Fast and Robust
Circular Object Detection
with Probabilistic Pairwise Voting


Lili Pan, Wen-Sheng Chu, Jason M. Saragih, Fernando De la Torre and Mei Xie



Move your cursor on the image to take a closer view!
(a) Detecting circular objects in natural images

(b) Localizing iris in face images


  Introduction

  • Accurate and efficient detection of circular objects in images is a challenging computer vision problem. Existing circular object detection methods can be broadly classified into two categories:
    1. Hough Transform based: robust to noise, however the computational complexity and memory requirement are high.
    2. Maximum likelihood (ML) estimation based: more computationally efficient but sensitive to noise, and cannot detect multiple circles (e.g., robust least squares fitting).

  • This letter proposes Probabilistic Pairwise Voting (PPV), a fast and robust algorithm for circular object detection based on an extension of Hough Transform. The main contributions are three fold:
    1. We formulate the problem of circular object detection as finding the intersection of lines in the three dimensional parameter space (i.e. center and radius of the circle).
    2. We propose a probabilistic pairwise voting scheme to robustly discover circular objects under occlusion, image noise and moderate shape deformations.
    3. We use a mode-finding algorithm to efficiently find multiple circular objects.

  • We demonstrate the benefits of our approach on two real-world problems: (a) detecting circular objects in natural images, and (b) localizing iris in face images.


  Download


  Reference


  Copyright

  • These images are collected from Google image search without permission from the original copyright holders. By downloading these files, you agree not to hold the authors liable for any damage, lawsuits, or other loss resulting from the possession or use of files. If you are the copyright owner of one of these images and would like it removed from the dataset, please contact Lili Pan (panlili8255[at]hotmail.com) or Wen-Sheng Chu (wschu[at]cmu.edu).