Automatic recognition of two-dimensional vortices by digital image processing

Shyh Ming Sheu, Jung Hua Chou

Research output: Contribution to journalArticlepeer-review


In this study, a digital image processing method has been developed to detect automatically the vortex structure behind a circular cylinder. The vortex recognition algorithm is based on the special features of a typical vortex; namely, the circular-like, elliptic-like, and'or the spiral-like characteristics extracted from the geometrical features of the vortex. The algorithm includes two parts: namely, pre-processing and vortex structure recognition. The results indicate that for a 256 × 240 image, all the vortices in the image can be detected in about 100 iterations if the noise level is low or moderate. But if the noise level is high and the pre-processing is poorly done, the algorithm is not reliable. The same situation occurs also if there are break points due to thinning or skeletonizing. These are the limitations of the present algorithm.

Original languageEnglish
Pages (from-to)269-279
Number of pages11
JournalJournal of Flow Visualization and Image Processing
Issue number4
Publication statusPublished - 2004

All Science Journal Classification (ASJC) codes

  • Condensed Matter Physics
  • Mechanical Engineering
  • Computer Science Applications

Fingerprint Dive into the research topics of 'Automatic recognition of two-dimensional vortices by digital image processing'. Together they form a unique fingerprint.

Cite this