How to serialize OpenCV Mat with boost XML archive

Question

I am trying to serialize an object which has a cv::Mat as a member, I've found this SO page but when I try it I get the following error:

usr/include/boost/archive/basic_xml_iarchive.hpp:70:9: error: no matching function for call to ‘assertion_failed(mpl_::failed************ boost::serialization::is_wrapper::************) note: template int mpl_::assertion_failed(typename mpl_::assert::type) /usr/include/boost/archive/basic_xml_iarchive.hpp: In member function ‘void boost::archive::basic_xml_iarchive::load_override(T&, int) [with T = unsigned char, Archive = boost::archive::xml_iarchive]’:’

In the example on the linked page they use a binary archive, whereas I'm using an xml one, could this be causing the problem?

philipvr · Accepted Answer · 2018-04-27 17:47:28Z

Need to add this somewhere in a included header:

namespace boost {
  namespace serialization {


    template<class Archive>
    inline void serialize(Archive & ar, cv::Mat& m, const unsigned int version) {
      int cols = m.cols;
      int rows = m.rows;
      size_t elemSize = m.elemSize();
      size_t elemType = m.type();

      ar & BOOST_SERIALIZATION_NVP(cols);
      ar & BOOST_SERIALIZATION_NVP(rows);
      ar & BOOST_SERIALIZATION_NVP(elemSize);
      ar & BOOST_SERIALIZATION_NVP(elemType); // element type.

      if(m.type() != elemType || m.rows != rows || m.cols != cols) {
        m = cv::Mat(rows, cols, elemType, cv::Scalar(0));
      }

      size_t dataSize = cols * rows * elemSize;
      cout << " datasize is " << dataSize;


      for (size_t dc = 0; dc < dataSize; dc++) {
        std::stringstream ss;
        ss << "elem_"<<dc;
        ar & boost::serialization::make_nvp(ss.str().c_str(), m.data[dc]);
      }

    }
  }
}

Collectives™ on Stack Overflow

How to serialize OpenCV Mat with boost XML archive

1 Answer 1

Comments

Your Answer

Linked

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

Comments

Your Answer

Sign up or log in

Post as a guest

Linked

Related