Directional sub-pixel edge detection based on limited window local entropy

  • Wang Qiang*
  • , Zhang Zhijing
  • , Ye Xin
  • , Jin Xin
  • , Zhou Zhi
  • *Corresponding author for this work

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

Abstract

To realize precise micro-vision positioning based on edge detection of micro-structure parts, a pixel-level edge points detection method based on limit local entropy was proposed. 2-3 pixel were extended in the gradient direction to both sides from the edge points to construct the 3 cubic spline interpolation function of edge transition zone. Using function derivation to obtain the floating value of sub-pixel coordinates, then, sub-pixel coordinate was located according to the positioning principle of directional sub-pixel coordinates. Finally, the fine grinding standard gauge block and LIGA micro-spring were taken as example, using traditional gradient edge detection and the method proposed in this paper to realize image edge detection, and achieve the width measurement. The results show that the edge positioning accuracy of the method proposed in this paper is higher than the traditional gradient edge detection method. And whether the pixel level edge positioning accuracy or sub-pixel edge positioning accuracy are much higher than the positioning accuracy required of micro-assembly systems (5μm).

Original languageEnglish
Title of host publicationProceedings - 4th International Conference on Intelligent Computation Technology and Automation, ICICTA 2011
Pages523-526
Number of pages4
DOIs
Publication statusPublished - 2011
Event2011 4th International Conference on Intelligent Computation Technology and Automation, ICICTA 2011 - Shenzhen, Guangdong, China
Duration: 28 Mar 201129 Mar 2011

Publication series

NameProceedings - 4th International Conference on Intelligent Computation Technology and Automation, ICICTA 2011
Volume2

Conference

Conference2011 4th International Conference on Intelligent Computation Technology and Automation, ICICTA 2011
Country/TerritoryChina
CityShenzhen, Guangdong
Period28/03/1129/03/11

Keywords

  • Limited window
  • Local entropy
  • Optical properties
  • Sub-pixel edge detection

Fingerprint

Dive into the research topics of 'Directional sub-pixel edge detection based on limited window local entropy'. Together they form a unique fingerprint.

Cite this