c# - 使用像素数组初始化 BitmapSource

Question

我正在尝试将 native 图像数据结构编码到 BitmapSource。

这是原生结构:

struct Bitmap_8Bit
        {
        public:
            unsigned char* data;
            int stride;
            int rows;
            int cols;
            int nChannels; //either 1 or 3. in case of three the order of the channels is BGR
        };

这是我的编码函数:

template<>
        inline
            System::Windows::Media::Imaging::BitmapSource^ marshal_as<System::Windows::Media::Imaging::BitmapSource^, Bitmap_8Bit>(const Bitmap_8Bit& nativeBitmap)
        {
            System::Windows::Media::PixelFormat format = System::Windows::Media::PixelFormats::Default;
            System::Windows::Media::Imaging::BitmapPalette^ palette = nullptr;
            if (nativeBitmap.nChannels == 1)
            {
                format = System::Windows::Media::PixelFormats::Gray8;
                palette = System::Windows::Media::Imaging::BitmapPalettes::Gray256;
            }
            else if (nativeBitmap.nChannels == 3)
            {
                format = System::Windows::Media::PixelFormats::Bgr24;
                palette = System::Windows::Media::Imaging::BitmapPalettes::Halftone256;
            }
            else
            {
                throw gcnew System::InvalidOperationException("Unsupported number of channels " + nativeBitmap.nChannels);
            }

            //copy data
            int pixelDataLength = nativeBitmap.rows*nativeBitmap.stride;
            System::IntPtr source = System::IntPtr(nativeBitmap.data);

            cli::array<unsigned char>^ buffer = gcnew cli::array<unsigned char>(pixelDataLength);
            System::Runtime::InteropServices::Marshal::Copy(source, buffer, 0, pixelDataLength);

             System::Windows::Media::Imaging::BitmapSource^ managedBitmap = 
                 System::Windows::Media::Imaging::BitmapSource::Create(nativeBitmap.cols, nativeBitmap.rows, 
                 96, 96, 
                 format, palette, 
                 buffer, nativeBitmap.stride);

            return managedBitmap;
        }

我正在一个简单的单 channel 案例中测试我的功能:

BSII::IP::Bitmap_8Bit native;
native.cols = 8;
native.rows = 4;
native.nChannels = 1;
native.stride = 8;
int dataLength = native.stride * native.rows;
native.data = new unsigned char[dataLength];

unsigned char gray = 0;
for (int i = 0; i < native.rows; ++i)
{
    for (int j = 0; j < native.cols; ++j)
    {
        native.data[native.stride*i + j] = gray;
        gray += 10;
    }
}

System::Windows::Media::Imaging::BitmapSource^ managed = msclr::interop::marshal_as<System::Windows::Media::Imaging::BitmapSource^>(native);

System::IO::FileStream^ stream = gcnew System::IO::FileStream("c:\\workspace\\temp\\managed.bmp", System::IO::FileMode::Create);
System::Windows::Media::Imaging::BmpBitmapEncoder^ encoder = gcnew System::Windows::Media::Imaging::BmpBitmapEncoder();
System::Windows::Media::Imaging::BitmapFrame^ bitmapFrame = System::Windows::Media::Imaging::BitmapFrame::Create(managed);
encoder->Frames->Add(bitmapFrame);
encoder->Save(stream);
stream->Close();

这行不通。当我调试它时，我看到缓冲区数组内部有正确的值，但是当我将 BitmapSource 保存到图像时，它看起来不像预期的那样。

我还尝试使用 nullptr 作为调色板而不是 Grey256 或 Halftone256，但它也没有用。

我想我在 BitmapSource 的使用中遗漏了一些东西。有什么想法吗？

谢谢，迪娜

Answer 1

使用 PngBitmapEncoder 而不是 BmpBitmapEncoder。 BmpBitmapEncoder 使图像变形。使用 PngBitmapEncoder 示例可以正常工作。

同样出于性能原因尝试使用:

BitmapSource.Create(Int32, Int32, Double, Double, PixelFormat, BitmapPalette, 
    IntPtr, Int32, Int32) 

代替

BitmapSource.Create(Int32, Int32, Double, Double, PixelFormat, BitmapPalette, 
    Array, Int32)