mirror/opencv
1
0
Fork 0
mirror of https://github.com/opencv/opencv.git synced 2024-11-29 22:00:25 +08:00
Code Issues Actions Packages Projects Releases Wiki Activity

improve RGB2YCrCb_i<uchar>()

Browse Source
This commit is contained in:
k-shinotsuka 2016-08-07 12:05:27 +09:00
parent b34272f8a2
commit e37c53d13e
1 changed files with 99 additions and 89 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

188
modules/imgproc/src/color.cpp
View File

@ -2174,54 +2174,60 @@ struct RGB2YCrCb_i<uchar>
if (blueIdx==0) if (blueIdx==0)
std::swap(coeffs[0], coeffs[2]); std::swap(coeffs[0], coeffs[2]);
v_c0 = _mm_set1_epi32(coeffs[0]); short delta = 1 << (yuv_shift - 1);
v_c1 = _mm_set1_epi32(coeffs[1]); v_delta_16 = _mm_set1_epi16(delta);
v_c2 = _mm_set1_epi32(coeffs[2]); v_delta_32 = _mm_set1_epi32(delta);
v_c3 = _mm_set1_epi32(coeffs[3]); short delta2 = 1 + ColorChannel<uchar>::half() * 2;
v_c4 = _mm_set1_epi32(coeffs[4]); v_coeff = _mm_set_epi16(delta2, (short)coeffs[4], delta2, (short)coeffs[3], delta2, (short)coeffs[4], delta2, (short)coeffs[3]);
v_delta2 = _mm_set1_epi32(1 << (yuv_shift - 1)); v_shuffle2 = _mm_set_epi8(0x0, 0x0, 0x0, 0x0, 0xf, 0xe, 0xc, 0xb, 0xa, 0x8, 0x7, 0x6, 0x4, 0x3, 0x2, 0x0);
v_delta = _mm_set1_epi32(ColorChannel<uchar>::half()*(1 << yuv_shift));
v_delta = _mm_add_epi32(v_delta, v_delta2);
v_zero = _mm_setzero_si128();
haveSIMD = checkHardwareSupport(CV_CPU_SSE4_1); haveSIMD = checkHardwareSupport(CV_CPU_SSE4_1);
} }
// 16u x 8 // 16u x 8
void process(__m128i v_r, __m128i v_g, __m128i v_b, void process(__m128i* v_rgb, __m128i & v_crgb,
__m128i & v_y, __m128i & v_cr, __m128i & v_cb) const __m128i* v_rb, uchar * dst) const
{ {
__m128i v_r_p = _mm_unpacklo_epi16(v_r, v_zero); v_rgb[0] = _mm_madd_epi16(v_rgb[0], v_crgb);
__m128i v_g_p = _mm_unpacklo_epi16(v_g, v_zero); v_rgb[1] = _mm_madd_epi16(v_rgb[1], v_crgb);
__m128i v_b_p = _mm_unpacklo_epi16(v_b, v_zero); v_rgb[2] = _mm_madd_epi16(v_rgb[2], v_crgb);
v_rgb[3] = _mm_madd_epi16(v_rgb[3], v_crgb);
v_rgb[0] = _mm_hadd_epi32(v_rgb[0], v_rgb[1]);
v_rgb[2] = _mm_hadd_epi32(v_rgb[2], v_rgb[3]);
v_rgb[0] = _mm_add_epi32(v_rgb[0], v_delta_32);
v_rgb[2] = _mm_add_epi32(v_rgb[2], v_delta_32);
v_rgb[0] = _mm_srai_epi32(v_rgb[0], yuv_shift);
v_rgb[2] = _mm_srai_epi32(v_rgb[2], yuv_shift);
__m128i v_y = _mm_packs_epi32(v_rgb[0], v_rgb[2]);
__m128i v_y0 = _mm_add_epi32(_mm_mullo_epi32(v_r_p, v_c0), v_rb[0] = _mm_cvtepu8_epi16(v_rb[0]);
_mm_add_epi32(_mm_mullo_epi32(v_g_p, v_c1), v_rb[1] = _mm_cvtepu8_epi16(v_rb[1]);
_mm_mullo_epi32(v_b_p, v_c2))); v_rb[0] = _mm_sub_epi16(v_rb[0], _mm_unpacklo_epi16(v_y, v_y));
v_y0 = _mm_srli_epi32(_mm_add_epi32(v_delta2, v_y0), yuv_shift); v_rb[1] = _mm_sub_epi16(v_rb[1], _mm_unpackhi_epi16(v_y, v_y));
v_rgb[0] = _mm_unpacklo_epi16(v_rb[0], v_delta_16);
v_rgb[1] = _mm_unpackhi_epi16(v_rb[0], v_delta_16);
v_rgb[2] = _mm_unpacklo_epi16(v_rb[1], v_delta_16);
v_rgb[3] = _mm_unpackhi_epi16(v_rb[1], v_delta_16);
v_rgb[0] = _mm_madd_epi16(v_rgb[0], v_coeff);
v_rgb[1] = _mm_madd_epi16(v_rgb[1], v_coeff);
v_rgb[2] = _mm_madd_epi16(v_rgb[2], v_coeff);
v_rgb[3] = _mm_madd_epi16(v_rgb[3], v_coeff);
v_rgb[0] = _mm_srai_epi32(v_rgb[0], yuv_shift);
v_rgb[1] = _mm_srai_epi32(v_rgb[1], yuv_shift);
v_rgb[2] = _mm_srai_epi32(v_rgb[2], yuv_shift);
v_rgb[3] = _mm_srai_epi32(v_rgb[3], yuv_shift);
v_rgb[0] = _mm_packs_epi32(v_rgb[0], v_rgb[1]);
v_rgb[2] = _mm_packs_epi32(v_rgb[2], v_rgb[3]);
v_rgb[0] = _mm_packus_epi16(v_rgb[0], v_rgb[2]);
__m128i v_cr0 = _mm_mullo_epi32(_mm_sub_epi32(blueIdx == 2 ? v_r_p : v_b_p, v_y0), v_c3); v_rb[0] = _mm_unpacklo_epi16(v_y, v_rgb[0]);
__m128i v_cb0 = _mm_mullo_epi32(_mm_sub_epi32(blueIdx == 0 ? v_r_p : v_b_p, v_y0), v_c4); v_rb[1] = _mm_unpackhi_epi16(v_y, v_rgb[0]);
v_cr0 = _mm_srai_epi32(_mm_add_epi32(v_delta, v_cr0), yuv_shift);
v_cb0 = _mm_srai_epi32(_mm_add_epi32(v_delta, v_cb0), yuv_shift);
v_r_p = _mm_unpackhi_epi16(v_r, v_zero); v_rb[0] = _mm_shuffle_epi8(v_rb[0], v_shuffle2);
v_g_p = _mm_unpackhi_epi16(v_g, v_zero); v_rb[1] = _mm_shuffle_epi8(v_rb[1], v_shuffle2);
v_b_p = _mm_unpackhi_epi16(v_b, v_zero); v_rb[1] = _mm_alignr_epi8(v_rb[1], _mm_slli_si128(v_rb[0], 4), 12);
__m128i v_y1 = _mm_add_epi32(_mm_mullo_epi32(v_r_p, v_c0), _mm_storel_epi64((__m128i *)(dst), v_rb[0]);
_mm_add_epi32(_mm_mullo_epi32(v_g_p, v_c1), _mm_storeu_si128((__m128i *)(dst + 8), v_rb[1]);
_mm_mullo_epi32(v_b_p, v_c2)));
v_y1 = _mm_srli_epi32(_mm_add_epi32(v_delta2, v_y1), yuv_shift);
__m128i v_cr1 = _mm_mullo_epi32(_mm_sub_epi32(blueIdx == 2 ? v_r_p : v_b_p, v_y1), v_c3);
__m128i v_cb1 = _mm_mullo_epi32(_mm_sub_epi32(blueIdx == 0 ? v_r_p : v_b_p, v_y1), v_c4);
v_cr1 = _mm_srai_epi32(_mm_add_epi32(v_delta, v_cr1), yuv_shift);
v_cb1 = _mm_srai_epi32(_mm_add_epi32(v_delta, v_cb1), yuv_shift);
v_y = _mm_packs_epi32(v_y0, v_y1);
v_cr = _mm_packs_epi32(v_cr0, v_cr1);
v_cb = _mm_packs_epi32(v_cb0, v_cb1);
} }
void operator()(const uchar * src, uchar * dst, int n) const void operator()(const uchar * src, uchar * dst, int n) const
@ -2233,63 +2239,67 @@ struct RGB2YCrCb_i<uchar>
if (haveSIMD) if (haveSIMD)
{ {
for ( ; i <= n - 96; i += 96, src += scn * 32) __m128i v_shuffle;
{ __m128i v_crgb;
__m128i v_r0 = _mm_loadu_si128((__m128i const *)(src));
__m128i v_r1 = _mm_loadu_si128((__m128i const *)(src + 16));
__m128i v_g0 = _mm_loadu_si128((__m128i const *)(src + 32));
__m128i v_g1 = _mm_loadu_si128((__m128i const *)(src + 48));
__m128i v_b0 = _mm_loadu_si128((__m128i const *)(src + 64));
__m128i v_b1 = _mm_loadu_si128((__m128i const *)(src + 80));
if (scn == 4) if (scn == 4)
{ {
__m128i v_a0 = _mm_loadu_si128((__m128i const *)(src + 96)); if (bidx == 0)
__m128i v_a1 = _mm_loadu_si128((__m128i const *)(src + 112)); {
_mm_deinterleave_epi8(v_r0, v_r1, v_g0, v_g1, v_shuffle = _mm_set_epi8(0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0xc, 0xe, 0x8, 0xa, 0x4, 0x6, 0x0, 0x2);
v_b0, v_b1, v_a0, v_a1);
} }
else else
_mm_deinterleave_epi8(v_r0, v_r1, v_g0, v_g1, v_b0, v_b1); {
v_shuffle = _mm_set_epi8(0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0xe, 0xc, 0xa, 0x8, 0x6, 0x4, 0x2, 0x0);
}
v_crgb = _mm_set_epi16(0, (short)C2, (short)C1, (short)C0, 0, (short)C2, (short)C1, (short)C0);
for ( ; i <= n - 24; i += 24, src += scn * 8)
{
__m128i v_src[2];
v_src[0] = _mm_loadu_si128((__m128i const *)(src));
v_src[1] = _mm_loadu_si128((__m128i const *)(src + 16));
__m128i v_y0 = v_zero, v_cr0 = v_zero, v_cb0 = v_zero; __m128i v_rgb[4];
process(_mm_unpacklo_epi8(v_r0, v_zero), v_rgb[0] = _mm_cvtepu8_epi16(v_src[0]);
_mm_unpacklo_epi8(v_g0, v_zero), v_rgb[1] = _mm_cvtepu8_epi16(_mm_srli_si128(v_src[0], 8));
_mm_unpacklo_epi8(v_b0, v_zero), v_rgb[2] = _mm_cvtepu8_epi16(v_src[1]);
v_y0, v_cr0, v_cb0); v_rgb[3] = _mm_cvtepu8_epi16(_mm_srli_si128(v_src[1], 8));
__m128i v_y1 = v_zero, v_cr1 = v_zero, v_cb1 = v_zero; __m128i v_rb[2];
process(_mm_unpackhi_epi8(v_r0, v_zero), v_rb[0] = _mm_shuffle_epi8(v_src[0], v_shuffle);
_mm_unpackhi_epi8(v_g0, v_zero), v_rb[1] = _mm_shuffle_epi8(v_src[1], v_shuffle);
_mm_unpackhi_epi8(v_b0, v_zero),
v_y1, v_cr1, v_cb1);
__m128i v_y_0 = _mm_packus_epi16(v_y0, v_y1); process(v_rgb, v_crgb, v_rb, dst + i);
__m128i v_cr_0 = _mm_packus_epi16(v_cr0, v_cr1); }
__m128i v_cb_0 = _mm_packus_epi16(v_cb0, v_cb1); }
else
{
if (bidx == 0)
{
v_shuffle = _mm_set_epi8(0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x9, 0xb, 0x6, 0x8, 0x3, 0x5, 0x0, 0x2);
}
else
{
v_shuffle = _mm_set_epi8(0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0xb, 0x9, 0x8, 0x6, 0x5, 0x3, 0x2, 0x0);
}
v_crgb = _mm_set_epi16(0, (short)C2, (short)C1, (short)C0, (short)C2, (short)C1, (short)C0, 0);
for ( ; i <= n - 24; i += 24, src += scn * 8)
{
__m128i v_src[2];
v_src[0] = _mm_loadu_si128((__m128i const *)(src));
v_src[1] = _mm_loadl_epi64((__m128i const *)(src + 16));
process(_mm_unpacklo_epi8(v_r1, v_zero), __m128i v_rgb[4];
_mm_unpacklo_epi8(v_g1, v_zero), v_rgb[0] = _mm_cvtepu8_epi16(_mm_slli_si128(v_src[0], 1));
_mm_unpacklo_epi8(v_b1, v_zero), v_rgb[1] = _mm_cvtepu8_epi16(_mm_srli_si128(v_src[0], 5));
v_y0, v_cr0, v_cb0); v_rgb[2] = _mm_cvtepu8_epi16(_mm_alignr_epi8(v_src[1], v_src[0], 11));
v_rgb[3] = _mm_cvtepu8_epi16(_mm_srli_si128(v_src[1], 1));
process(_mm_unpackhi_epi8(v_r1, v_zero), __m128i v_rb[2];
_mm_unpackhi_epi8(v_g1, v_zero), v_rb[0] = _mm_shuffle_epi8(v_src[0], v_shuffle);
_mm_unpackhi_epi8(v_b1, v_zero), v_rb[1] = _mm_shuffle_epi8(_mm_alignr_epi8(v_src[1], v_src[0], 12), v_shuffle);
v_y1, v_cr1, v_cb1);
__m128i v_y_1 = _mm_packus_epi16(v_y0, v_y1); process(v_rgb, v_crgb, v_rb, dst + i);
__m128i v_cr_1 = _mm_packus_epi16(v_cr0, v_cr1); }
__m128i v_cb_1 = _mm_packus_epi16(v_cb0, v_cb1);
_mm_interleave_epi8(v_y_0, v_y_1, v_cr_0, v_cr_1, v_cb_0, v_cb_1);
_mm_storeu_si128((__m128i *)(dst + i), v_y_0);
_mm_storeu_si128((__m128i *)(dst + i + 16), v_y_1);
_mm_storeu_si128((__m128i *)(dst + i + 32), v_cr_0);
_mm_storeu_si128((__m128i *)(dst + i + 48), v_cr_1);
_mm_storeu_si128((__m128i *)(dst + i + 64), v_cb_0);
_mm_storeu_si128((__m128i *)(dst + i + 80), v_cb_1);
} }
} }
@ -2304,10 +2314,10 @@ struct RGB2YCrCb_i<uchar>
} }
} }
__m128i v_delta_16, v_delta_32;
__m128i v_coeff;
__m128i v_shuffle2;
int srccn, blueIdx, coeffs[5]; int srccn, blueIdx, coeffs[5];
__m128i v_c0, v_c1, v_c2;
__m128i v_c3, v_c4, v_delta, v_delta2;
__m128i v_zero;
bool haveSIMD; bool haveSIMD;
}; };