in the AUTHORS file in the root of the source tree.
|
|
*/
|
|
|
|
#include <assert.h>
|
|
#include <limits.h>
|
|
|
|
#include "./vpx_scale_rtcd.h"
|
|
|
|
#include "vpx_dsp/vpx_dsp_common.h"
|
|
#include "vpx_mem/vpx_mem.h"
|
|
#include "vpx_ports/mem.h"
|
|
|
|
#include "vp10/common/loopfilter.h"
|
|
#include "vp10/common/onyxc_int.h"
|
|
#include "vp10/common/quant_common.h"
|
|
|
|
#include "vp10/encoder/encoder.h"
|
|
#include "vp10/encoder/picklpf.h"
|
|
#include "vp10/encoder/quantize.h"
|
|
|
|
int vp10_get_max_filter_level(const VP10_COMP *cpi) {
|
|
if (cpi->oxcf.pass == 2) {
|
|
return cpi->twopass.section_intra_rating > 8 ? MAX_LOOP_FILTER * 3 / 4
|
|
: MAX_LOOP_FILTER;
|
|
} else {
|
|
return MAX_LOOP_FILTER;
|
|
}
|
|
}
|
|
|
|
static int64_t try_filter_frame(const YV12_BUFFER_CONFIG *sd,
|
|
VP10_COMP *const cpi,
|
|
int filt_level, int partial_frame) {
|
|
VP10_COMMON *const cm = &cpi->common;
|
|
int64_t filt_err;
|
|
|
|
#if CONFIG_VAR_TX
|
|
vp10_loop_filter_frame(cm->frame_to_show, cm, &cpi->td.mb.e_mbd, filt_level,
|
|
1, partial_frame);
|
|
#else
|
|
if (cpi->num_workers > 1)
|
|
vp10_loop_filter_frame_mt(cm->frame_to_show, cm, cpi->td.mb.e_mbd.plane,
|
|
filt_level, 1, partial_frame,
|
|
cpi->workers, cpi->num_workers,
|
|
&cpi->lf_row_sync);
|
|
else
|
|
vp10_loop_filter_frame(cm->frame_to_show, cm, &cpi->td.mb.e_mbd, filt_level,
|
|
1, partial_frame);
|
|
#endif
|
|
|
|
#if CONFIG_VP9_HIGHBITDEPTH
|
|
if (cm->use_highbitdepth) {
|
|
filt_err = vp10_highbd_get_y_sse(sd, cm->frame_to_show);
|
|
} else {
|
|
filt_err = vp10_get_y_sse(sd, cm->frame_to_show);
|
|
}
|
|
#else
|
|
filt_err = vp10_get_y_sse(sd, cm->frame_to_show);
|
|
#endif // CONFIG_VP9_HIGHBITDEPTH
|
|
|
|
// Re-instate the unfiltered frame
|
|
vpx_yv12_copy_y(&cpi->last_frame_uf, cm->frame_to_show);
|
|
|
|
return filt_err;
|
|
}
|
|
|
|
int vp10_search_filter_level(const YV12_BUFFER_CONFIG *sd, VP10_COMP *cpi,
|
|
int partial_frame, double *best_cost_ret) {
|
|
const VP10_COMMON *const cm = &cpi->common;
|
|
const struct loopfilter *const lf = &cm->lf;
|
|
const int min_filter_level = 0;
|
|
const int max_filter_level = vp10_get_max_filter_level(cpi);
|
|
int filt_direction = 0;
|
|
int64_t best_err;
|
|
int filt_best;
|
|
MACROBLOCK *x = &cpi->td.mb;
|
|
|
|
// Start the search at the previous frame filter level unless it is now out of
|
|
// range.
|
|
int filt_mid = clamp(lf->filter_level, min_filter_level, max_filter_level);
|
|
int filter_step = filt_mid < 16 ? 4 : filt_mid / 4;
|
|
// Sum squared error at each filter level
|
|
int64_t ss_err[MAX_LOOP_FILTER + 1];
|
|
|
|
// Set each entry to -1
|
|
memset(ss_err, 0xFF, sizeof(ss_err));
|
|
|
|
// Make a copy of the unfiltered / processed recon buffer
|
|
vpx_yv12_copy_y(cm->frame_to_show, &cpi->last_frame_uf);
|
|
|
|
best_err = try_filter_frame(sd, cpi, filt_mid, partial_frame);
|
|
filt_best = filt_mid;
|
|
ss_err[filt_mid] = best_err;
|
|
|
|
while (filter_step > 0) {
|
|
const int filt_high = VPXMIN(filt_mid + filter_step, max_filter_level);
|
|
const int filt_low = VPXMAX(filt_mid - filter_step, min_filter_level);
|
|
|
|
// Bias against raising loop filter in favor of lowering it.
|
|
int64_t bias = (best_err >> (15 - (filt_mid / 8))) * filter_step;
|
|
|
|
if ((cpi->oxcf.pass == 2) && (cpi->twopass.section_intra_rating < 20))
|
|
bias = (bias * cpi->twopass.section_intra_rating) / 20;
|
|
|
|
// yx, bias less for large block size
|
|
if (cm->tx_mode != ONLY_4X4)
|
|
bias >>= 1;
|
|
|
|
if (filt_direction <= 0 && filt_low != filt_mid) {
|
|
// Get Low filter error score
|
|
if (ss_err[filt_low] < 0) {
|
|
ss_err[filt_low] = try_filter_frame(sd, cpi, filt_low, partial_frame);
|
|
}
|
|
// If value is close to the best so far then bias towards a lower loop
|
|
// filter value.
|
|
if ((ss_err[filt_low] - bias) < best_err) {
|
|
// Was it actually better than the previous best?
|
|
if (ss_err[filt_low] < best_err)
|
|
best_err = ss_err[filt_low];
|
|
|
|
filt_best = filt_low;
|
|
}
|
|
}
|
|
|
|
// Now look at filt_high
|
|
if (filt_direction >= 0 && filt_high != filt_mid) {
|
|
if (ss_err[filt_high] < 0) {
|
|
ss_err[filt_high] = try_filter_frame(sd, cpi, filt_high, partial_frame);
|
|
}
|
|
// Was it better than the previous best?
|
|
if (ss_err[filt_high] < (best_err - bias)) {
|
|
best_err = ss_err[filt_high];
|
|
filt_best = filt_high;
|
|
}
|
|
}
|
|
|
|
// Half the step distance if the best filter value was the same as last time
|
|
if (filt_best == filt_mid) {
|
|
filter_step /= 2;
|
|
filt_direction = 0;
|
|
} else {
|
|
filt_direction = (filt_best < filt_mid) ? -1 : 1;
|
|
filt_mid = filt_best;
|
|
}
|
|
}
|
|
if (best_cost_ret)
|
|
*best_cost_ret = RDCOST_DBL(x->rdmult, x->rddiv, 0, best_err);
|
|
return filt_best;
|
|
}
|
|
|
|
#if !CONFIG_LOOP_RESTORATION
|
|
void vp10_pick_filter_level(const YV12_BUFFER_CONFIG *sd, VP10_COMP *cpi,
|
|
LPF_PICK_METHOD method) {
|
|
VP10_COMMON *const cm = &cpi->common;
|
|
struct loopfilter *const lf = &cm->lf;
|
|
|
|
lf->sharpness_level = cm->frame_type == KEY_FRAME ? 0
|
|
: cpi->oxcf.sharpness;
|
|
|
|
if (method == LPF_PICK_MINIMAL_LPF && lf->filter_level) {
|
|
lf->filter_level = 0;
|
|
} else if (method >= LPF_PICK_FROM_Q) {
|
|
const int min_filter_level = 0;
|
|
const int max_filter_level = vp10_get_max_filter_level(cpi);
|
|
const int q = vp10_ac_quant(cm->base_qindex, 0, cm->bit_depth);
|
|
// These values were determined by linear fitting the result of the
|
|
// searched level, filt_guess = q * 0.316206 + 3.87252
|
|
#if CONFIG_VP9_HIGHBITDEPTH
|
|
int filt_guess;
|
|
switch (cm->bit_depth) {
|
|
case VPX_BITS_8:
|
|
filt_guess = ROUND_POWER_OF_TWO(q * 20723 + 1015158, 18);
|
|
break;
|
|
case VPX_BITS_10:
|
|
filt_guess = ROUND_POWER_OF_TWO(q * 20723 + 4060632, 20);
|
|
break;
|
|
case VPX_BITS_12:
|
|
filt_guess = ROUND_POWER_OF_TWO(q * 20723 + 16242526, 22);
|
|
break;
|
|
default:
|
|
assert(0 && "bit_depth should be VPX_BITS_8, VPX_BITS_10 "
|
|
"or VPX_BITS_12");
|
|
return;
|
|
}
|
|
#else
|
|
int filt_guess = ROUND_POWER_OF_TWO(q * 20723 + 1015158, 18);
|
|
#endif // CONFIG_VP9_HIGHBITDEPTH
|
|
if (cm->frame_type == KEY_FRAME)
|
|
filt_guess -= 4;
|
|
lf->filter_level = clamp(filt_guess, min_filter_level, max_filter_level);
|
|
} else {
|
|
lf->filter_level = vp10_search_filter_level(
|
|
sd, cpi, method == LPF_PICK_FROM_SUBIMAGE, NULL);
|
|
}
|
|
}
|
|
#endif // !CONFIG_LOOP_RESTORATION
|