How to get the boundary vertices of image chunking using Python and OpenCV?

Efficiently extract image chunking boundary vertices using Python and OpenCV

In image processing, it is often necessary to segment the image into blocks and extract the boundary vertices of each block. Suppose there is a single channel image, and after processing, it forms a mesh of h×m, and the value of each mesh block is incremented from 1. This article will explain how to use Python and OpenCV libraries to efficiently find the boundary vertices of each block.

Methods and code examples

We will use OpenCV to read the image and perform chunking processing (assuming the chunked image already exists, stored in a two-dimensional NumPy array, each element represents the label of the chunk). Then, use the power of NumPy to find the boundary vertices of each block.

The following code demonstrates how to traverse each block, find its boundary pixel point and record its coordinates:

 import cv2
import numpy as np

# Sample chunked image (replace with your actual chunked image data)
segmented_image = np.array([
    [1, 1, 1, 2, 2],
    [1, 1, 1, 2, 2],
    [3, 3, 3, 2, 2],
    [3, 3, 3, 4, 4],
    [3, 3, 3, 4, 4]
])

# Dictionary block_boundaries = {}

# traverse each block label for label in np.unique(segmented_image):
    # Create a mask, keep only the current block mask = (segmented_image == label).astype(np.uint8)

    # Use OpenCV to find contours, _ = cv2.findContours(mask, cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_SIMPLE)

    # Extract boundary vertices if contours:
        contour = contours[0]
        vertices = contour.reshape(-1, 2)
        block_boundaries[label] = vertices

# Print the boundary vertices of each block for label, vertices in block_boundaries.items():
    print(f"Border vertex coordinates of block {label}:")
    print(vertices)

This code uses OpenCV's findContours function to efficiently find the outer contour of each block, and then extract the vertex coordinates on the contour. The block_boundaries dictionary stores the mapping relationship between each block label and its corresponding boundary vertex coordinates.

Improvements and extensions

This method is suitable for simple rectangular block segmentation. For more complex shapes, more advanced image processing techniques may be required, such as segmentation methods based on region growth or more complex contour analysis algorithms. In addition, the code can be modified according to actual needs, such as adding error handling or optimizing performance. If there are more specific needs or questions, further requests are welcome.

The above is the detailed content of How to get the boundary vertices of image chunking using Python and OpenCV?. For more information, please follow other related articles on the PHP Chinese website!