In this class, we will learn about 'Image 'Filtering'.
● Image
What is an image?
An image is an array of numbers.
Color images is made up of 3 colors, Red, Green, Blue.
And actual intensity values per pixel is 0-255.
● 2D function
It is helpful to think of an image as a 2D function.
So, what is the range of f(x)?
For example, 8-bit image would have 256 values.
What kind of image transformations can we perform?
As you see in the image above, filtering changes the pixel values.
And Warping changes the pixel location.
To implement those image transformation, some function calculation is needed.
In filtering case, we should change the range of image.
In warping case, we should change the domain of image.
So, what kind of image filtering can we perform?
Point operation, which processes point, can be possible.
Neighborhood operation, which uses filtering, can also be possible.
The pictures above are examples of Point Processing.
The pictures above are the examples of filtering.
Next time, we will look Point Processing in detail.
'AI > Computer Vision Materials' 카테고리의 다른 글
2.2 Box Filter (0) | 2020.08.03 |
---|---|
2.1 Point Processing (0) | 2020.08.03 |
1.2 Application of computer vision (0) | 2020.07.31 |
1.1 What is Computer Vision? (0) | 2020.07.31 |
8.X TF-IDF (0) | 2020.07.31 |