dilation.md

Dilation

Introduction

Here in this image dilation op, we take a kernel of size 3 x 3, which has its center at (1, 1). This kernel to be applied on each pixel of image, and if at least one pixel of image has intensity value 1 ( for binary images ), then pixel at which currently kernel centered at, will get a value 1 ( foreground pixel intensity value ), resulting into increased size of foreground region of image.

Here we consider 1 to be foreground color & 0 to be background color.

But we're interested in applying dilation on color & grayscale images too. All we need to do is to consider max pixel intensity from order-X neighborhood around current center pixel & put that as new intensity at current center pixel.

Dilation is not exactly Mode Filter, because shape of kernel not confined to square matrices. Though square shaped masks only supported as of this writing.

For color image, we need to do this same thing for three different color components. Only RGB color model supported at time of writing.

Usage

• Make sure you download in.itzmeanjan.filterit.jar & add it as your project dependency.

• Code for applying Dilation

import in.itzmeanjan.filterit.ImportExportImage;
import in.itzmeanjan.filterit.morphology.Dilation;


public class Main{

	public static void main(String [] args){
		System.out.println(
            ImportExportImage.exportImage(
                new Dilation().dilate("dream.jpg", 1, 1), 
                "dilated.jpg"));
	}

}

Result

Change may not be that much prominent, but taking difference of these two images reveals change in better fashion.

Original	Dilated

Thanking you