如何保存24bit BMP

Question

I have a function to capture screen and save it to bitmap (32 bit). This function works great but I also need 24 bit bitmap. I have no idea how to convert this function.

        HWND okno=GetDesktopWindow();       
        HDC _dc = GetWindowDC(okno); 
        RECT re;
        GetWindowRect( okno, & re );
        unsigned int w = re.right, h = re.bottom;

        HDC dc = CreateCompatibleDC( 0 );
        HBITMAP bm = CreateCompatibleBitmap( _dc, w, h );
        SelectObject( dc, bm );
        StretchBlt( dc, 0, 0, w, h, _dc, 0, 0, w, h, SRCCOPY );

        void * file = CreateFile(file_name, GENERIC_WRITE, 0, NULL, CREATE_ALWAYS, 0, 0 ); //create file
        void * buf = new char[ w * h * 3 ]; //buffor

        GetObject( bm, 84, buf );
        HDC ddd = GetDC( 0 );
        HDC dc2 = CreateCompatibleDC( ddd );

        tagBITMAPINFO bit_info; //bitmapinfo
        bit_info.bmiHeader.biSize = sizeof( bit_info.bmiHeader );  
        bit_info.bmiHeader.biWidth = w;
        bit_info.bmiHeader.biHeight = h;  
        bit_info.bmiHeader.biPlanes = 1;  
        bit_info.bmiHeader.biBitCount = 32;
        bit_info.bmiHeader.biCompression = 0; 
        bit_info.bmiHeader.biSizeImage = 0; 
        CreateDIBSection( dc, & bit_info, DIB_RGB_COLORS, & buf, 0, 0 );
        GetDIBits( dc, bm, 0, h, buf, & bit_info, DIB_RGB_COLORS );

        BITMAPFILEHEADER bit_header;
        bit_header.bfType = MAKEWORD( 'B', 'M' );
        bit_header.bfSize = w * h * 4 + 54;
        bit_header.bfOffBits = 54;

        BITMAPINFOHEADER bit_info_header;
        bit_info_header.biSize = 40;
        bit_info_header.biWidth = w;
        bit_info_header.biHeight = h;
        bit_info_header.biPlanes = 0;
        bit_info_header.biBitCount = 32;
        bit_info_header.biCompression = 0;
        bit_info_header.biSizeImage = w * h *4;

        DWORD r;
        WriteFile( file, & bit_header, sizeof( bit_header ), & r, NULL );
        WriteFile( file, & bit_info_header, sizeof( bit_info_header ), & r, NULL );
        WriteFile( file, buf, w * h * 4, & r, NULL );

Sorry for my english :-)

Answer 1

Study about 32 bit and 24 bit presentation of images. Every image file has two parts header and data. In normal images a upto a fixed size of bytes contains header (Usually 1024byte). Then the data starts. If the image size is WxH then you will get WxHx32 byte data. Each 32 bit contains a single pixel information so you will get R, G, B and alpha information (4x8). you write it in 24 format with only RGB data. That's all you need. I haven't find any build in function for it.