How to detect polygons in image using OpenCV Python?

Polygon detection is a common computer vision task that involves finding multi-sided shapes in images. OpenCV provides powerful tools to detect polygons by first finding contours and then approximating them to identify shapes with 5 or more vertices.

How It Works

The polygon detection process follows these key steps:

1. Contour Detection - Find all object boundaries in the image
2. Contour Approximation - Simplify contours to approximate polygonal shapes
3. Vertex Counting - Count vertices to identify polygons (?5 vertices)
4. Visualization - Draw detected polygons on the original image

Step-by-Step Process

• Import the required library OpenCV

• Read the input image using cv2.imread() and convert it to grayscale

• Apply thresholding cv2.threshold() on the grayscale image to create a binary image

• Find contours using cv2.findContours() function

• Loop over all detected contours

• Compute approximate contour points using cv2.approxPolyDP() function

• If the approximate contour has 5 or more vertices, classify it as a polygon

• Draw the detected polygons and display the result

Example

Here's a complete example that detects polygons in an image and highlights them ?

import cv2
import numpy as np

# Create a sample image with polygons for demonstration
img = np.ones((400, 500, 3), dtype=np.uint8) * 255

# Draw some sample polygons
pentagon = np.array([[100, 100], [150, 80], [180, 120], [140, 160], [80, 140]], np.int32)
hexagon = np.array([[250, 100], [300, 90], [330, 120], [310, 160], [260, 170], [230, 140]], np.int32)
octagon = np.array([[350, 100], [380, 110], [400, 140], [390, 170], [360, 180], [330, 170], [320, 140], [330, 110]], np.int32)

cv2.fillPoly(img, [pentagon], (200, 200, 200))
cv2.fillPoly(img, [hexagon], (150, 150, 150))
cv2.fillPoly(img, [octagon], (100, 100, 100))

# Convert to grayscale
gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)

# Apply thresholding
ret, thresh = cv2.threshold(gray, 127, 255, cv2.THRESH_BINARY_INV)

# Find contours
contours, hierarchy = cv2.findContours(thresh, cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_SIMPLE)

print(f"Number of contours detected: {len(contours)}")

# Process each contour
for cnt in contours:
    # Approximate the contour
    epsilon = 0.02 * cv2.arcLength(cnt, True)
    approx = cv2.approxPolyDP(cnt, epsilon, True)
    
    # Get coordinates for text placement
    x, y = cnt[0, 0]
    
    # Check if it's a polygon (5 or more vertices)
    if len(approx) >= 5:
        cv2.drawContours(img, [approx], -1, (0, 255, 255), 3)
        cv2.putText(img, f'Polygon ({len(approx)} vertices)', (x-20, y-10),
                   cv2.FONT_HERSHEY_SIMPLEX, 0.5, (255, 0, 0), 2)

print("Polygon detection completed!")

Number of contours detected: 3
Polygon detection completed!

Key Parameters

Common Use Cases

• Shape Recognition - Identifying geometric shapes in images

• Object Detection - Finding rectangular or polygonal objects

• Document Processing - Detecting tables, forms, and structured elements

• Quality Control - Inspecting manufactured parts with specific shapes

Conclusion

Polygon detection in OpenCV involves finding contours, approximating them with cv2.approxPolyDP(), and counting vertices. Shapes with 5 or more vertices are classified as polygons. Adjust the epsilon parameter and threshold values for optimal results based on your specific image characteristics.