[python] How to add noise (Gaussian/salt and pepper etc) to image in Python with OpenCV

I am wondering if there exists some functions in Python with OpenCV or any other python image processing library that adds Gaussian or salt and pepper noise to an image? For example, in MATLAB there exists straight-forward functions that do the same job.

Or, how to add noise to an image using Python with OpenCV?

This question is related to python opencv

The answer is


The Function adds gaussian , salt-pepper , poisson and speckle noise in an image

Parameters
----------
image : ndarray
    Input image data. Will be converted to float.
mode : str
    One of the following strings, selecting the type of noise to add:

    'gauss'     Gaussian-distributed additive noise.
    'poisson'   Poisson-distributed noise generated from the data.
    's&p'       Replaces random pixels with 0 or 1.
    'speckle'   Multiplicative noise using out = image + n*image,where
                n is uniform noise with specified mean & variance.


import numpy as np
import os
import cv2
def noisy(noise_typ,image):
   if noise_typ == "gauss":
      row,col,ch= image.shape
      mean = 0
      var = 0.1
      sigma = var**0.5
      gauss = np.random.normal(mean,sigma,(row,col,ch))
      gauss = gauss.reshape(row,col,ch)
      noisy = image + gauss
      return noisy
   elif noise_typ == "s&p":
      row,col,ch = image.shape
      s_vs_p = 0.5
      amount = 0.004
      out = np.copy(image)
      # Salt mode
      num_salt = np.ceil(amount * image.size * s_vs_p)
      coords = [np.random.randint(0, i - 1, int(num_salt))
              for i in image.shape]
      out[coords] = 1

      # Pepper mode
      num_pepper = np.ceil(amount* image.size * (1. - s_vs_p))
      coords = [np.random.randint(0, i - 1, int(num_pepper))
              for i in image.shape]
      out[coords] = 0
      return out
  elif noise_typ == "poisson":
      vals = len(np.unique(image))
      vals = 2 ** np.ceil(np.log2(vals))
      noisy = np.random.poisson(image * vals) / float(vals)
      return noisy
  elif noise_typ =="speckle":
      row,col,ch = image.shape
      gauss = np.random.randn(row,col,ch)
      gauss = gauss.reshape(row,col,ch)        
      noisy = image + image * gauss
      return noisy

I don't know is there any method in Python API.But you can use this simple code to add Salt-and-Pepper noise to an image.

import numpy as np
import random
import cv2

def sp_noise(image,prob):
    '''
    Add salt and pepper noise to image
    prob: Probability of the noise
    '''
    output = np.zeros(image.shape,np.uint8)
    thres = 1 - prob 
    for i in range(image.shape[0]):
        for j in range(image.shape[1]):
            rdn = random.random()
            if rdn < prob:
                output[i][j] = 0
            elif rdn > thres:
                output[i][j] = 255
            else:
                output[i][j] = image[i][j]
    return output

image = cv2.imread('image.jpg',0) # Only for grayscale image
noise_img = sp_noise(image,0.05)
cv2.imwrite('sp_noise.jpg', noise_img)

just look at cv2.randu() or cv.randn(), it's all pretty similar to matlab already, i guess.

let's play a bit ;) :

import cv2
import numpy as np

>>> im = np.empty((5,5), np.uint8) # needs preallocated input image
>>> im
array([[248, 168,  58,   2,   1],  # uninitialized memory counts as random, too ?  fun ;) 
       [  0, 100,   2,   0, 101],
       [  0,   0, 106,   2,   0],
       [131,   2,   0,  90,   3],
       [  0, 100,   1,   0,  83]], dtype=uint8)
>>> im = np.zeros((5,5), np.uint8) # seriously now.
>>> im
array([[0, 0, 0, 0, 0],
       [0, 0, 0, 0, 0],
       [0, 0, 0, 0, 0],
       [0, 0, 0, 0, 0],
       [0, 0, 0, 0, 0]], dtype=uint8)
>>> cv2.randn(im,(0),(99))         # normal
array([[  0,  76,   0, 129,   0],
       [  0,   0,   0, 188,  27],
       [  0, 152,   0,   0,   0],
       [  0,   0, 134,  79,   0],
       [  0, 181,  36, 128,   0]], dtype=uint8)
>>> cv2.randu(im,(0),(99))         # uniform
array([[19, 53,  2, 86, 82],
       [86, 73, 40, 64, 78],
       [34, 20, 62, 80,  7],
       [24, 92, 37, 60, 72],
       [40, 12, 27, 33, 18]], dtype=uint8)

to apply it to an existing image, just generate noise in the desired range, and add it:

img = ...
noise = ...

image = img + noise