[PATCH v3 3/3] ipa: rkisp1: algorithms: Add crosstalk algorithm
Paul Elder
paul.elder at ideasonboard.com
Fri May 17 08:09:06 CEST 2024
On Wed, May 15, 2024 at 11:42:18AM +0200, Stefan Klug wrote:
> Hi Paul,
>
> thanks for the patch.
>
> On Tue, May 07, 2024 at 03:10:16PM +0900, Paul Elder wrote:
> > Add an algorithm module to the rkisp1 IPA for crosstalk correction.
> >
> > Signed-off-by: Paul Elder <paul.elder at ideasonboard.com>
> >
> > ---
> > This patch depends on "[PATCH] libcamera: utils: Add a helper to convert
> > floating-point to fixed-point"
> >
> > Changes in v3:
> > - read ccm offsets from tuning data, and write these offsets to the
> > parameters buffer
> > - make parseYaml return void, as it should fill in default data if
> > unable to read, thus never failing
> >
> > Changes in v2:
> > - rename ctk to ccm
> > - reset the matrix interpolator to identity matrix if failed to read
> > from tuning file
> > ---
> > src/ipa/rkisp1/algorithms/ccm.cpp | 116 ++++++++++++++++++++++++++
> > src/ipa/rkisp1/algorithms/ccm.h | 44 ++++++++++
> > src/ipa/rkisp1/algorithms/meson.build | 1 +
> > 3 files changed, 161 insertions(+)
> > create mode 100644 src/ipa/rkisp1/algorithms/ccm.cpp
> > create mode 100644 src/ipa/rkisp1/algorithms/ccm.h
> >
> > diff --git a/src/ipa/rkisp1/algorithms/ccm.cpp b/src/ipa/rkisp1/algorithms/ccm.cpp
> > new file mode 100644
> > index 00000000..11fa5ec2
> > --- /dev/null
> > +++ b/src/ipa/rkisp1/algorithms/ccm.cpp
> > @@ -0,0 +1,116 @@
> > +/* SPDX-License-Identifier: LGPL-2.1-or-later */
> > +/*
> > + * Copyright (C) 2024, Ideas On Board
> > + *
> > + * ccm.cpp - RkISP1 Cross Talk Correction control algorithm
> > + */
> > +
> > +#include "ccm.h"
> > +
> > +#include <algorithm>
> > +#include <chrono>
> > +#include <cmath>
> > +#include <tuple>
> > +#include <vector>
> > +
> > +#include <libcamera/base/log.h>
> > +#include <libcamera/base/utils.h>
> > +
> > +#include <libcamera/ipa/core_ipa_interface.h>
> > +
> > +#include "libcamera/internal/yaml_parser.h"
> > +
> > +#include "libipa/matrix_interpolator.h"
> > +
> > +/**
> > + * \file ctk.h
> > + */
> > +
> > +namespace libcamera {
> > +
> > +namespace ipa::rkisp1::algorithms {
> > +
> > +/**
> > + * \class Ccm
> > + * \brief A cross talk correction algorithm
> > + */
> > +
> > +LOG_DEFINE_CATEGORY(RkISP1Ccm)
> > +
> > +void Ccm::parseYaml(const YamlObject &tuningData)
> > +{
> > + int ret = ccm_.readYaml(tuningData["ccms"]);
> > + if (ret < 0) {
> > + LOG(RkISP1Ccm, Warning)
> > + << "Failed to parse 'ccms' "
> > + << "parameter from tuning file; falling back to unit matrix";
> > + ccm_.reset();
> > + }
> > +
> > + ret = offsets_.readYaml(tuningData["offsets"]);
>
> Ahh I should have read that patch first. This explains how you solved
> the "more data per temperature" question. We can also postpone that for
> now and see what we really need.
>
> > + if (ret < 0) {
> > + LOG(RkISP1Ccm, Warning)
> > + << "Failed to parse 'offsets' "
> > + << "parameter from tuning file; falling back to zero offsets";
> > + const std::map<unsigned int, Matrix<int16_t, 3, 1>> offsets = {
> > + { 0, Matrix<int16_t, 3, 1>({ 0, 0, 0 }) }
> > + };
>
> Oh, beeing able to use the interpolator for the offsets is nice. Having
> to do manual handling of the "identity" case doesn't feel too good. As
> the identity matrix is only defined for square matrices what about
> initializing the matrix to identity only if R==C in the constructor and
> to zero otherwise? Or is that too much magic?
Oh, good idea. We can do that.
>
> > + MatrixInterpolator<int16_t, 3, 1> interpolator(offsets);
> > + offsets_ = interpolator;
> > + }
> > +}
> > +
> > +/**
> > + * \copydoc libcamera::ipa::Algorithm::init
> > + */
> > +int Ccm::init([[maybe_unused]] IPAContext &context, const YamlObject &tuningData)
> > +{
> > + parseYaml(tuningData);
> > + return 0;
> > +}
> > +
> > +void Ccm::setParameters(rkisp1_params_cfg *params,
> > + const Matrix<double, 3, 3> &matrix,
> > + const Matrix<int16_t, 3, 1> &offsets)
> > +{
> > + struct rkisp1_cif_isp_ctk_config &config = params->others.ctk_config;
> > +
> > + /*
> > + * 4 bit integer and 7 bit fractional, ranging from -8 (0x400) to
> > + * +7.992 (0x3FF)
> > + */
> > + for (unsigned int i = 0; i < 3; i++)
> > + for (unsigned int j = 0; j < 3; j++)
> > + config.coeff[i][j] =
> > + utils::floatingToFixedPoint<4, 7, uint16_t, double>(matrix.get(i, j));
> > +
> > + for (unsigned int i = 0; i < 3; i++)
> > + config.ct_offset[i] = offsets.get(i, 0) & 0xFFF;
>
> If we go for floats in the matrix, we should handle the offsets the same
> way.
The offsets are integers between (and including) -2048 to 2047 so the
floats don't really apply. The tuning tool with output these values
directly.
>
> > +
> > + LOG(RkISP1Ccm, Debug) << "Setting matrix " << matrix.toString();
> > + LOG(RkISP1Ccm, Debug) << "Setting offsets " << offsets.toString();
>
> Just having to write
> LOG(RkISP1Ccm, Debug) << "Setting matrix " << matrix;
> would be beautiful :-)
Indeed :)
Thanks,
Paul
>
> > +
> > + params->module_en_update |= RKISP1_CIF_ISP_MODULE_CTK;
> > + params->module_ens |= RKISP1_CIF_ISP_MODULE_CTK;
> > + params->module_cfg_update |= RKISP1_CIF_ISP_MODULE_CTK;
> > +}
> > +
> > +/**
> > + * \copydoc libcamera::ipa::Algorithm::prepare
> > + */
> > +void Ccm::prepare(IPAContext &context, [[maybe_unused]] const uint32_t frame,
> > + [[maybe_unused]] IPAFrameContext &frameContext,
> > + rkisp1_params_cfg *params)
> > +{
> > + uint32_t ct = context.activeState.awb.temperatureK;
> > + Matrix<double, 3, 3> ccm = ccm_.get(ct);
> > + Matrix<int16_t, 3, 1> offsets = offsets_.get(ct);
> > +
> > + setParameters(params, ccm, offsets);
> > +}
> > +
> > +REGISTER_IPA_ALGORITHM(Ccm, "Ccm")
> > +
> > +} /* namespace ipa::rkisp1::algorithms */
> > +
> > +} /* namespace libcamera */
> > diff --git a/src/ipa/rkisp1/algorithms/ccm.h b/src/ipa/rkisp1/algorithms/ccm.h
> > new file mode 100644
> > index 00000000..dcab5cd3
> > --- /dev/null
> > +++ b/src/ipa/rkisp1/algorithms/ccm.h
> > @@ -0,0 +1,44 @@
> > +/* SPDX-License-Identifier: LGPL-2.1-or-later */
> > +/*
> > + * Copyright (C) 2024, Ideas On Board
> > + *
> > + * ccm.h - RkISP1 Cross Talk Correction control algorithm
> > + */
> > +
> > +#pragma once
> > +
> > +#include <linux/rkisp1-config.h>
> > +
> > +#include "libipa/matrix.h"
> > +#include "libipa/matrix_interpolator.h"
> > +
> > +#include "algorithm.h"
> > +
> > +namespace libcamera {
> > +
> > +namespace ipa::rkisp1::algorithms {
> > +
> > +class Ccm : public Algorithm
> > +{
> > +public:
> > + Ccm(){};
> > + ~Ccm() = default;
> > +
> > + int init(IPAContext &context, const YamlObject &tuningData) override;
> > + void prepare(IPAContext &context, const uint32_t frame,
> > + IPAFrameContext &frameContext,
> > + rkisp1_params_cfg *params) override;
> > +
> > +private:
> > + void parseYaml(const YamlObject &tuningData);
> > + void setParameters(rkisp1_params_cfg *params,
> > + const Matrix<double, 3, 3> &matrix,
> > + const Matrix<int16_t, 3, 1> &offsets);
> > +
> > + MatrixInterpolator<double, 3, 3> ccm_;
> > + MatrixInterpolator<int16_t, 3, 1> offsets_;
> > +};
> > +
> > +} /* namespace ipa::rkisp1::algorithms */
> > +
> > +} /* namespace libcamera */
> > diff --git a/src/ipa/rkisp1/algorithms/meson.build b/src/ipa/rkisp1/algorithms/meson.build
> > index 93a48329..16de7133 100644
> > --- a/src/ipa/rkisp1/algorithms/meson.build
> > +++ b/src/ipa/rkisp1/algorithms/meson.build
> > @@ -4,6 +4,7 @@ rkisp1_ipa_algorithms = files([
> > 'agc.cpp',
> > 'awb.cpp',
> > 'blc.cpp',
> > + 'ccm.cpp',
> > 'cproc.cpp',
> > 'dpcc.cpp',
> > 'dpf.cpp',
> > --
> > 2.39.2
> >
More information about the libcamera-devel
mailing list