Uniform image noise

import cv2
import numpy as np
import math
 
def uniform_noise_add(img, prob_noise, mean, variance):
 
    height, width = img.shape[:2]
 
    height, width, depth = np.shape(img)
 
    noise_array = np.zeros(20, np.float)
 
    noise_mat = np.zeros(shape = (img.shape[0],img.shape[1],img.shape[2]),dtype = np.float)
 
    count = 0
 
    b = ((2 * mean) + math.sqrt(12 * variance)) / 2
 
    a = (2 * mean) - b
 
    for i in range(-10,10):
 
        if i >= a and i <= b:
            fx = 1 / (b - a)
        else:
            fx = 0
 
        noise_array[count] = fx
 
        count += 1
 
    cdf_g = np.cumsum(noise_array)
 
    for i in range(img.shape[0]):
        for j in range(img.shape[1]):
 
            random_number = np.random.rand()
 
            count = -10
 
            for k in cdf_g:
 
                if random_number <= k:
 
                    noise_mat[i,j] = count*23
 
                    break
 
                count += 1
 
    return noise_mat
 
if  __name__ == "__main__":
 
    img = cv2.imread("Lenna.jpg")
 
    mean = 2
    variance = 2.28
    prob_noise = 0.50
 
    noise_mat = uniform_noise_add(img, prob_noise, mean, variance)
 
    img = img + noise_mat
 
    cv2.imwrite("Lenna_uniform_noise.jpg", img)