How to process images using OpenCV

Key takeaways:

  • Image processing lets us analyze and change digital images, from smartphone filters to medical imaging.

  • OpenCV is a Python library that makes image processing tasks like resizing and detecting objects easy.

  • An image matrix represents each pixel’s color intensity, forming the basis for image manipulation.

  • Image filters apply matrix convolutions to modify images, such as blurring or sharpening.

  • Image transformations allow you to resize, crop, blur, or sharpen images without using filters.

Image processing

Image processing is an exciting field that allows us to analyze and change digital images. It is used everywhere, from smartphones with cool filters to advanced medical imaging. Python offers a powerful library called OpenCV, which makes image processing easier for everyone, including beginners. With OpenCV, you can perform various tasks like resizing pictures, detecting faces, and even recognizing objects.

This Answer will introduce you to the basics of image processing using OpenCV and show you how to get started on your own projects.

Image matrix

An image is typically represented as a two-dimensional matrix, where each element corresponds to a pixel value within the image. These pixel values store information regarding the color intensity of each pixel. In grayscale images, each pixel value ranges from.

For example, a grayscaled image of size 4×34 \times 3 pixels will be:

[ 

  [135, 150, 120, 100],  

  [50, 80, 200, 90],  

  [200, 180, 220, 75] 

]
Image matrix of a grayscaled image

Using Python OpenCV for image processing

The Python OpenCV library opens up a world of possibilities for image processing. Using the following image as an example, let’s explore key concepts, such as filtering, transformation, etc., for manipulating and enhancing images.

The sample image used in this answer
The sample image used in this answer

Image filters

An image filter is a matrix convolved over the image matrix to manipulate it in different ways, such as blurring, sharpening, and noise reduction. Filters are also very useful in edge detection. For example, in the following code, we’re using the famous Gaussian Blur to blur the image and the Canny edge detection filter to detect the edges of the shapes.

import cv2
import numpy as np
# Load an image
image = cv2.imread('image.png')
# Preprocess the image
gray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)
blur = cv2.GaussianBlur(gray, (5, 5), 0)
# Edge detection
edges = cv2.Canny(blur, 30, 150)  # Adjust threshold values as needed
# Saving the processed images
cv2.imwrite('output/BlurredImage.png', blur)
cv2.imwrite('output/EdgesImage.png', edges)

Image transformation

Image transformation is a crucial technique in image processing. It involves applying a specific transformation matrix to the image matrix. However, unlike filters, the transformation matrix is not convolved across the image matrix. Instead, it directly gets multiplied with it.

We can also apply miscellaneous transformations such as the ones discussed and demonstrated below.

Crop the image

This code crops a specific region from the image.png (from row number 50 to 200 and column number 100 to 300), and then saves the cropped portion as CroppedImage.png in the output folder.

import cv2
import numpy as np
# Load an image
image = cv2.imread('image.png')
# Crop the image
cropped_image = image[50:200, 100:300]
# Saving the processed images
cv2.imwrite('output/CroppedImage.png', cropped_image)

Resize the image

This code loads an image named image.png, resizes it to a width of 300300 pixels and a height of 200200 pixels, and saves the resized image as ResizedImage.png in the output folder:

import cv2
import numpy as np
# Load an image
image = cv2.imread('image.png')
# Resize the image
resized_image = cv2.resize(image, (300, 200))
cv2.imwrite('output/ResizedImage.png', resized_image)

Blur the image

Here the code loads an image named image.png, blurs the image by applying a Gaussian blur with a 5x55x5 kernel to reduce noise and smooth the image, and saves the blurred result as BlurredImage.png in the output folder.

import cv2
import numpy as np
# Load an Image
image = cv2.imread('image.png')
# Apply Gaussian Blur
blurred_image = cv2.GaussianBlur(image, (5, 5), 0)
cv2.imwrite('output/BlurredImage.png', blurred_image)

Sharpen the image

This code loads the image that we’ve saved as image.png, applies a sharpening filter to enhance edges and details, and saves the sharpened image as SharpenedImage.png in the output folder.

import cv2
import numpy as np
# Load an image
image = cv2.imread('image.png')
# Apply sharpening filter
sharpening_filter = np.array([[0, -1, 0],
                              [-1, 5, -1],
                              [0, -1, 0]])
sharpened_image = cv2.filter2D(image, -1, sharpening_filter)
cv2.imwrite('output/SharpenedImage.png', sharpened_image)

Saving the processed image

Once we have processed the image, we would also like to save our image locally on our devices. For that, OpenCV provides an imwrite() method that lets us save the processed image in any desired format.

Three-dimensional transformation

Since images are a two-dimensional representation of the three-dimensional real world, 3D transformation sometimes comes in handy in image processing. In 3D transformation, we must take care of the aspects such as perspective geometry, vanishing point, etc.

Applications of image processing

Using the operations mentioned above, OpenCV lets us manipulate images for some beneficial use cases, such as face detection, object detection, or image stitching.

Image processing with OpenCV has many real-world applications that make our lives easier and more interesting. For example, you can build a forgery detector to check if images have been manipulated, identify and differentiate faces in photos, track moving objects, or even in healthcare, where doctors can use image processing for medical imaging, helping them see inside the human body more clearly. The possibilities are vast, making image processing an essential tool in today’s world.

Conclusion

OpenCV is a versatile library that simplifies image processing tasks for everyone, whether you are a beginner or an experienced programmer. It offers a wide range of tools and functions that let you manipulate images easily. By learning how to use OpenCV, you can create exciting projects in fields like computer vision, robotics, and medical imaging. We hope this Answer has inspired you to explore image processing further and try out your own ideas with OpenCV.

Want to enhance your images or automate visual tasks? OpenCV offers a versatile toolkit for processing images. Whether you’re resizing, filtering, or transforming them. Get hands-on with practical examples and explore exciting projects such as:

Frequently asked questions

Haven’t found what you were looking for? Contact Us

How to do image processing in Python?

You can use libraries like OpenCV or Pillow in Python to load, edit, and analyze images by applying techniques like filtering, transformations, and edge detection.

Is OpenCV best for image processing?

OpenCV is one of the best and most widely used libraries for image processing, but depending on your project needs, other libraries like Pillow or scikit-image may also be useful.

What is the difference between a Gaussian filter and box filter?

A Gaussian filter smooths an image by giving more weight to pixels near the center, creating a softer blur, while a box filter gives equal weight to all pixels in the area, resulting in a more uniform, less natural blur.

Free Resources

Copyright ©2025 Educative, Inc. All rights reserved