Skip to main navigation Skip to search Skip to main content

Intelligent detection of convex polygon based on Hough transformation and set classifier

  • Xudong Yang*
  • , Peng Dai
  • , Ping He
  • , Pan Li
  • *Corresponding author for this work
  • Harbin Institute of Technology

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

This paper proposes a shape recognition method for detecting convex polygons (CVPs) in image planes, which is characterized by using Hough transformation (HT) and designing a set classifier. HT is used to extract the sides of a CVP by searching the peaks of accumulators in the Hough Parametric Space (HPS). A total set of the intersection points formed by all sides of CVP and their extending lines can be established based on the peaks in HPS, which includes a subset only containing the vertex of CVP. Based on the gradient distribution of the elements in the subset, the classifier for extracting the subset is designed. Compared with conventional method, the detection method has advantage of detecting the CVP shape with discontinuity and broken edges, and thus worthy of being promoted.

Original languageEnglish
Title of host publicationProceedings - 2010 2nd International Workshop on Intelligent Systems and Applications, ISA 2010
DOIs
StatePublished - 2010
Event2nd International Workshop on Intelligent Systems and Applications, ISA2010 - Wuhan, China
Duration: 22 May 201023 May 2010

Publication series

NameProceedings - 2010 2nd International Workshop on Intelligent Systems and Applications, ISA 2010

Conference

Conference2nd International Workshop on Intelligent Systems and Applications, ISA2010
Country/TerritoryChina
CityWuhan
Period22/05/1023/05/10

Keywords

  • Convex polygon
  • Gradient distribution
  • Hough transformation
  • Set classifier

Fingerprint

Dive into the research topics of 'Intelligent detection of convex polygon based on Hough transformation and set classifier'. Together they form a unique fingerprint.

Cite this