People tracking in dense crowds using optical flow clustering based superpixel

Shota Takayama, Hirokatsu Kataoka, Sho Isobe, Naoki Kurita, Makoto Masuda, Yoshimitsu Aoki

Research output: Contribution to journalArticlepeer-review


People tracking in crowded scenes should be focused since the problem is beneficial and challenging in computer vision field. However, the problem of "crowd tracking" is extremely difficult because hard occlusions, various motions and posture changes. Especially in the crowd tracking, we need to handle occlusion for more robust tracking. This paper tackles a robust crowd tracking based on combination of superpixel and optical flow tracking. The SLIC based superpixel algorithm adaptively estimates a boundary between person and background, therefore the combination of superpixel and optical flow tracking becomes a highly confident tracking for crowd tracking. The tracking experiments show significant results on the UCF crowd dataset in terms of performance rate with comparison.

Original languageEnglish
Pages (from-to)259-265
Number of pages7
JournalSeimitsu Kogaku Kaishi/Journal of the Japan Society for Precision Engineering
Issue number3
Publication statusPublished - 2016


  • Dense crowd
  • Kmeans clustering
  • Optical flow
  • People tracking
  • Superpixel

ASJC Scopus subject areas

  • Mechanical Engineering


Dive into the research topics of 'People tracking in dense crowds using optical flow clustering based superpixel'. Together they form a unique fingerprint.

Cite this