Create Image Histogram Manually and Efficiently in Python

Question

I want to write codes that can show the histogram of an image without using built in Matplotlib hist function.

Here is my codes:

import cv2 as cv
import numpy as np
from matplotlib import pyplot as plt

def manHist(img):
   row, col = img.shape # img is a grayscale image
   y = np.zeros((256), np.uint64)
   for i in range(0,row):
      for j in range(0,col):
         y[img[i,j]] += 1
   x = np.arange(0,256)
   plt.bar(x,y,color="gray",align="center")
   plt.show()

def main():
   img = cv.imread("C:/Users/Kadek/Documents/MATLAB/2GS.jpg")
   manHist(img)

main()

My question is, is there a more efficent way to make an array of pixel value frequency without using for loop?

There's no other way than walking through every value in the image, though a vectorized numpy solution is going to be some orders of magnitude faster — Francisco
– Francisco, Commented Oct 16, 2016 at 17:09
So, you are not using the third axis of channels as you have : y[img[i,j]]? — Divakar
– Divakar, Commented Oct 16, 2016 at 17:14
You could use collections.Counter for this, although that might not be "manual" enough for you. — kindall
– kindall, Commented Oct 16, 2016 at 17:16
@FranciscoCouzo thank you for your advice, but i don't understand how to use vectorized numpy. — Kadek Dwi Budi Utama
– Kadek Dwi Budi Utama, Commented Oct 16, 2016 at 17:21

Divakar · Accepted Answer · 2016-10-16 17:33:12Z

2

A NumPy based vectorized solution would be with np.bincount -

out = np.bincount(img.ravel(),minlength=256)

Another vectorized approach based on .sum() -

out = (img.ravel() == np.arange(256)[:,None]).sum(1)

Sample run to verify results -

In [155]: # Input image (512x512) as array
     ...: img = np.random.randint(0,255,(512,512))
     ...: 
     ...: # Original code
     ...: row, col = img.shape
     ...: y = np.zeros((256), np.uint64)
     ...: for i in range(0,row):
     ...:     for j in range(0,col):
     ...:         y[img[i,j]] += 1
     ...:         

In [156]: out1 = np.bincount(img.ravel(),minlength=256)

In [157]: out2 = (img.ravel() == np.arange(256)[:,None]).sum(1)

In [158]: np.allclose(y,out1)
Out[158]: True

In [159]: np.allclose(y,out2)
Out[159]: True

edited Oct 16, 2016 at 17:33

answered Oct 16, 2016 at 17:24

Divakar

222k19 gold badges273 silver badges374 bronze badges

Sign up to request clarification or add additional context in comments.

5 Comments

Kadek Dwi Budi Utama Over a year ago

both returning false in my codes. i wrote exactly same like yours.

Divakar Over a year ago

@KadekDwiBudiUtama You haven't answered by question in the comments, whether that should be row, col = img.shape instead or not?

Kadek Dwi Budi Utama Over a year ago

Oh, sorry. It's work after using only 1 channel! Thank's!

Divakar Over a year ago

@KadekDwiBudiUtama Kindly update your code in the question with that correction.

Divakar Over a year ago

@KadekDwiBudiUtama Took the liberty to make the edit myself. If not coherent, feel free to make additional edits.

Collectives™ on Stack Overflow

Create Image Histogram Manually and Efficiently in Python

1 Answer 1

5 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

5 Comments

Your Answer

Sign up or log in

Post as a guest

Related