Circle Detection in Computer
Vision
Detailed Notes with Diagrams (PPT
Format)
Introduction
• • Circle detection is the process of identifying
circular shapes in digital images.
• • Circles are important geometric primitives in
computer vision tasks such as object
detection, medical imaging, and robotics.
• • Unlike lines, circles have three parameters:
center (x, y) and radius (r).
Steps in Circle Detection
• 1. Image Preprocessing – grayscale conversion,
noise removal.
• 2. Edge Detection
• 3. Circle Extraction – applying algorithms like
Hough Circle Transform, RANSAC.
• 4. Post-Processing – refine circles, remove
false detections.
Methods of Circle Detection
• • Hough Circle Transform – Classic method for
detecting circles by voting in parameter space
(x, y, r).
• • Gradient-Based Methods – Use edge
gradients to estimate circle centers and radii.
• • RANSAC for Circle Fitting – Robust model
fitting for noisy/occluded circles.
• • Deep Learning Approaches – CNN-based
detectors for complex scenes.
Hough Circle Transform
• A circle in Cartesian coordinates is defined as:
• (x - a)² + (y - b)² = r²
• where (a, b) = center, r = radius.
• Each edge pixel votes in a 3D accumulator (a, b, r).
• Peaks in accumulator correspond to detected circles.
• Pros: Can detect multiple circles, robust.
• Cons: Computationally expensive due to 3D
parameter space.
Applications of Circle Detection
• • Medical imaging – detecting cells, tumors, or
circular structures.
• • Robotics – detecting wheels, circular objects.
• • Industrial inspection – detecting pipes, holes,
and gears.
• • Traffic systems – detecting circular traffic
signs.
• • Astronomy – detecting celestial bodies like
planets and stars.