Automatic tracking of player locations from video image of football game

Masaaki Honda, Noriyuki Oosaka

Research output: Chapter in Book/Report/Conference proceedingChapter

Abstract

In this paper, we describe a method for automatically tracking the location of a player on the pitch from the video image of a football game. In this method, the player location on the video image is determined by employing binary image transformation based on the color features of the player’s uniform and a centroid computation of the binary image. The image area for the position tracking and the color threshold value in the binarization are automatically adapted in the tracking process to accommodatethe motion of the player as well as the lighting condition, which is dependent on the pitch location. The field coordinates of the player location are determined from the video coordinates by using the two dimensional DLT (Direct Linear Transformation) method. The automatic tracking method was evaluated for video data which were collected fora football game. The results indicated that the correctly detected rate of the tracking method was 99.93 %.

Original languageEnglish
Title of host publicationSports Performance
PublisherSpringer Japan
Pages353-365
Number of pages13
ISBN (Print)9784431553151, 9784431553144
DOIs
Publication statusPublished - 2015 Jan 1

Keywords

  • Automatic tracking
  • Football game
  • Image processing

ASJC Scopus subject areas

  • Medicine(all)
  • Engineering(all)
  • Social Sciences(all)

Fingerprint Dive into the research topics of 'Automatic tracking of player locations from video image of football game'. Together they form a unique fingerprint.

  • Cite this