flow-dev/ctypes_for_Image_processing

Sample code for basically using python and writing only image processing algorithm in C ++

ctypes_for_Image_processing

基本はpythonで組んで,画像処理アルゴリズムだけC++で書きたい用のサンプルコード

Point

• ctypesを使う
• Windows/LinuxでC++ファイルの作り方,呼び方が異なる.
• numpy配列をポインタ型にしてC++関数に渡せるが,numpyの型指定必須.
• .cppの関数は,"extern "C" __declspec(dllexport)"つけないとpythonから外部参照できない
• ctypesポインタに変換した配列をnumpy配列にする方法はとりあえず,xrangeにした.
• Image_processing.pyから実行

DLLの作り方Tips

• Windows OS

  • Download https://sourceforge.net/projects/mingw-w64/files/Toolchains%20targetting%20Win64/Personal%20Builds/rubenvb/gcc-4.8-release/
  • "x86_64-w64-mingw32-gcc-4.8.0-win64_rubenvb.7z"
  • Build Dll for Windows 64bit System.
  • Reference https://kakurasan.blogspot.jp/2015/07/debianubuntu-mingw-crosscompile.html
  • Command

    • x86_64-w64-mingw32-g++ -c Image_processing.cpp

    • x86_64-w64-mingw32-g++ -shared -o Image_processing.dll Image_processing.o

• Linux OS

  • Build .so for linux System.
  • Command

    • g++ -g -Wall -fPIC -o Image_processing.o -c Image_processing.cpp

    • g++ -shared Image_processing.o -o Image_processing.so

実行結果

python .\Image_processing.py (array([10, 20, 30], dtype=uint8), array([40, 50, 60], dtype=uint8), array([10, 20, 30, 40, 50, 60, 70, 80, 90], dtype=uint8), array([0, 0, 0, 0, 0, 0, 0, 0, 0], dtype=uint8)) (<main.LP_c_ubyte object at 0x000000000210FBC8>, <main.LP_c_ubyte object at 0x0000000002A12A48>, <main.LP_c_ubyte object at 0x0000000002AB0AC8>, <main.LP_c_ubyte object at 0x0000000002B548C8>) 210 [11 21 31 41 51 61 71 81 91]