Commit 89ee460e authored by Jingning Han's avatar Jingning Han Committed by Gerrit Code Review
Browse files

Merge "Refactor integral projection based motion estimation"

parents c4cb8059 73a00d32
......@@ -519,13 +519,14 @@ void vp9_set_vbp_thresholds(VP9_COMP *cpi, int q) {
#endif
#if GLOBAL_MOTION
static int vector_match(int16_t *ref, int16_t *src) {
static int vector_match(int16_t *ref, int16_t *src, int length) {
int best_sad = INT_MAX;
int this_sad;
int d;
int center, offset = 0;
for (d = 0; d <= 64; d += 16) {
this_sad = vp9_vector_sad(&ref[d], src, 64);
int bw = length; // redundant variable, to be changed in the experiments.
for (d = 0; d <= bw; d += 16) {
this_sad = vp9_vector_sad(&ref[d], src, length);
if (this_sad < best_sad) {
best_sad = this_sad;
offset = d;
......@@ -536,9 +537,9 @@ static int vector_match(int16_t *ref, int16_t *src) {
for (d = -8; d <= 8; d += 16) {
int this_pos = offset + d;
// check limit
if (this_pos < 0 || this_pos > 64)
if (this_pos < 0 || this_pos > bw)
continue;
this_sad = vp9_vector_sad(&ref[this_pos], src, 64);
this_sad = vp9_vector_sad(&ref[this_pos], src, length);
if (this_sad < best_sad) {
best_sad = this_sad;
center = this_pos;
......@@ -549,9 +550,9 @@ static int vector_match(int16_t *ref, int16_t *src) {
for (d = -4; d <= 4; d += 8) {
int this_pos = offset + d;
// check limit
if (this_pos < 0 || this_pos > 64)
if (this_pos < 0 || this_pos > bw)
continue;
this_sad = vp9_vector_sad(&ref[this_pos], src, 64);
this_sad = vp9_vector_sad(&ref[this_pos], src, length);
if (this_sad < best_sad) {
best_sad = this_sad;
center = this_pos;
......@@ -562,9 +563,9 @@ static int vector_match(int16_t *ref, int16_t *src) {
for (d = -2; d <= 2; d += 4) {
int this_pos = offset + d;
// check limit
if (this_pos < 0 || this_pos > 64)
if (this_pos < 0 || this_pos > bw)
continue;
this_sad = vp9_vector_sad(&ref[this_pos], src, 64);
this_sad = vp9_vector_sad(&ref[this_pos], src, length);
if (this_sad < best_sad) {
best_sad = this_sad;
center = this_pos;
......@@ -575,16 +576,16 @@ static int vector_match(int16_t *ref, int16_t *src) {
for (d = -1; d <= 1; d += 2) {
int this_pos = offset + d;
// check limit
if (this_pos < 0 || this_pos > 64)
if (this_pos < 0 || this_pos > bw)
continue;
this_sad = vp9_vector_sad(&ref[this_pos], src, 64);
this_sad = vp9_vector_sad(&ref[this_pos], src, length);
if (this_sad < best_sad) {
best_sad = this_sad;
center = this_pos;
}
}
return (center - 32);
return (center - (bw >> 1));
}
static const MV search_pos[9] = {
......@@ -592,16 +593,18 @@ static const MV search_pos[9] = {
{1, -1}, {1, 0}, {1, 1},
};
static void motion_estimation(VP9_COMP *cpi, MACROBLOCK *x) {
static void motion_estimation(VP9_COMP *cpi, MACROBLOCK *x,
BLOCK_SIZE bsize) {
MACROBLOCKD *xd = &x->e_mbd;
DECLARE_ALIGNED(16, int16_t, hbuf[128]);
DECLARE_ALIGNED(16, int16_t, vbuf[128]);
DECLARE_ALIGNED(16, int16_t, src_hbuf[64]);
DECLARE_ALIGNED(16, int16_t, src_vbuf[64]);
int idx;
const int stride = 64;
const int search_width = 128;
const int search_height = 128;
const int bw = 4 << b_width_log2_lookup[bsize];
const int bh = 4 << b_height_log2_lookup[bsize];
const int search_width = bw << 1;
const int search_height = bh << 1;
const int src_stride = x->plane[0].src.stride;
const int ref_stride = xd->plane[0].pre[0].stride;
uint8_t const *ref_buf, *src_buf;
......@@ -610,34 +613,33 @@ static void motion_estimation(VP9_COMP *cpi, MACROBLOCK *x) {
MV this_mv;
// Set up prediction 1-D reference set
ref_buf = xd->plane[0].pre[0].buf + (-32);
ref_buf = xd->plane[0].pre[0].buf - (bw >> 1);
for (idx = 0; idx < search_width; idx += 16) {
vp9_int_pro_row(&hbuf[idx], ref_buf, ref_stride, 64);
vp9_int_pro_row(&hbuf[idx], ref_buf, ref_stride, bh);
ref_buf += 16;
}
ref_buf = xd->plane[0].pre[0].buf + (-32) * ref_stride;
ref_buf = xd->plane[0].pre[0].buf - (bh >> 1) * ref_stride;
for (idx = 0; idx < search_height; ++idx) {
vbuf[idx] = vp9_int_pro_col(ref_buf, 64);
vbuf[idx] = vp9_int_pro_col(ref_buf, bw);
ref_buf += ref_stride;
}
// Set up src 1-D reference set
for (idx = 0; idx < stride; idx += 16) {
for (idx = 0; idx < bw; idx += 16) {
src_buf = x->plane[0].src.buf + idx;
vp9_int_pro_row(&src_hbuf[idx], src_buf, src_stride, 64);
vp9_int_pro_row(&src_hbuf[idx], src_buf, src_stride, bh);
}
src_buf = x->plane[0].src.buf;
for (idx = 0; idx < stride; ++idx) {
src_vbuf[idx] = vp9_int_pro_col(src_buf, 64);
for (idx = 0; idx < bh; ++idx) {
src_vbuf[idx] = vp9_int_pro_col(src_buf, bw);
src_buf += src_stride;
}
// Find the best match per 1-D search
tmp_mv->col = vector_match(hbuf, src_hbuf);
tmp_mv->row = vector_match(vbuf, src_vbuf);
tmp_mv->col = vector_match(hbuf, src_hbuf, bw);
tmp_mv->row = vector_match(vbuf, src_vbuf, bh);
best_sad = INT_MAX;
this_mv = *tmp_mv;
......@@ -648,8 +650,8 @@ static void motion_estimation(VP9_COMP *cpi, MACROBLOCK *x) {
(search_pos[idx].row + this_mv.row) * ref_stride +
(search_pos[idx].col + this_mv.col);
this_sad = cpi->fn_ptr[BLOCK_64X64].sdf(src_buf, src_stride,
ref_buf, ref_stride);
this_sad = cpi->fn_ptr[bsize].sdf(src_buf, src_stride,
ref_buf, ref_stride);
if (this_sad < best_sad) {
best_sad = this_sad;
tmp_mv->row = search_pos[idx].row + this_mv.row;
......@@ -717,7 +719,7 @@ static void choose_partitioning(VP9_COMP *cpi,
mbmi->interp_filter = BILINEAR;
#if GLOBAL_MOTION
motion_estimation(cpi, x);
motion_estimation(cpi, x, BLOCK_64X64);
#endif
vp9_build_inter_predictors_sb(xd, mi_row, mi_col, BLOCK_64X64);
......
......@@ -100,22 +100,14 @@ int16_t vp9_int_pro_col_sse2(uint8_t const *ref, const int width) {
__m128i src_line = _mm_load_si128((const __m128i *)ref);
__m128i s0 = _mm_sad_epu8(src_line, zero);
__m128i s1;
(void) width; // width = 64
int i;
ref += 16;
src_line = _mm_load_si128((const __m128i *)ref);
s1 = _mm_sad_epu8(src_line, zero);
s0 = _mm_adds_epu16(s0, s1);
ref += 16;
src_line = _mm_load_si128((const __m128i *)ref);
s1 = _mm_sad_epu8(src_line, zero);
s0 = _mm_adds_epu16(s0, s1);
ref += 16;
src_line = _mm_load_si128((const __m128i *)ref);
s1 = _mm_sad_epu8(src_line, zero);
s0 = _mm_adds_epu16(s0, s1);
for (i = 16; i < width; i += 16) {
ref += 16;
src_line = _mm_load_si128((const __m128i *)ref);
s1 = _mm_sad_epu8(src_line, zero);
s0 = _mm_adds_epu16(s0, s1);
}
s1 = _mm_srli_si128(s0, 8);
s0 = _mm_adds_epu16(s0, s1);
......@@ -136,8 +128,6 @@ int vp9_vector_sad_sse2(int16_t const *ref, int16_t const *src,
diff = _mm_xor_si128(diff, sign);
sum = _mm_sub_epi16(diff, sign);
(void) width; // width = 64;
ref += 8;
src += 8;
......@@ -145,7 +135,7 @@ int vp9_vector_sad_sse2(int16_t const *ref, int16_t const *src,
v1 = _mm_unpackhi_epi16(sum, zero);
sum = _mm_add_epi32(v0, v1);
for (idx = 1; idx < 8; ++idx) {
for (idx = 8; idx < width; idx += 8) {
v0 = _mm_loadu_si128((const __m128i *)ref);
v1 = _mm_load_si128((const __m128i *)src);
diff = _mm_subs_epi16(v0, v1);
......
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment