From 29892f1c56c6657a9d650aff203912212c295bd0 Mon Sep 17 00:00:00 2001 From: Laurent Pinchart Date: Mon, 18 Nov 2024 21:07:01 +0200 Subject: ipa: libipa: colour: Use the RGB class to model RGB values The rec601LuminanceFromRGB() and estimateCCT() functions take RGB triplets as three variables. Replace them with instances of the RGB class and adapt the users accordingly. Only variables passed directly to these functions are converted to RGB instances, further conversion of IPA modules to the RGB class will be performed separately. While at it, fix a typo in the documentation of the estimateCCT() function. Signed-off-by: Laurent Pinchart Reviewed-by: Milan Zamazal --- src/ipa/ipu3/algorithms/agc.cpp | 14 ++++----- src/ipa/ipu3/algorithms/awb.cpp | 2 +- src/ipa/libipa/colours.cpp | 22 ++++++-------- src/ipa/libipa/colours.h | 6 ++-- src/ipa/rkisp1/algorithms/awb.cpp | 47 ++++++++++++++++-------------- src/ipa/rpi/controller/rpi/agc_channel.cpp | 21 +++++++------ 6 files changed, 55 insertions(+), 57 deletions(-) (limited to 'src') diff --git a/src/ipa/ipu3/algorithms/agc.cpp b/src/ipa/ipu3/algorithms/agc.cpp index 46669203..39d0aebb 100644 --- a/src/ipa/ipu3/algorithms/agc.cpp +++ b/src/ipa/ipu3/algorithms/agc.cpp @@ -178,18 +178,16 @@ Histogram Agc::parseStatistics(const ipu3_uapi_stats_3a *stats, */ double Agc::estimateLuminance(double gain) const { - double redSum = 0, greenSum = 0, blueSum = 0; + RGB sum{ 0.0 }; for (unsigned int i = 0; i < rgbTriples_.size(); i++) { - redSum += std::min(std::get<0>(rgbTriples_[i]) * gain, 255.0); - greenSum += std::min(std::get<1>(rgbTriples_[i]) * gain, 255.0); - blueSum += std::min(std::get<2>(rgbTriples_[i]) * gain, 255.0); + sum.r() += std::min(std::get<0>(rgbTriples_[i]) * gain, 255.0); + sum.g() += std::min(std::get<1>(rgbTriples_[i]) * gain, 255.0); + sum.b() += std::min(std::get<2>(rgbTriples_[i]) * gain, 255.0); } - double ySum = rec601LuminanceFromRGB(redSum * rGain_, - greenSum * gGain_, - blueSum * bGain_); - + RGB gains{{ rGain_, gGain_, bGain_ }}; + double ySum = rec601LuminanceFromRGB(sum * gains); return ySum / (bdsGrid_.height * bdsGrid_.width) / 255; } diff --git a/src/ipa/ipu3/algorithms/awb.cpp b/src/ipa/ipu3/algorithms/awb.cpp index 7c6bff09..55de05d9 100644 --- a/src/ipa/ipu3/algorithms/awb.cpp +++ b/src/ipa/ipu3/algorithms/awb.cpp @@ -412,7 +412,7 @@ void Awb::awbGreyWorld() blueGain = sumBlue.g() / (sumBlue.b() + 1); /* Color temperature is not relevant in Grey world but still useful to estimate it :-) */ - asyncResults_.temperatureK = estimateCCT(sumRed.r(), sumRed.g(), sumBlue.b()); + asyncResults_.temperatureK = estimateCCT({{ sumRed.r(), sumRed.g(), sumBlue.b() }}); /* * Gain values are unsigned integer value ranging [0, 8) with 13 bit diff --git a/src/ipa/libipa/colours.cpp b/src/ipa/libipa/colours.cpp index 9fcb53b0..6c597093 100644 --- a/src/ipa/libipa/colours.cpp +++ b/src/ipa/libipa/colours.cpp @@ -21,9 +21,7 @@ namespace ipa { /** * \brief Estimate luminance from RGB values following ITU-R BT.601 - * \param[in] r The red value - * \param[in] g The green value - * \param[in] b The blue value + * \param[in] rgb The RGB value * * This function estimates a luminance value from a triplet of Red, Green and * Blue values, following the formula defined by ITU-R Recommendation BT.601-7 @@ -31,21 +29,19 @@ namespace ipa { * * \return The estimated luminance value */ -double rec601LuminanceFromRGB(double r, double g, double b) +double rec601LuminanceFromRGB(const RGB &rgb) { - return (r * .299) + (g * .587) + (b * .114); + return (rgb.r() * .299) + (rgb.g() * .587) + (rgb.b() * .114); } /** * \brief Estimate correlated colour temperature from RGB color space input - * \param[in] red The input red value - * \param[in] green The input green value - * \param[in] blue The input blue value + * \param[in] rgb The RGB value * * This function estimates the correlated color temperature RGB color space * input. In physics and color science, the Planckian locus or black body locus * is the path or locus that the color of an incandescent black body would take - * in a particular chromaticity space as the blackbody temperature changes. + * in a particular chromaticity space as the black body temperature changes. * * If a narrow range of color temperatures is considered (those encapsulating * daylight being the most practical case) one can approximate the Planckian @@ -56,12 +52,12 @@ double rec601LuminanceFromRGB(double r, double g, double b) * * \return The estimated color temperature */ -uint32_t estimateCCT(double red, double green, double blue) +uint32_t estimateCCT(const RGB &rgb) { /* Convert the RGB values to CIE tristimulus values (XYZ) */ - double X = (-0.14282) * (red) + (1.54924) * (green) + (-0.95641) * (blue); - double Y = (-0.32466) * (red) + (1.57837) * (green) + (-0.73191) * (blue); - double Z = (-0.68202) * (red) + (0.77073) * (green) + (0.56332) * (blue); + double X = (-0.14282) * rgb.r() + (1.54924) * rgb.g() + (-0.95641) * rgb.b(); + double Y = (-0.32466) * rgb.r() + (1.57837) * rgb.g() + (-0.73191) * rgb.b(); + double Z = (-0.68202) * rgb.r() + (0.77073) * rgb.g() + (0.56332) * rgb.b(); /* Calculate the normalized chromaticity values */ double x = X / (X + Y + Z); diff --git a/src/ipa/libipa/colours.h b/src/ipa/libipa/colours.h index b42ed0ac..fa6a8b57 100644 --- a/src/ipa/libipa/colours.h +++ b/src/ipa/libipa/colours.h @@ -9,12 +9,14 @@ #include +#include "vector.h" + namespace libcamera { namespace ipa { -double rec601LuminanceFromRGB(double r, double g, double b); -uint32_t estimateCCT(double red, double green, double blue); +double rec601LuminanceFromRGB(const RGB &rgb); +uint32_t estimateCCT(const RGB &rgb); } /* namespace ipa */ diff --git a/src/ipa/rkisp1/algorithms/awb.cpp b/src/ipa/rkisp1/algorithms/awb.cpp index 5c1d9511..dbeaf815 100644 --- a/src/ipa/rkisp1/algorithms/awb.cpp +++ b/src/ipa/rkisp1/algorithms/awb.cpp @@ -192,9 +192,7 @@ void Awb::process(IPAContext &context, const rkisp1_cif_isp_stat *params = &stats->params; const rkisp1_cif_isp_awb_stat *awb = ¶ms->awb; IPAActiveState &activeState = context.activeState; - double greenMean; - double redMean; - double blueMean; + RGB rgbMeans; metadata.set(controls::AwbEnable, frameContext.awb.autoEnabled); metadata.set(controls::ColourGains, { @@ -209,9 +207,11 @@ void Awb::process(IPAContext &context, } if (rgbMode_) { - greenMean = awb->awb_mean[0].mean_y_or_g; - redMean = awb->awb_mean[0].mean_cr_or_r; - blueMean = awb->awb_mean[0].mean_cb_or_b; + rgbMeans = {{ + static_cast(awb->awb_mean[0].mean_y_or_g), + static_cast(awb->awb_mean[0].mean_cr_or_r), + static_cast(awb->awb_mean[0].mean_cb_or_b) + }}; } else { /* Get the YCbCr mean values */ double yMean = awb->awb_mean[0].mean_y_or_g; @@ -233,9 +233,11 @@ void Awb::process(IPAContext &context, yMean -= 16; cbMean -= 128; crMean -= 128; - redMean = 1.1636 * yMean - 0.0623 * cbMean + 1.6008 * crMean; - greenMean = 1.1636 * yMean - 0.4045 * cbMean - 0.7949 * crMean; - blueMean = 1.1636 * yMean + 1.9912 * cbMean - 0.0250 * crMean; + rgbMeans = {{ + 1.1636 * yMean - 0.0623 * cbMean + 1.6008 * crMean, + 1.1636 * yMean - 0.4045 * cbMean - 0.7949 * crMean, + 1.1636 * yMean + 1.9912 * cbMean - 0.0250 * crMean + }}; /* * Due to hardware rounding errors in the YCbCr means, the @@ -243,9 +245,7 @@ void Awb::process(IPAContext &context, * negative gains, messing up calculation. Prevent this by * clamping the means to positive values. */ - redMean = std::max(redMean, 0.0); - greenMean = std::max(greenMean, 0.0); - blueMean = std::max(blueMean, 0.0); + rgbMeans = rgbMeans.max(0.0); } /* @@ -253,19 +253,22 @@ void Awb::process(IPAContext &context, * divide by the gains that were used to get the raw means from the * sensor. */ - redMean /= frameContext.awb.gains.red; - greenMean /= frameContext.awb.gains.green; - blueMean /= frameContext.awb.gains.blue; + RGB gains{{ + frameContext.awb.gains.red, + frameContext.awb.gains.green, + frameContext.awb.gains.blue + }}; + rgbMeans /= gains; /* * If the means are too small we don't have enough information to * meaningfully calculate gains. Freeze the algorithm in that case. */ - if (redMean < kMeanMinThreshold && greenMean < kMeanMinThreshold && - blueMean < kMeanMinThreshold) + if (rgbMeans.r() < kMeanMinThreshold && rgbMeans.g() < kMeanMinThreshold && + rgbMeans.b() < kMeanMinThreshold) return; - activeState.awb.temperatureK = estimateCCT(redMean, greenMean, blueMean); + activeState.awb.temperatureK = estimateCCT(rgbMeans); /* Metadata shall contain the up to date measurement */ metadata.set(controls::ColourTemperature, activeState.awb.temperatureK); @@ -275,8 +278,8 @@ void Awb::process(IPAContext &context, * gain is hardcoded to 1.0. Avoid divisions by zero by clamping the * divisor to a minimum value of 1.0. */ - double redGain = greenMean / std::max(redMean, 1.0); - double blueGain = greenMean / std::max(blueMean, 1.0); + double redGain = rgbMeans.g() / std::max(rgbMeans.r(), 1.0); + double blueGain = rgbMeans.g() / std::max(rgbMeans.b(), 1.0); /* * Clamp the gain values to the hardware, which expresses gains as Q2.8 @@ -298,8 +301,8 @@ void Awb::process(IPAContext &context, LOG(RkISP1Awb, Debug) << std::showpoint - << "Means [" << redMean << ", " << greenMean << ", " << blueMean - << "], gains [" << activeState.awb.gains.automatic.red << ", " + << "Means " << rgbMeans + << ", gains [" << activeState.awb.gains.automatic.red << ", " << activeState.awb.gains.automatic.green << ", " << activeState.awb.gains.automatic.blue << "], temp " << activeState.awb.temperatureK << "K"; diff --git a/src/ipa/rpi/controller/rpi/agc_channel.cpp b/src/ipa/rpi/controller/rpi/agc_channel.cpp index 14335943..79c45973 100644 --- a/src/ipa/rpi/controller/rpi/agc_channel.cpp +++ b/src/ipa/rpi/controller/rpi/agc_channel.cpp @@ -13,6 +13,7 @@ #include #include "libipa/colours.h" +#include "libipa/vector.h" #include "../awb_status.h" #include "../device_status.h" @@ -681,12 +682,13 @@ static double computeInitialY(StatisticsPtr &stats, AwbStatus const &awb, * Note that the weights are applied by the IPA to the statistics directly, * before they are given to us here. */ - double rSum = 0, gSum = 0, bSum = 0, pixelSum = 0; + ipa::RGB sum{ 0.0 }; + double pixelSum = 0; for (unsigned int i = 0; i < stats->agcRegions.numRegions(); i++) { auto ®ion = stats->agcRegions.get(i); - rSum += std::min(region.val.rSum * gain, (maxVal - 1) * region.counted); - gSum += std::min(region.val.gSum * gain, (maxVal - 1) * region.counted); - bSum += std::min(region.val.bSum * gain, (maxVal - 1) * region.counted); + sum.r() += std::min(region.val.rSum * gain, (maxVal - 1) * region.counted); + sum.g() += std::min(region.val.gSum * gain, (maxVal - 1) * region.counted); + sum.b() += std::min(region.val.bSum * gain, (maxVal - 1) * region.counted); pixelSum += region.counted; } if (pixelSum == 0.0) { @@ -694,14 +696,11 @@ static double computeInitialY(StatisticsPtr &stats, AwbStatus const &awb, return 0; } - double ySum; /* Factor in the AWB correction if needed. */ - if (stats->agcStatsPos == Statistics::AgcStatsPos::PreWb) { - ySum = ipa::rec601LuminanceFromRGB(rSum * awb.gainR, - gSum * awb.gainG, - bSum * awb.gainB); - } else - ySum = ipa::rec601LuminanceFromRGB(rSum, gSum, bSum); + if (stats->agcStatsPos == Statistics::AgcStatsPos::PreWb) + sum *= ipa::RGB{{ awb.gainR, awb.gainR, awb.gainB }}; + + double ySum = ipa::rec601LuminanceFromRGB(sum); return ySum / pixelSum / (1 << 16); } -- cgit v1.2.1