CODEWITHSUNDEEP
pythonpython-3.x
bramastaVic
bramastaVic

Reputation: 83

mirror image in python

I want to mirror an image in python but this error occur 

Exception has occurred: IndexError
index -751 is out of bounds for axis 1 with size 750
File "D:\PART 1\Miror.py", line 20, in <module>
d=img.item(i,mn,0)

this is my code

img = cv2.imread('D:\\PART 1\\gwk.jpg')

tinggi = img.shape[0]
lebar  = img.shape[1]
brightness = 100
nm=int(tinggi-1)
mn=int(lebar-1)
lebarBaru= int(lebar/2)
tinggiBaru= int(tinggi/2)
for i in np.arange(tinggiBaru):
for j in np.arange(lebarBaru):
    a=img.item(i,j,0) 
    b=img.item(i,j,1) 
    c=img.item(i,j,2) 
    d=img.item(i,mn,0) 
    e=img.item(i,mn,1) 
    f=img.item(i,mn,2) 
    img.itemset((i,j,0),d)
    img.itemset((i,j,1),e)
    img.itemset((i,j,2),f)
    img.itemset((i,mn,0),a)
    img.itemset((i,mn,1),b)
    img.itemset((i,mn,2),c)
    mn-=1

I want to mirror an image in python without using OpenCV function for mirroring an image

Upvotes: 2

Views: 12604

Answers (2)

SpghttCd
SpghttCd

Reputation: 10890

Just invert the direction of one (or both if you like) dimensions of the image array.

Example with imageio, but should work the same with similar image arrays, too, like e.g. in opencv:

import matplotlib.pyplot as plt
import imageio
im = imageio.imread('imageio:chelsea.png') 

fig, axs = plt.subplots(1, 3, sharey=True)

axs[0].imshow(im)
axs[1].imshow(im[:, ::-1, :])
axs[2].imshow(im[::-1, :, :])

enter image description here

Upvotes: 5

HansHirse
HansHirse

Reputation: 18925

If you're willing to use NumPy (OpenCV uses NumPy arrays for storing images), use it's flip method. Otherwise, you can also just use the underlying array indexing using ::-1 notation.

Here's an example for mirroring in x direction (flip horizontally), y direction (flip vertically), and both directions (flip horizontally and vertically):

import cv2
import numpy as np

# Read input image
img = cv2.imread('path/to/your/image.png')

# Mirror in x direction (flip horizontally)
imgX = np.flip(img, axis=1)
# imgX = imgX = img[:, ::-1, :]

# Mirror in y direction (flip vertically)
imgY = np.flip(img, axis=0)
# imgY = img[::-1, :, :]

# Mirror in both directions (flip horizontally and vertically)
imgXY = np.flip(img, axis=(0, 1))
# imgXY = img[::-1, ::-1, :]

# Outputs
cv2.imshow('img', img)
cv2.imshow('imgX', imgX)
cv2.imshow('imgY', imgY)
cv2.imshow('imgXY', imgXY)
cv2.waitKey(0)

(Exemplary input and outputs are omitted here...)

Hope that helps!

Upvotes: 4

Related Questions