“Python-OpenCV” How to Add Noise on Image. Degree Adjustable .

higashi

Hi, I’m higashi.

This page introduces how to add noise on image by using python-openCV as shown below.

The above image has large noise, but it is adjustable to without being noticed.

I think it is useful for making teacher data for AI.

So, let’s get started!”

Necessary Libraries.

This work is conducted by using below libraries.

★OpenCV

★Numpy

If it is not installed on your python, please install first.

Sample Code of Adding Noise on Image

The shown below is the program.

On this program, original image file is saved as ‘base.jpg’ and added noise image will save as ‘noise.jpg’ in program execution folder.

#import libraries
import numpy as np
import cv2

#loading original image 
img=cv2.imread('base.jpg',cv2.IMREAD_GRAYSCALE)
h,w=img.shape[:2]

#adding noise(please adjust noise_level)
noise_level=150
noise=np.random.randint(0,noise_level,(h,w))
img=img+noise
cv2.imwrite('noise.jpg',img)

By changing ‘noise_level’ on the program, you can adjust the noise degree.

* Note that too large value of ‘noise_level’ will be white image.

Execution Result of Sample Code

At first, this is original Image.(noise_level=0)

noise_level=20.

noise_lebel=150

noise_lebel=300

higashi

It seems to be fine.

That’s all. Thank you!!