Image Transformations using OpenCV in Python

Image transformations are fundamental operations in image processing that modify an image's position, orientation, size, or shape. OpenCV (Open Source Computer Vision Library) is a popular open-source library that provides efficient functions for performing these transformations in Python.

• Translate images by shifting them horizontally or vertically.
• Modify image geometry through reflection, rotation, scaling, and shearing operations.
• Extract important regions from images using cropping techniques.

Image Translation

In computer vision and image processing, image translation refers to shifting an image from one position to another. It changes the location of objects in the image without altering their shape, size, or orientation.

• The image is loaded using the imread() function after importing the NumPy and OpenCV modules.
• The warpAffine()method performs the translation using the transformation matrix M.
• In the matrix, x = 100 shifts the image 100 pixels to the right and y = 50 shifts it 50 pixels downward, while (cols, rows) preserves the original image dimensions.

Output:

Image Reflection

Image reflection is used to flip the image vertically or horizontally. For reflection along the x-axis, we set the value of Sy to -1, Sx to 1, and vice-versa for the y-axis reflection.

• For horizontal reflection, the value of Sy is set to -1 while Sx remains 1, causing the image to flip along the x-axis.
• For vertical reflection, the value of Sx is set to -1 while Sy remains 1, causing the image to flip along the y-axis.

Output:

Image Rotation

Image rotation is the process of rotating an image by a specified angle around a fixed point, usually its center. It is widely used in image processing, computer vision, and data augmentation tasks.

• The getRotationMatrix2D() function creates the transformation matrix required for rotation.
• The image is rotated by 30 degrees around its center point (cols/2, rows/2).
• The scale factor 0.6 reduces the image size to 60% of its original dimensions while applying the rotation.

Output:

Image Scaling 

Image scaling is a process used to resize a digital image. We perform two things in the image scaling either we enlarge the image or we shrink the image, OpenCV has a built-in function cv2.resize() for image scaling.

• The image is shrunk to 350 × 300 pixels using the resize() function with INTER_AREA interpolation.
• The shrunk image is enlarged by 1.5× along both axes using INTER_CUBIC interpolation for smoother scaling.

Output:

Image Cropping

Image cropping is the process of removing unwanted portions of an image to focus on a specific region.

• OpenCV stores images as NumPy arrays, allowing cropping through array indexing.
• The expression img[100:300, 100:300] extracts a 200 × 200 pixel region from the image.

Output:

Image Shearing in X-Axis

Image shearing in the x-axis shifts image pixels horizontally, causing the image to appear slanted while preserving parallelism.

• The transformation matrix [[1, 0.5, 0], [0, 1, 0], [0, 0, 1]] applies a shearing factor of 0.5 along the x-axis.
• The warpPerspective() function performs the transformation, and the output size is increased to accommodate the sheared image.

Output:

Image Shearing in Y-Axis

Image shearing in the y-axis shifts image pixels vertically, causing the image to appear slanted while preserving parallelism.

• The transformation matrix [[1, 0, 0], [0.5, 1, 0], [0, 0, 1]] applies a shearing factor of 0.5 along the y-axis.
• The warpPerspective() function performs the transformation, and the output size is increased to accommodate the sheared image.

Output: