Scribd
Upload
7 views35 pages

Image Processing Techniques in Python

The document provides practical exercises on computer vision, including image manipulation techniques such as brightness adjustment, flipping, color component display, grayscale conversion, and image negation. It also covers histogram analysis, image filtering methods, edge detection, thresholding, and background subtraction. Each section includes code examples using OpenCV and Matplotlib for visual representation of the results.

Uploaded by

aniketyad2121
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
7 views35 pages

Image Processing Techniques in Python

The document provides practical exercises on computer vision, including image manipulation techniques such as brightness adjustment, flipping, color component display, grayscale conversion, and image negation. It also covers histogram analysis, image filtering methods, edge detection, thresholding, and background subtraction. Each section includes code examples using OpenCV and Matplotlib for visual representation of the results.

Uploaded by

aniketyad2121
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

computer vision

Practical 1 : - Basic Operation on Image


1A) Program to change the Brightness of Image.
Code : -
import cv2
import [Link] as plt
image = [Link]('[Link]')
image_rgb = [Link](image, cv2.COLOR_BGR2RGB)
# Adjust brightness by increasing beta by 50
adjusted = [Link](image, beta=50)
adjusted_rgb = [Link](adjusted, cv2.COLOR_BGR2RGB)
[Link](1, 2, 1)
[Link]('Original')
[Link](image_rgb)
[Link]('off')
[Link](1, 2, 2)
[Link]('Brightness Adjusted')
[Link](adjusted_rgb)
[Link]('off')
plt.tight_layout()
[Link]()
Output : -

Sheth. N. K. T. T. College T. Y. Bsc. Data Science


Computer Vision

1B) : - To flip the image around the vertical and horizontal line.
Code : -
import cv2
import [Link] as plt
img = [Link](r"C:\Users\Suyash\Desktop\vs studio code\[Link]")
if img is None:
print("Error: Image not found or path is incorrect.")
else:
img_org = [Link](img, cv2.COLOR_BGR2RGB)
[Link](1, 3, 1)
[Link]('Original')
[Link](img_org)
[Link]('off')
v = [Link](img_org, 0)
[Link](1, 3, 2)
[Link]('Vertical Flip')
[Link](v)
[Link]('off')
h = [Link](img_org, 1)
[Link](1, 3, 3)
[Link]('Horizontal Flip')
[Link](h)
[Link]('off')
plt.tight_layout()
[Link]()

Sheth. N. K. T. T. College T. Y. Bsc. Data Science


Computer Vision

Output : -

Sheth. N. K. T. T. College T. Y. Bsc. Data Science


Computer Vision

1.C) : - Display color components of image.


Code : -
import cv2
import [Link] as plt
image = [Link]("[Link]")
# Check if the image was loaded successfully
if image is None:
print("Error: Could not load image. Check the file path.")
exit()
# Convert image from BGR (OpenCV default) to RGB (for matplotlib display)
image_rgb = [Link](image, cv2.COLOR_BGR2RGB)
# Create Blue channel image (set Green and Red to 0)
b = [Link]()
b[:, :, 1] = 0 # Zero Green
b[:, :, 2] = 0 # Zero Red
blue_rgb = [Link](b, cv2.COLOR_BGR2RGB)
# Create Green channel image (set Blue and Red to 0)
g = [Link]()
g[:, :, 0] = 0 # Zero Blue
g[:, :, 2] = 0 # Zero Red
green_rgb = [Link](g, cv2.COLOR_BGR2RGB)
# Create Red channel image (set Blue and Green to 0)
r = [Link]()
r[:, :, 0] = 0 # Zero Blue
r[:, :, 1] = 0 # Zero Green
red_rgb = [Link](r, cv2.COLOR_BGR2RGB)
# Plot the original and color channel images
[Link](figsize=(12, 4))
[Link](1, 4, 1)
Sheth. N. K. T. T. College T. Y. Bsc. Data Science
Computer Vision

[Link]('Original Image')
[Link](image_rgb)
[Link]('off')
[Link](1, 4, 2)
[Link]('Blue Channel')
[Link](blue_rgb)
[Link]('off')
[Link](1, 4, 3)
[Link]('Green Channel')
[Link](green_rgb)
[Link]('off')
[Link](1, 4, 4)
[Link]('Red Channel')
[Link](red_rgb)
[Link]('off')
plt.tight_layout()
[Link]()

Output : -

Sheth. N. K. T. T. College T. Y. Bsc. Data Science


Computer Vision

1.D) : - Display of gray scale images


Code : -
import [Link] as plt
import cv2
image = [Link]("[Link]")
image_rgb = [Link](image, cv2.COLOR_BGR2RGB)
[Link](1,2,1)
[Link](image_rgb)
[Link]("Original Image")
[Link](1,2,2)

gray_image = [Link](image_rgb, cv2.COLOR_RGB2GRAY) # Corrected here


[Link](gray_image, cmap='gray')
[Link]("Gray Scaled")
[Link]()

output : -

Sheth. N. K. T. T. College T. Y. Bsc. Data Science


Computer Vision

1.E) : - To find the negative of an image


Code : -
import cv2
import [Link] as plt
image=[Link]("[Link]")
[Link](1,2,1)
[Link]([Link](image, cv2.COLOR_BGR2RGB))
[Link]("original image")
[Link](1,2,2)
negative_image=155 - image
[Link]([Link](negative_image, cv2.COLOR_BGR2RGB))
[Link]('Negative Image')
[Link]()

Output : -

Sheth. N. K. T. T. College T. Y. Bsc. Data Science


Computer Vision

Practical 2 : - Using histogram for image quality analysis


2.A) : - Calculate the histogram of a given imae
from matplotlib import pyplot as plt
img = [Link]("[Link]",0)
histr = [Link]([img],[0],None,[256],[0,256])
[Link](histr)
[Link]()

output : -

Sheth. N. K. T. T. College T. Y. Bsc. Data Science


Computer Vision

2.B) : - Histogram Equalization


Code : -
import cv2
from matplotlib import pyplot as plt
img = [Link]('[Link]',0)

[Link](1,2,1)
histr = [Link]([img],[0],None,[256],[0,256])
[Link](histr)
[Link](1,2,2)
equ = [Link](img)
[Link](equ)
[Link]()

Output : -

Sheth. N. K. T. T. College T. Y. Bsc. Data Science


Computer Vision

Practical 3 : - Program for Image Filtering


3.A) : - Low pass filter
1) Average filter
Code : -
import cv2
import [Link] as plt
import numpy as np
img=[Link]("[Link]")
imagee=[Link](img,cv2.COLOR_RGB2BGR)
[Link](1,2,1)
[Link]("Original")
[Link](imagee)
kernel = [Link]((7,7),np.float32)/25
img2=cv2.filter2D(imagee,-1,kernel)
[Link](1,2,2)
[Link]("Average Filter")
[Link](img2)
[Link]()

Output : -

Sheth. N. K. T. T. College T. Y. Bsc. Data Science


Computer Vision

2) Weighted average filter


Code : -
import cv2
import [Link] as plt
import numpy as np
img=[Link]("[Link]")
imagee=[Link](img, cv2.COLOR_RGB2BGR)
[Link](1,2,1)
[Link]("original")
[Link](imagee)
img2= [Link](imagee,(5,7),100)
[Link](1,2,2)
[Link]("Weighted Average Filter")
[Link](img2)
[Link]()

Output : -

Sheth. N. K. T. T. College T. Y. Bsc. Data Science


Computer Vision

3) Median Filter
Code : -
import cv2
import [Link] as plt
import numpy as np
img=[Link]("[Link]")
imagee=[Link](img, cv2.COLOR_RGB2BGR)
[Link](1,2,1)
[Link]("original")
[Link](imagee)
img2= [Link](imagee,7)
[Link](1,2,2)
[Link]("median Filter")
[Link](img2)
[Link]()

Output : -

Sheth. N. K. T. T. College T. Y. Bsc. Data Science


Computer Vision

3.A) High Pass Filter


Code : -
import cv2
import [Link] as plt
import numpy as np
image = [Link]('[Link]')
gray_image = [Link](image, cv2.COLOR_BGR2GRAY)
sobelx = [Link](gray_image, cv2.CV_64F, 1, 0, ksize=5)
sobely = [Link](gray_image, cv2.CV_64F, 0, 1, ksize=5)
sobel_combined = [Link]([Link](sobelx), 0.5,
[Link](sobely), 0.5, 0)
[Link](2, 2, 1)
[Link]([Link](image, cv2.COLOR_BGR2RGB))
[Link]('Original Image')
[Link]('off')
[Link](2, 2, 2)
[Link](gray_image, cmap='gray')
[Link]('Grayscale Image')
[Link]('off')
[Link](2, 2, 3)
[Link](sobelx, cmap='gray')
[Link]('Sobel X')
[Link]('off')
[Link](2, 2, 4)
[Link](sobely, cmap='gray')
[Link]('Sobel Y')
[Link]('off')
[Link]()

Sheth. N. K. T. T. College T. Y. Bsc. Data Science


Computer Vision

Output : -

Sheth. N. K. T. T. College T. Y. Bsc. Data Science


Computer Vision

2) Laplacian operator
Code : -
import cv2
import numpy as np
import [Link] as plt
image = [Link]('[Link]')
gray_image = [Link](image, cv2.COLOR_BGR2GRAY)
laplacian = [Link](gray_image, cv2.CV_64F)
laplacian_abs = np.uint8([Link](laplacian))
[Link](1, 2, 1)
[Link]([Link](image, cv2.COLOR_BGR2RGB))
[Link]('Original Image')
[Link]('off')
[Link](1, 2, 2)
[Link](laplacian_abs, cmap='gray')
[Link]('Laplacian Edge Detection')
[Link]('off')
plt.tight_layout()
[Link]()
Output : -

Sheth. N. K. T. T. College T. Y. Bsc. Data Science


Computer Vision

Practical 3B) Design non-linear filtering


Code : -
import cv2
import [Link] as plt
image = [Link]('[Link]')
image_rgb = [Link](image, cv2.COLOR_BGR2RGB)
filtered_image = [Link](image_rgb, d=9, sigmaColor=75, sigmaSpace=75)
[Link](1,2,1)
[Link](image_rgb)
[Link]('Original Image')
[Link](1,2,2)
[Link](filtered_image)
[Link]('Bilateral Filtered Image')
[Link]()

Output : -

Sheth. N. K. T. T. College T. Y. Bsc. Data Science


Computer Vision

Practical 4) : -Edge detection with gradient and convolution of ana Image.


Code : -
import cv2
import numpy as np
import [Link] as plt
image = [Link]('[Link]')
gray_image = [Link](image, cv2.COLOR_BGR2GRAY)
sobel_x = [Link](gray_image, cv2.CV_64F, 1, 0, ksize=3)
sobel_y = [Link](gray_image, cv2.CV_64F, 0, 1, ksize=3)
magnitude = [Link](sobel_x, sobel_y)
threshold_value = 100
_, gradient_edges = [Link](magnitude, threshold_value, 255,
cv2.THRESH_BINARY)
gradient_edges = np.uint8(gradient_edges)
lower_threshold = 100
upper_threshold = 200
canny_edges = [Link](gray_image, lower_threshold, upper_threshold)
[Link](figsize=(15, 6))
[Link](1, 3, 1)
[Link]('Original Image')
[Link]([Link](image, cv2.COLOR_BGR2RGB))
[Link]('off')
[Link](1, 3, 2)
[Link]('Gradient-Based Edges')
[Link](gradient_edges, cmap='gray')
[Link]('off')

Sheth. N. K. T. T. College T. Y. Bsc. Data Science


Computer Vision

[Link](1, 3, 3)
[Link]('Canny Edges')
[Link](canny_edges, cmap='gray')
[Link]('off')
plt.tight_layout()
[Link]()
Output : -

Sheth. N. K. T. T. College T. Y. Bsc. Data Science


Computer Vision

Practical 5) : - Finding Threshold of Image


5.A) Program to find threshold of gray scale image
Code : -
import cv2
import numpy as np
import [Link] as plt
image = [Link]('[Link]')
gray_image = [Link](image, cv2.COLOR_BGR2GRAY)
threshold_value = 127
max_value = 255
ret, thresh_image = [Link](gray_image, threshold_value, max_value,
cv2.THRESH_BINARY)
[Link](1, 2, 1)
[Link]('Original Image')
[Link](gray_image, cmap='gray')
[Link](1, 2, 2)
[Link]('Thresholded Image')
[Link](thresh_image, cmap='gray')
plt.tight_layout()
[Link]()

Output : -

Sheth. N. K. T. T. College T. Y. Bsc. Data Science


Computer Vision

5.B.) Program to find threshold of RGB image.

Code : -
import cv2
import numpy as np
import [Link] as plt
image_bgr = [Link]('[Link]')
image_rgb = [Link](image_bgr, cv2.COLOR_BGR2RGB)
threshold_value = 127
max_value = 255
r, g, b = [Link](image_rgb)
_, thresh_r = [Link](r, threshold_value, max_value, cv2.THRESH_BINARY)
_, thresh_g = [Link](g, threshold_value, max_value, cv2.THRESH_BINARY)
_, thresh_b = [Link](b, threshold_value, max_value, cv2.THRESH_BINARY)
thresh_rgb = [Link]([thresh_r, thresh_g, thresh_b])
[Link](1, 2, 1)
[Link]('Original Image')
[Link](image_rgb)
[Link](1, 2, 2)
[Link]('Thresholded RGB Image')
[Link](thresh_rgb)
[Link]()
Output : -

Sheth. N. K. T. T. College T. Y. Bsc. Data Science


Computer Vision

Practical 6 : - Program to estimate and subtract the background of an


image.
Code : -
import cv2
import [Link] as plt
image = [Link]('[Link]')
gray = [Link](image, cv2.COLOR_BGR2GRAY)
_,mask = [Link](gray, 127, 255, cv2.THRESH_BINARY)
background = cv2.bitwise_and(image, image, mask=cv2.bitwise_not(mask))
foreground = cv2.bitwise_and(image, image, mask=mask)
[Link](1,3,1)
[Link]([Link](image, cv2.COLOR_BGR2RGB))
[Link]('Original')
[Link](1,3,2)
[Link]([Link](background, cv2.COLOR_BGR2RGB))
[Link]('Background')
[Link](1,3,3)
[Link]([Link](foreground, cv2.COLOR_BGR2RGB))
[Link]('Foreground')
[Link]()

Output : -

Sheth. N. K. T. T. College T. Y. Bsc. Data Science


Computer Vision

Practical 7 : - Program to convert color image to gray and hsv.


Code : -
import cv2
import numpy as np
import [Link] as plt
image_path = '[Link]'
image = [Link](image_path)
gray_image = [Link](image, cv2.COLOR_BGR2GRAY)
hsv_image = [Link](image, cv2.COLOR_BGR2HSV)
[Link](1,3,1)
[Link]('Original Color Image')
[Link]([Link](image, cv2.COLOR_BGR2RGB))
[Link](1,3,2)
[Link]('Grayscale Image')
[Link](gray_image, cmap='gray')
[Link](1,3,3)
[Link]('HSV Image')
[Link](hsv_image[:,:,0],cmap='hsv')
[Link]()

Output : -

Sheth. N. K. T. T. College T. Y. Bsc. Data Science


Computer Vision

Practical 8A:- Determination of edge detection using operators.


Code : -
import cv2
import numpy as np
import [Link] as plt
image = [Link]('[Link]')
gray_image = [Link](image, cv2.COLOR_BGR2GRAY)
kernel_size = 5
sigma = 1.0
blurred_image = [Link](gray_image, (kernel_size, kernel_size), sigma)
log_edges = [Link](blurred_image, cv2.CV_64F)
log_edges = [Link](log_edges)
sigma1 = 1.0
sigma2 = 2.0
blurred_image1 = [Link](gray_image, (0, 0), sigma1)
blurred_image2 = [Link](gray_image, (0, 0), sigma2)
dog_edges = blurred_image1 - blurred_image2
dog_edges = [Link](dog_edges)
[Link](figsize=(18, 6))
[Link](1, 3, 1)
[Link]('Original Grayscale Image')
[Link](gray_image, cmap='gray')
[Link]('off')
[Link](1, 3, 2)
[Link]('LoG Edge Detection')
[Link](log_edges, cmap='gray')
[Link]('off')
[Link](1, 3, 3)
[Link]('DoG Edge Detection')
Sheth. N. K. T. T. College T. Y. Bsc. Data Science
Computer Vision

[Link](dog_edges, cmap='gray')
[Link]('off')
plt.tight_layout()
[Link]()

Output : -

Sheth. N. K. T. T. College T. Y. Bsc. Data Science


Computer Vision

Practical 8.B) :- 2-D DFT and DCT


Code : -
import cv2
import numpy as np
import [Link] as plt
image = [Link]('[Link]', cv2.IMREAD_GRAYSCALE)
image_float = np.float32(image)
dft = [Link](image_float, flags=cv2.DFT_COMPLEX_OUTPUT)
dft_shift = [Link](dft)
magnitude_spectrum = 20 * [Link]([Link](dft_shift[:,:,0], dft_shift[:,:,1]) + 1)
[Link](magnitude_spectrum, cmap='gray')
[Link]('Magnitude Spectrum')
[Link]()

Output : -

Sheth. N. K. T. T. College T. Y. Bsc. Data Science


Computer Vision

Sheth. N. K. T. T. College T. Y. Bsc. Data Science


Computer Vision

8.B.) DCT
Code : -
import cv2
import numpy as np
import [Link] as plt
image = [Link]('[Link]', cv2.IMREAD_GRAYSCALE)
image = np.float32(image)
dct_result = [Link](image)
[Link](figsize=(12, 6))
[Link](1, 2, 1)
[Link]('Original Image')
[Link](image, cmap='gray')
[Link]('off')
[Link](1, 2, 2)
[Link]('DCT Result')
[Link]([Link]([Link](dct_result) + 1), cmap='gray')
[Link]('off')
[Link]()

Output : -

Sheth. N. K. T. T. College T. Y. Bsc. Data Science


Computer Vision

Practical 8.C) : - Filtering in Frequency domain.


Code : -
import cv2
import numpy as np
import [Link] as plt
f = [Link]('[Link]', 0)
[Link](figsize=(15, 6))
[Link](1, 5, 1)
[Link](f, cmap='gray')
[Link]('Original')
F = [Link].fft2(f)
Fshift = [Link](F)
[Link](1, 5, 2)
[Link](np.log1p([Link](Fshift)), cmap='gray')
[Link]('FFT Spectrum')
M, N = [Link]
H = [Link]((M, N), dtype=np.float32)
D0 = 10
for u in range(M):
for v in range(N):
D = [Link]((u - M/2)**2 + (v - N/2)**2)
H[u, v] = [Link](-(D**2) / (2 * (D0**2)))
Gshift = Fshift * H
G = [Link](Gshift)
g = [Link]([Link].ifft2(G))

[Link](1, 5, 3)
[Link](g, cmap='gray')
[Link]('Low-Pass Filtered')
HPF = 1 - H
Sheth. N. K. T. T. College T. Y. Bsc. Data Science
Computer Vision

Gshift = Fshift * HPF


G = [Link](Gshift)
g = [Link]([Link].ifft2(G))
[Link](1, 5, 4)
[Link](g, cmap='gray')
[Link]('High-Pass Filtered')
[Link](1, 5, 5)
[Link](np.log1p([Link](Gshift)), cmap='gray')
[Link]('HPF Spectrum')
plt.tight_layout()
[Link]()

Output : -

Sheth. N. K. T. T. College T. Y. Bsc. Data Science


Computer Vision

Practical 9.A.) : - Display of colour images


Code : -
import cv2
import [Link] as plt
img = [Link]('[Link]')
if img is None: exit()
[Link](1, 2, 1)
[Link]("BGR")
[Link](img)
[Link]('off')
img_rgb = [Link](img, cv2.COLOR_BGR2RGB)
[Link](1, 2, 2)
[Link]("RGB")
[Link](img_rgb)
[Link]('off')
[Link]()

Output : -

Sheth. N. K. T. T. College T. Y. Bsc. Data Science


Computer Vision

Practical 9.B.) : - Conversion between colour spaces.


Code : -
import cv2
import numpy as np
import [Link] as plt
image_path = '[Link]'
image_bgr = [Link](image_path)
image_rgb = [Link](image_bgr, cv2.COLOR_BGR2RGB)
image_hsv = [Link](image_bgr, cv2.COLOR_BGR2HSV)
image_YCRB = [Link](image_bgr, cv2.COLOR_BGR2YCrCb)
[Link](1,3,1)
[Link]('Original BGR')
[Link]([Link](image_bgr, cv2.COLOR_BGR2RGB))
[Link]('off')
[Link](1,3,2)
[Link]('HSV')
[Link](image_hsv)
[Link]('off')
[Link](1,3,3)
[Link]('YCRB')
[Link](image_YCRB)
[Link]('off')
[Link]()
Output : -

Sheth. N. K. T. T. College T. Y. Bsc. Data Science


Computer Vision

Practical 10.A) :- DWT of images


Code : -
import cv2
import pywt
import numpy as np
import [Link] as plt
image = [Link]('[Link]', cv2.IMREAD_GRAYSCALE)
if image is None:
print("Image not found. Check the path or file name.")
exit()
image = np.float32(image)
coeffs2 = pywt.dwt2(image, 'haar')
cA, (cH, cV, cD) = coeffs2
[Link](figsize=(6, 6))
[Link](2, 2, 1)
[Link]('Approximation')
[Link](cA, cmap='gray')
[Link]('off')
[Link](2, 2, 2)
[Link]('Horizontal')
[Link](cH, cmap='gray')
[Link]('off')
[Link](2, 2, 3)

Sheth. N. K. T. T. College T. Y. Bsc. Data Science


Computer Vision

[Link]('Vertical')
[Link](cV, cmap='gray')
[Link]('off')
[Link](2, 2, 4)
[Link]('Diagonal')
[Link](cD, cmap='gray')
[Link]('off')
plt.tight_layout()
[Link]()
Output : -

Sheth. N. K. T. T. College T. Y. Bsc. Data Science


Computer Vision

Practical 10.B.) : - Segmentation using watershed transfrom


Code : -
import cv2
import numpy as np
import [Link] as plt
from skimage import morphology
image = [Link]('[Link]')
gray = [Link](image, cv2.COLOR_BGR2GRAY)
_, binary = [Link](gray, 127, 255, cv2.THRESH_BINARY_INV)
dist_transform = [Link](binary, cv2.DIST_L2, 5)
local_maxi = morphology.local_maxima(dist_transform)
markers = np.zeros_like(dist_transform, dtype=np.int32)
markers[local_maxi] = [Link](1, [Link](local_maxi) + 1)
markers[binary == 0] = 0
markers = [Link](image, markers)
image[markers == -1] = [0, 0, 255]
[Link](figsize=(12, 4))
[Link](1, 3, 1)
[Link]([Link](image, cv2.COLOR_BGR2RGB))
[Link]('Watershed Result')
[Link]('off')
[Link](1, 3, 2)
[Link](dist_transform, cmap='jet')
[Link]('Distance Transform')
[Link]('off')
[Link](1, 3, 3)
[Link](local_maxi, cmap='gray')
[Link]('Local Maxima')
[Link]('off')
plt.tight_layout()
Sheth. N. K. T. T. College T. Y. Bsc. Data Science
Computer Vision

[Link]()

Output : -

Sheth. N. K. T. T. College T. Y. Bsc. Data Science

You might also like