AlanLuSun/High-quality-ellipse-detection

A high-quality ellipse detector based on arc-support line segments which can both accurately and efficiently detect ellipses in images.

What it solves

This project provides an efficient and high-quality ellipse detection algorithm designed to solve the long-standing challenge of accurately locating ellipses in real-world images, providing high localization accuracy, robustness, and stability.

How it works

The system uses an approach based on "Arc-support Line Segments" to identify ellipses. It allows users to control detection quality through two primary extrinsic parameters: elliptic angular coverage ($T_{ac}$) and the ratio of support inliers ($T_{r}$). It also includes a specified_polarity option to filter ellipses based on their polarity.

Who it’s for

This tool is intended for researchers and engineers working in computer vision and robot vision, specifically those using MATLAB and OpenCV on Windows.

Highlights

  • High Performance: Combines high localization accuracy with computational efficiency.
  • Polarity Control: Ability to detect ellipses with specific positive or negative polarity.
  • Practical Applications: Proven useful for PCB inspection, car wheel hub recognition, object fingerprinting, and robot vision.
  • Simplified Configuration: Only requires two main parameters for tuning.

Related

  • Project
    JEOresearch/EyeTracker

    A lightweight Python-based 3D eye tracking algorithm that detects the pupil ellipse from video or image data.

  • Project
    PRBonn/MapClosures

    A robotics library for detecting loop closures in mapping by using point cloud density maps to improve map accuracy and reduce drift.

  • Project
    ultralytics/ultralytics

    A comprehensive computer vision framework featuring YOLO models for real-time object detection, segmentation, classification, and pose estimation.

  • Project
    iago-suarez/ELSED

    ELSED is a high-speed line segment detector designed for resource-limited devices like drones and smartphones.

  • Project
    PaddlePaddle/PaddleClas

    A comprehensive image recognition and classification toolkit providing ultra-lightweight models and industrial-grade backbones for high-performance vision applications.