Scribd
Upload
17 views3 pages

Deep Learning Course - Session 12

This document discusses fundamentals of computer images including RGBA values, coordinates, and box tuples. It describes how to manipulate images through cropping, resizing, rotating, and flipping. Specific Python code examples are given to crop an image, resize it, rotate it 90 degrees, and insert a logo onto multiple images in a directory.

Uploaded by

Nguyễn Nam Anh
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
17 views3 pages

Deep Learning Course - Session 12

This document discusses fundamentals of computer images including RGBA values, coordinates, and box tuples. It describes how to manipulate images through cropping, resizing, rotating, and flipping. Specific Python code examples are given to crop an image, resize it, rotate it 90 degrees, and insert a logo onto multiple images in a directory.

Uploaded by

Nguyễn Nam Anh
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
You are on page 1/ 3

Computer Image fundamentals:

- Colors and RGBA Values


- Coordinates and Box Tuples
Manipulate Images:
- Cropping Images
- Resizing an Image
- Rotating and Flipping Images
2 extension thường dùng là .png, .jpeg
(image datatype) - - value -> Tuple(R,G,B,A)
R: Red
G: Green
B: Blue
A: Alpha ( cường độ của màu)

We use Pillow module


From PIL import ImageColor (Image Color is a sub module of Pillow)

Pixel đầu tiên là trái trên, pixel cuối cùng là phải dưới

1 tấm ảnh 3 x 3 resolution sẽ có dữ liệu là:


List[ List[Tuples] ] = [ [(RGBA) (0,0) , (RGBA) (0,1) , (RGBA) (0,2) ]
[(RGBA) (1,0) , (RGBA) (1,1) , (RGBA) (1,2) ]
[(RGBA) (2,0) , (RGBA) (2,1) , (RGBA) (2,2) ] ]
Module PIL

Box Tuples: (Left, Top, Right, Bottom)

note*: there are 2 type of path, which are ABSOLUTE PATH and RELATIVE PATH
There are some type of built-in functions in this module:

Here are the results:


=> 4000 * 6000

Example of a program to crop image:


From PIL import Image

Dog_img = image.open(“./dog.jpg“)

Cropped_dog_img = dog.img.crop( (2000, 1000, 4000, 6000) )

cropped_dog_img.save( “./cropped_dog_img.jpg” )

dog.img.paste( cropped_dog_img, (100,100) )


(here is another img), (here is the top left pixel)

We can resize an image as following:


width , height = dom_img.size

Compressed_10_dog_img = dog_img.resize( (int(width/10) , int(height/10) ))


Compressed_10_dog_img.save (“./image/dog_resize.jpg”)

Or, we can rotate the image as following:


Dog_90 = dog_img.rotate(90) (rotate 90 degree)
However, as the image is 4000 * 6000, when you rotate, some pixels may be deleted.
Therefore, you should expand the pixels:
Dog_90 = dog_img.rotate(90, expand = True)

We can get the stats of a particular pixel as following:


Pixel_at_100_100 = dog_img.getpixel( (100,100) )
print (pixel_at_100_100)

Now, let’s see an assignment to insert logo to several images:


From PIL import Image
Import os

Src_path = “your directory”


For image in os.listdir(src_path):
Img_path = os.path.join(image, src_path)
Img = image.open(img_path)
img.paste( logo, (0,0) )
If not os.path.exist(“ ./logo.image”):
os.makedirs(“ ./logo.image”)
Img.save( )
Logo = image.open(‘./image/logo1.png’) (insert your logo directory here)
Img = image.open( )

You might also like