Question

I have a design question…

I want to create my own image class called EfficientImage which will include a Load which will take a System.Drawing.Bitmap and use lock bits to convert it into a 2D byte array.

class EfficientImage
{

    private byte[,] image;
    private int width;
    private int height;

    public Load(System.Drawing.Bitmap inImage)
    {
    }

    //etc...

}

so i might have

EfficentImage ef = new EfficientImage();
ef.Load(myBitmap);

What is the best way to access the image array within the object without losing the performance gain of using LockBits in the first place? I believe having an accessor or a property would be replicating the GetPixel/SetPixel in a normal Bitmap object which is not efficient and setting the image field to public would break encapsulation

It may be that this design is entirely the wrong way to go but I would be very pleased to learn of a better/correct way.

Thanks for your help

Answer 1

I assume you want to apply some kind of filter to images, or have some other kind of processing that modifies multiple pixels.

If that is the case, I would call your class something like ImageFilter, for instance:

public abstract class ImageFilter
{
    protected abstract void ApplyFilter(byte[,] imageData);

    public Bitmap Apply(Bitmap image)
    {
        // lockbits etc
        ApplyFilter(imageData);
        // unlockbits
    }
}

public class MyFilter : ImageFilter
{
    protected override ApplyFilter(byte[,] imageData)
    {
        // work on the bits
    }
}

Bitmap newBitmap = new MyFilter().Apply(oldBitmap);

That way you can write custom filters that deal with the raw data only, while the ImageFilter abstract class deals with loading the image and its data.

Apologies if I missed the point 🙂