Image Rotation Based on Largest Contour Detection

This function loads an image, converts it to a grayscale image, and uses Canny edge detection to find contours. Then, it finds the largest contour and calculates its bounding box. Finally, the function rotates the image around the center of the bounding box.

Technology Stack : scikit-image, numpy

Code Type : Image processing

Code Difficulty : Intermediate

                
                    
import numpy as np
from skimage import feature, io, transform

def find_contours_and_rotate_image(input_image_path, angle):
    # Load the image
    image = io.imread(input_image_path)
    
    # Convert the image to grayscale
    gray_image = feature.canny(image)
    
    # Find contours in the image
    contours, _ = feature.find_contours(gray_image, 0.5)
    
    # Calculate the bounding box for the largest contour
    if contours:
        largest_contour = max(contours, key=lambda c: np.sum(c[:, 1]))
        bounding_box = transform.bbox(largest_contour)
    else:
        bounding_box = (0, 0, image.shape[1], image.shape[0])
    
    # Rotate the image around the center of the bounding box
    rotated_image = transform.rotate(image, angle, resize=False, mode='edge', center=bounding_box[:2])
    
    return rotated_image