A Lagrange Programming Neural Network Approach for Robust Ellipse Fitting

Research output: Journal Publications and Reviews (RGC: 21, 22, 62)21_Publication in refereed journal

View graph of relations

Related Research Unit(s)

Detail(s)

Original languageEnglish
Pages (from-to)686-696
Journal / PublicationLecture Notes in Computer Science
Volume10636
Publication statusPublished - Nov 2017

Conference

Title24th International Conference on Neural Information Processing (ICONIP 2017)
PlaceChina
CityGuangzhou
Period14 - 18 November 2017

Abstract

Ellipse fitting aims at constructing an elliptical equation that best fits the scattering points collected from an edge detection process. However, the edge detection process may introduce some noisy scattering points. This paper proposes a robust ellipse fitting model based on the Lagrange programming neural network (LPNN) framework. We formulate the ellipse fitting problem as a constrained optimization problem. The objective function contains an l1-norm term which can effectively suppress the effect of outliers. Since the LPNN framework cannot handle non-differentiable objective functions, we introduce an approximation for the l1-norm term. Besides, the local stability of the proposed LPNN method is discussed. Simulation results show that the proposed ellipse fitting algorithm can effectively reduce the influence of outliers. 1-norm term which can effectively suppress the effect of outliers. Since the LPNN framework cannot handle non-differentiable objective functions, we introduce an approximation for the -norm term. Besides, the local stability of the proposed LPNN method is discussed. Simulation results show that the proposed ellipse fitting algorithm can effectively reduce the influence of outliers.

Research Area(s)

  • Ellipse fitting, LPNN, Outliers