[opencv learning] [image pyramid]

Today, learn about the image pyramid , To put it bluntly, some methods of image zooming in and out , But it is in equal proportion .

import cv2
import numpy as np


#  Show the image , Encapsulate as a function 
def cv_show_image(name, img):
    cv2.imshow(name, img)
    cv2.waitKey(0)  #  Waiting time , In milliseconds ,0 Represents any key termination 
    cv2.destroyAllWindows()

#  Image pyramid , above size Small , Below size Big 
#  For example, an image shape yes (H,W)
#  Then the upper layer of the pyramid is (1/2 * H, 1/2 * W)
#  Then the upper layer of the pyramid is (1/4 * H, 1/4 * W)
#  Then the upper layer of the pyramid is (1/8 * H, 1/8 * W)
#  Then the upper layer of the pyramid is (1/16 * H, 1/16 * W)
#  This is a shrinking process , The way is , First, convolute with Gaussian kernel , Let each pixel contain some information of adjacent pixels , Then remove all even rows and columns 
#  This process is called down sampling , Shrunk 


#  Same thing , Upsampling means that the image becomes larger .
#  Then the next floor of the pyramid is (2*H, 2*W)
#  Then the next level of the pyramid is (4*H, 4*W)
#  Then the next level of the pyramid is (8*H, 8*W)
#  For example, expand twice [[10, 20], [30, 40]] ===>>> [[10, 0, 20, 0],[0, 0, 0, 0],[30, 0, 40, 0],[0, 0, 0, 0]]

img = cv2.imread('images/saoge.jpg', cv2.IMREAD_GRAYSCALE)
print(img.shape)  # (337, 600)
up = cv2.pyrUp(img)
print(up.shape)   # (674, 1200)
cv_show_image('up', up)
down = cv2.pyrDown(img)
print(down.shape)  # (169, 300)
cv_show_image('down', down)