3.0 Image Subsampling

In this class, we will learn about Image Subsampling.

 

 

 

Sometimes, an image is too big to fit on the screen. We should reduce the image to show the whole image in the screen.

 

 

 

 

 

● Naive image sub-sampling

Naive image sub-sampling throws away even rows and columns.

 

By doing so, we can reduce the image's size by half and repeat again and again.

 

But what are the problems with this approach?

 

As the images scaled by half, the images become so blocky(pixelated, aliased).

 

 

 

 

 

 

● Gaussian (Lowpass) pre-filtering

We can add gaussian pre-filtering to prevent these problems.

 

 

When we add Gaussian pre-filtering while reducing by half,  the shape of the whole image is still maintained.

It only get blurred.

 

 

 

 

Naive Subsampling

 When we just apply Naive subsampling, the image loses its original shape.

 

 

 

Gaussian Image Pyramid

This sequence of subsampled images is called the Gaussian image Pyramid.

 

 

 

Next class, we will learn about Image Pyramid in more detail.