PyNpArr_2_CppCvMat

PyNpArr_2_CppCvMat is a simple interface to send OpenCV images from Python to C++ and vice-versa

OpenCV code in Python, although practical to write, can run quite slowly. Pixel-wise operations, if not done inside OpenCV's calls, can make real-time applications impossible.

On the other hand, while it runs much faster, C++ code can be quite a pain to debug, due to the lack of a practical console.

Implementing the heavy part in C++ and running it from Python is the best of both worlds.

While sending basic types and data structures between the two is easy, sending images is not. This tool focuses on this specific problem. Although there are solutions that could handle the task, such as Boost, they tend to be quite bulky. This code was written using the ctypes Python package, which is in its standard distribution.

cpp2py contains the tools to convert from a OpenCV C++ image(cv::Mat) to a Python OpenCV image(numpy array)

py2cpp does the opposite

FILES

cpp2py files:

-PyImgWriter.hpp and PyImgWriter.cpp: C++ implementation of the PyImgWriter class. It serializes a cv::Mat's raw data, so Python can read it as an array of bytes.

-CppImgReader.py: Python implementation of the CppImgReader class. It fetches the PyImgWriter's raw image data and the original image's spatial and type attributes and rebuilds the image as a numpy array.

-Mat2Ctype.py: Responsible for corresponding cv::Mat types (constants such as cv::CV_8UC3, for example) to their ctypes counterparts. Used internally by CppImgReader.

py2cpp files:

-PyImgReader.hpp and PyImgReader.cpp: C++ implementation of the PyImgReader class. It receives a byte array containing the raw image data, together with its spatial and type attributes, so it can construct a cv::Mat from it.

-CppImgWriter.py: Python Implementation of the CppImgWriter. It serializes a numpy array that contains an image and sends it to a PyImgReader.

NpArr2CArr.py: serializes a numpy array into a ctypes array of its corresponding type. Used internally by CppImgWriter.

Np2Ctype.py: Responsible for corresponding numpy array data types to their ctypes counterparts. Used internally by NpArr2CArr.py.

Np2MatType.py: Responsible for corresponding numpy array data types to their cv:Mat types counterparts (constants such as cv::CV_8UC3, for example). Used internally by NpArr2CArr.py.

USAGE

When designing a C++ function that has to return an image to python:

Instead of:

cv::Mat* someFunction(...)
{
    cv::Mat* img = new cv::Mat(...);
    ...
    return img;
}

Do:

extern "C"
{
    PyImgWriter* someFunction(...)
    {
        cv::Mat* img = new cv::Mat(...);
        ...
        return new PyImgWriter(img);
    }
}
/* Note that ctypes only deals with C functions, hence the extern declaration.
   In order to make a member of an object visible to python,
   make a C wrapper around it
*/

Receiving the image with Python:

import ctypes as ct

lib = ct.CDLL(path_to_C_lib_containing_someFunction)
lib.someFunction.restype = ct.c_void_p
...
reader = CppImgReader(lib.someFunction(someargs))

img = reader.getImg()

When designing a C++ function that needs to receive an image from Python:

Instead of:

returntype someFunction(cv::Mat* img)

Do:

extern "C"
{
    returntype someFunction(PyImgReader* reader)
    {
        cv::Mat* img = reader->getImg();
        ...
    }
}
/* Note that ctypes only deals with C functions, hence the extern declaration.
   In order to make a member of an object visible to python,
   make a C wrapper around it
*/

Sending the image with Python:

import ctypes as ct

lib = ct.CDLL(path_to_C_lib_containing_someFunction)
...
img = someimage OR imgpath = path_to_some_image
...
writer = CppImgWriter(img=img OR imgpath=imgpath)

someresult = lib.someFunction(writer.sendImg())

DEPENDENCIES:

Development was done in an enviroment with these characteristics:

    Python modules:
       Python 3.6.8
       OpenCV 4.1.0
       ctypes 1.1.0
       numpy 1.17.0

    C++ modules:
       OpenCV 4.1.1