开发者

What is the fastest way to convert an image to pure black and white in C#?

Before someone flames me for not checking the other posts, I have done that already. I have a specific question about one of the techniques that exist to convert an image to grayscale.

I have read through the other posts on SO and basically copied technique 3 (ColorMatrix) from the tutorial at this site. It works very, very, quickly.

The problem I'm having is that what I need is a pure BW image. (i.e.: if average(R,B,G) > 200 => white else black). I have implemented it the straightforward way, and it actually takes nearly 10x as long as the grayscale algorithm from the tutorial.

I am no expert with image processing, and was wondering if there is any way to modify the snippet from the tutorial to convert images to pure black and white. If not, any efficient way would do.

EDIT:

Here's the code I'm using for the black and white (no brain approach):

public Bitmap make_bw(Bitmap original) {

    Bitmap output = new Bitmap(original.Width, original.Height);

    for (int i = 0; i < original.Width; i++) {

        for (int j = 0; j < origi开发者_开发知识库nal.Height; j++) {

            Color c = original.GetPixel(i, j);

            int average = ((c.R + c.B + c.G) / 3);

            if (average < 200)
                output.SetPixel(i, j, Color.Black);

            else
                output.SetPixel(i, j, Color.White);

        }
    }

    return output;       

}


Read this, same problem: What would be a good TRUE black and white colormatrix?


I'm using this simple code:

public Bitmap GrayScaleFilter(Bitmap image)
{
    Bitmap grayScale = new Bitmap(image.Width, image.Height);

    for (Int32 y = 0; y < grayScale.Height; y++)
        for (Int32 x = 0; x < grayScale.Width; x++)
        {
            Color c = image.GetPixel(x, y);

            Int32 gs = (Int32)(c.R * 0.3 + c.G * 0.59 + c.B * 0.11);

            grayScale.SetPixel(x, y, Color.FromArgb(gs, gs, gs));
        }
    return grayScale;
}


This guy makes three tests, the fastests being the ColorMatrix as mentioned here. External link about image conversion

BUT one of the commenters claims that the reason the direct and unsafe pixel access method is slower is calls to .Width and .Height in the bitmaps and copying these values to local variables makes direct pixel access substantially faster.

I tested this and it seems that the unsafe method is approximately 2.5 times faster than using the ColorMatrix. But it is arguably also more involved and obviously it requires that unsafe code is allowed in the project which might not be desirable.

0

上一篇:

下一篇:

精彩评论

暂无评论...
验证码 换一张
取 消

最新问答

问答排行榜