Introduction

Digital Image Processing (DIP)

DIP is one emerging technology that is getting a lot of applications in various fields. It has its origins dating back to the early 20th century where it had a sole application in the press sector. Nowadays, it exists in photography, medicine, industry and astronomy. So, what do we mean by DIP and how do we see an image from an engineering perspective?

Well, simply stating we are doing some kinda of “processing” for a digital image. A digital image is the one where we see the image as a matrix of numbers. For instance, let’s consider an image that looks like this:

We actually see it as a matrix of numbers where each location of the image, or more technically each “pixel”, has a certain value that we call the “intensity”. Note that the number of pixels of a certain image differs based on how you sample the image for digital processing. So, the above image is actually seen as this:

So, a simple image with the word hi is treated as this matrix. For technical speaking, we’d say the intensity of the pixel in first row and first column, or mathematically (1,1), is 254 whereas I(3, 7) = 251 and in case you wonder how this relates to the image, consider looking at the below illustration where numbers representing black are highlighted in yellow. Of course you expect some errors.

One more thing to consider here. We’ll consider some operations which we need to know that some operations are linear and some others are nonlinear. For an operation to be linear, the following relation should be true:

Z (ax + by) = a Z(x) + b Z(y)

where Z is an operation, x and y are 2 images and a and b are arbitrary constants. In other words, if you apply this operation to the sum of 2 images, it’s identical to applying the operation on each image separately and then add the results together. Remember that images are actually matrices of numbers.