Scribd
Upload
60 views8 pages

Computer Vision Lab Exp 1: Group Members

The document summarizes a computer vision experiment conducted by a group of 4 students. The aim was to apply different probability density functions (PDFs) like exponential, Bernoulli, and normal to an image and observe the effects. They took an image of Lenna and applied the 3 PDFs to generate modified grayscale images. Histograms of image gradients were also computed by dividing the image into tiles and plotting pixel intensity frequencies. The results showed how each PDF transformed pixel magnitudes differently.

Uploaded by

SP Akaash Nidhiss
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
60 views8 pages

Computer Vision Lab Exp 1: Group Members

The document summarizes a computer vision experiment conducted by a group of 4 students. The aim was to apply different probability density functions (PDFs) like exponential, Bernoulli, and normal to an image and observe the effects. They took an image of Lenna and applied the 3 PDFs to generate modified grayscale images. Histograms of image gradients were also computed by dividing the image into tiles and plotting pixel intensity frequencies. The results showed how each PDF transformed pixel magnitudes differently.

Uploaded by

SP Akaash Nidhiss
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 8

Computer Vision Lab Exp 1

Group Members:
2K19/IT/008 (Akaash Nidhiss)
2K19/IT/018 (Anasuya Mithra)
2K19/SE/003 (Ahmad Habib Khan)
2K19/CO/068 (Anshuman Roy)

Aim:
Draw Probability density functions.
Take a given image or Hand drawing with black pen and apply these PDFs to 3x3 matrix of the given
image. Also compute histogram of the Image Gradients.

Code:

import cv2
import numpy as np
import matplotlib.pyplot as plt

# Lenna.png image produced by Ahmad Khan (2K19/SE/003)

img = cv2.imread("Lenna.png")[0:510, 0:510]


gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
gray1 = np.copy(gray)
gray2 = np.copy(gray)
gray3 = np.copy(gray)
x = np.linspace(0, 1, 510)

p = 0.4
lam = 0.7
mean = 0.53
dev = 0.14

y = [((dev*np.sqrt(2*np.pi))**(-1))*np.exp(-(i - mean)**2/(2 * dev**2))


for i in x]
y = [i/max(y) for i in y]
ex = [np.exp(i) for i in x]
bern = [p**i/(1 - p)**(1 - i) for i in x]

# e^x distribution made by Ahmad Khan (2K19/SE/003)

for i in range(len(gray1)):
for j in range(len(ex)):
gray1[j][i] *= ex[i]

# bernoulli distribution made by Anshuman Roy (2K19/CO/068)

for i in range(len(gray2)):
for j in range(len(bern)):
gray2[j][i] *= bern[i]

# normal distribution made by Anshuman Roy (2K19/CO/068)

for i in range(len(gray3)):
for j in range(len(y)):
gray3[j][i] *= y[i]

plt.imshow(gray,cmap = 'gray')
plt.show()
plt.imshow(gray1,cmap = 'gray')
plt.show()
plt.imshow(gray2,cmap = 'gray')
plt.show()
plt.imshow(gray3,cmap = 'gray')
plt.show()

windowsize_r = 170
windowsize_c = 170

# histogram made by Akaash Nidhiss (2K19/IT/008) and Anasuya Mithra


(2K19/IT/018)

tiles = []
for r in range(0,gray.shape[0], windowsize_r):
for c in range(0,gray.shape[1], windowsize_c):
window = gray[r:r+windowsize_r,c:c+windowsize_c]
tiles.append(window)
for image in tiles:
freq = [0]*256
pixels = image.flatten()
for i in pixels:
freq[i] += 1
x = [i for i in range(256)]
plt.plot(x, freq)
plt.show()

Output:

1)
e^x distribution:

2)
bernoulli distribution:

3)
normal distribution:

4)

Histograms:

5)
6)

7)

8)
9)

10)

11)
12)

13)

Conclusion:
When we apply the various probability density functions, such as e^x, bernoulli’s and Normal/Gaussian
distribution, from the series of gray colour mapped images obtained, we see how the magnitude of each
pixel was multiplied by the pdf function applied to it.

Using this, we then produce 9 histograms by dividing the image into 9 parts, and each histogram maps
out the frequency of every pixel intensity in that window tile.

You might also like