Histogram of the image with python

 We used python for creating a histogram of the image. I used the imread() function from the OpenCV library and matplotlib library to plot the histogram of the image. Also, we used the NumPy library to store the pixel value of the image and to do various operations on the image. 

Note: At the end of the article, I provided the whole code ...... and also join the telegram channel for more information : Join

This is the input image that I used  : 

• I used the below function(read_image()) to read the image from the source and convert into a row-major format from a matrix format. 

• def read_image():

    image = cv2.imread("/content/lena_face.jpg")

    print(image.shape)

    image_1d =image.flatten()

    print(image_1d.shape)

    return image_1d

            

• To generate a bin, I used one data structure of python that, is Dictionary.

• def generate_intensity():

    intensity = dict()

    for i in range(0,256):

         intensity[i] = 0

    return intensity               

• Find the probability for each bin.

• def findProbability(intensity,image):

    p = [i for i in range(256)]

    for i in range(256):

       p[i] = intensity[i]/image.shape[0]

  return p    

• Here, we generated bins and find the probability for each bin and plot it using the matplotlib library.

• #generate the x axis(bins)

  x = [i for i in range(256)]

  y = [ intensity_1[i] for i in range(256)]

  #find the probability for each bins

  p =findProbability(intensity_1,image_1)

  print(intensity_1)

  print(p)

  plt.plot(x,p)library

• After completing the above task, you should add the 100 constant value to each pixel, plot a histogram for it. Your histogram looks like the one below if you use an input image I used.

• If you want to do a different variety of tasks, please go through the below link where all kinds of operations have been explained and copy the whole program then also go through the below link. 

  Article link :

• What kind of arithmetic operation moves the whole histogram of the image to the left-hand side. Can you guess what happens with an image?  I have provided the result below for the same input image.

• Divide operation gives result. Once try it and comment below..

• What kind of changes will happen while doing down-sampling by factor-2. ?

• If you do the sampling nothing changes happen with the histogram except the count of pixels in each bin. Pixels in the histogram bin decrease respectively, therefore you couldn't see any changes in the histogram of the image.

   

Code : 

        
#Library
import numpy as np
import cv2
import math
from google.colab.patches import cv2_imshow
import matplotlib.pyplot as plt

#read image
def read_image():
  image = cv2.imread("/content/lena_face.jpg")
  print(image.shape)
  image_1d =image.flatten()
  print(image_1d.shape)
  return image_1d
  
  
def generate_intensity():
  intensity = dict()
  for i in range(0,256):
    intensity[i] = 0
  return intensity

def findProbability(intensity,image):
  p = [i for i in range(256)]
  for i in range(256):
   p[i] = intensity[i]/image.shape[0]
  return p

image_1 =read_image()
intensity_1 = generate_intensity()
print(type(intensity_1))
for i in range(image_1.shape[0]):
  intensity_1[image_1[i]] = intensity_1[image_1[i]] + 1
 
#generate the x axis(bins)
x = [i for i in range(256)]
y = [ intensity_1[i] for i in range(256)]
#find the probability for each bins
p =findProbability(intensity_1,image_1)
print(intensity_1)
print(p)
plt.plot(x,p)



#add the 100 constant value on this image.

image_2 = read_image()
intensity_2 = generate_intensity()
image_2 =image_2 +100
imag_2=255*image_2/np.max(image_2-np.min(image_2))
for i in range(image_2.shape[0]):
  intensity_2[image_2[i]] = intensity_2[image_2[i]] +1
y_probability = findProbability(intensity_2,image_2)
plt.plot(x,y_probability)

#move histogram toward left
image_3 = read_image()
intensity_3 = generate_intensity()
for i in range(image_3.shape[0]):
 image_3[i] = math.ceil(image_3[i]/10)
# imag_1d_left=255*image_3/np.max(image_3-np.min(image_3))
for i in range(image_3.shape[0]):
  intensity_3[image_3[i]] = intensity_3[image_3[i]] +1
y_probability =findProbability(intensity_3,image_3)
plt.plot(x,y_probability)

image_4 = cv2.imread('/content/lena_face.jpg')
intensity_4 = generate_intensity()
image_4 = cv2.resize(image_4,None,fx=1/2,fy=1/2,interpolation=cv2.INTER_CUBIC)
print(image_4.shape)
image_4 = image_4.flatten()
print(image_4.shape)
for i in range(image_4.shape[0]):
  intensity_4[image_4[i]] = intensity_4[image_4[i]] +1
y_probability_4 =findProbability(intensity_4,image_4)
plt.plot(x,y_probability_4)
        

The whole code runs into the google colab. It is the best platform where we can easily run the program and analyze the output.

I also write stories on  medium platform which is free for all users. Just visit once :
Medium with Debugger and follow it. 

Thanks for giving your golden time......

(Harsh Prajapati)