Image feature detection as robust model fitting

Authors:
Dengfeng Chai;Qunsheng Peng
Affiliations:
State Key Lab of CAD&CG, Zhejiang University, Hangzhoz, China;State Key Lab of CAD&CG, Zhejiang University, Hangzhoz, China
Venue:
ACCV'06 Proceedings of the 7th Asian conference on Computer Vision - Volume Part II
Year:
2006

Citing 11
Cited 0

Robust regression and outlier detection

Robust regression and outlier detection
A survey of the Hough transform

Computer Vision, Graphics, and Image Processing
A new curve detection method: randomized Hough transform (RHT)

Pattern Recognition Letters
A probabilistic Hough transform

Pattern Recognition
Which Hough transform?

CVGIP: Image Understanding
Procedural elements for computer graphics (2nd ed.)

Procedural elements for computer graphics (2nd ed.)
Random sample consensus: a paradigm for model fitting with applications to image analysis and automated cartography

Communications of the ACM
An efficient randomized algorithm for detecting circles

Computer Vision and Image Understanding
A new randomized algorithm for detecting lines

Real-Time Imaging
A Metric for Line Segments

IEEE Transactions on Pattern Analysis and Machine Intelligence
Algorithm for computer control of a digital plotter

IBM Systems Journal

Quantified Score

Hi-index	0.00

Visualization

Abstract

In this paper, we describe image feature as parameterized model and formulate feature detection as robust model fitting problem. It can detect global feature easily without parameter transformation, which is needed by Hough Transform methods. We adopt RANSAC paradigm to solve the problem. It is immune to outliers and can deal with image contains multiple features and noisy pixels. In the voting stage of RANSAC, in contrast with previous methods which need distance computation and comparison, we apply Bresenham algorithm to generate pixels in the inlier region of the feature and use the foreground pixels in this region to vote the potential feature. It greatly improves the efficiency and can detect spatially-linked features easily. Experimental results with both synthetic and real images are reported.