/*
-----------------------------------------------------------------------
Copyright: 2010-2016, iMinds-Vision Lab, University of Antwerp
2014-2016, CWI, Amsterdam
Contact: astra@uantwerpen.be
Website: http://www.astra-toolbox.com/
This file is part of the ASTRA Toolbox.
The ASTRA Toolbox is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation, either version 3 of the License, or
(at your option) any later version.
The ASTRA Toolbox is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License
along with the ASTRA Toolbox. If not, see .
-----------------------------------------------------------------------
*/
#ifdef ASTRA_CUDA
#include "astra/CudaDataOperationAlgorithm.h"
#include "../cuda/2d/algo.h"
#include "../cuda/2d/darthelper.h"
#include "../cuda/2d/arith.h"
#include "astra/AstraObjectManager.h"
using namespace std;
namespace astra {
// type of the algorithm, needed to register with CAlgorithmFactory
std::string CCudaDataOperationAlgorithm::type = "DataOperation_CUDA";
//----------------------------------------------------------------------------------------
// Constructor
CCudaDataOperationAlgorithm::CCudaDataOperationAlgorithm()
{
m_pMask = NULL;
m_bIsInitialized = false;
}
//----------------------------------------------------------------------------------------
// Destructor
CCudaDataOperationAlgorithm::~CCudaDataOperationAlgorithm()
{
}
//---------------------------------------------------------------------------------------
// Initialize - Config
bool CCudaDataOperationAlgorithm::initialize(const Config& _cfg)
{
ASTRA_ASSERT(_cfg.self);
ConfigStackCheck CC("CCudaDataOperationAlgorithm", this, _cfg);
// operation
XMLNode node = _cfg.self.getSingleNode("Operation");
ASTRA_CONFIG_CHECK(node, "CCudaDataOperationAlgorithm", "No Operation tag specified.");
m_sOperation = node.getContent();
m_sOperation.erase(std::remove(m_sOperation.begin(), m_sOperation.end(), ' '), m_sOperation.end());
CC.markNodeParsed("Operation");
// data
node = _cfg.self.getSingleNode("DataId");
ASTRA_CONFIG_CHECK(node, "CCudaDataOperationAlgorithm", "No DataId tag specified.");
vector data = node.getContentArray();
for (vector::iterator it = data.begin(); it != data.end(); ++it){
int id = StringUtil::stringToInt(*it);
m_pData.push_back(dynamic_cast(CData2DManager::getSingleton().get(id)));
}
CC.markNodeParsed("DataId");
// scalar
node = _cfg.self.getSingleNode("Scalar");
ASTRA_CONFIG_CHECK(node, "CCudaDataOperationAlgorithm", "No Scalar tag specified.");
m_fScalar = node.getContentNumericalArray();
CC.markNodeParsed("Scalar");
// Option: GPU number
m_iGPUIndex = (int)_cfg.self.getOptionNumerical("GPUindex", -1);
m_iGPUIndex = (int)_cfg.self.getOptionNumerical("GPUIndex", m_iGPUIndex);
CC.markOptionParsed("GPUindex");
if (!_cfg.self.hasOption("GPUindex"))
CC.markOptionParsed("GPUIndex");
if (_cfg.self.hasOption("MaskId")) {
int id = _cfg.self.getOptionInt("MaskId");
m_pMask = dynamic_cast(CData2DManager::getSingleton().get(id));
}
CC.markOptionParsed("MaskId");
_check();
if (!m_bIsInitialized)
return false;
return true;
}
//---------------------------------------------------------------------------------------
// Initialize - C++
//bool CCudaDartMaskAlgorithm::initialize(CFloat32VolumeData2D* _pSegmentation, int _iConn)
//{
// return false;
//}
//----------------------------------------------------------------------------------------
// Iterate
void CCudaDataOperationAlgorithm::run(int _iNrIterations)
{
// check initialized
ASTRA_ASSERT(m_bIsInitialized);
astraCUDA::setGPUIndex(m_iGPUIndex);
astraCUDA::SDimensions dims;
// We slightly abuse dims here: width/height is not necessarily a volume
dims.iVolWidth = m_pData[0]->getWidth();
dims.iVolHeight = m_pData[0]->getHeight();
if (m_sOperation == "$1*s1" || m_sOperation == "$1.*s1") // data * scalar
{
if (m_pMask == NULL)
astraCUDA::processVolCopy(m_pData[0]->getData(), m_fScalar[0], dims);
else
astraCUDA::processVolCopy(m_pData[0]->getData(), m_pMask->getDataConst(), m_fScalar[0], dims);
}
else if (m_sOperation == "$1/s1" || m_sOperation == "$1./s1") // data / scalar
{
if (m_pMask == NULL)
astraCUDA::processVolCopy(m_pData[0]->getData(), 1.0f/m_fScalar[0], dims);
else
astraCUDA::processVolCopy(m_pData[0]->getData(), m_pMask->getDataConst(), 1.0f/m_fScalar[0], dims);
}
else if (m_sOperation == "$1+s1") // data + scalar
{
astraCUDA::processVolCopy(m_pData[0]->getData(), m_fScalar[0], dims);
}
else if (m_sOperation == "$1-s1") // data - scalar
{
astraCUDA::processVolCopy(m_pData[0]->getData(), -m_fScalar[0], dims);
}
else if (m_sOperation == "$1.*$2") // data .* data
{
astraCUDA::processVolCopy(m_pData[0]->getData(), m_pData[1]->getDataConst(), dims);
}
else if (m_sOperation == "$1+$2") // data + data
{
astraCUDA::processVolCopy(m_pData[0]->getData(), m_pData[1]->getDataConst(), dims);
}
}
//----------------------------------------------------------------------------------------
// Check
bool CCudaDataOperationAlgorithm::_check()
{
// s*: 1 data + 1 scalar
// success
m_bIsInitialized = true;
return true;
}
//---------------------------------------------------------------------------------------
// Information - All
map CCudaDataOperationAlgorithm::getInformation()
{
map res;
// TODO: add PDART-specific options
return mergeMap(CAlgorithm::getInformation(), res);
}
//---------------------------------------------------------------------------------------
// Information - Specific
boost::any CCudaDataOperationAlgorithm::getInformation(std::string _sIdentifier)
{
return NULL;
}
} // namespace astra
#endif // ASTRA_CUDA