Virtual lines for offside situations analysis in football

Abstract

Offside is one of the situations that is analyzed by the Video Assistant Referee (VAR). However, it has caused some controversy due to the delay in the analysis and definition of the irregularity. This work proposes a method that helps in the analysis of offside situations and also makes it available for nonprofessional matches. Here, image processing algorithms were used to determine offside situations in football matches from TV videos, of course, in accordance with the game regulation. The method includes the image vanishing point identification, camera calibration and the virtual offside line drawing. The method presented good results from 10 videos selected for analysis, with five from the right side of the field and five on from the left side. Among the videos, one was chosen as the basis for explaining the development of the method and demonstrate a situation with a virtual line drawn automatically, therefore determining an offside situation. As a result, the virtual line is identified by the color red when the manually selected player is in offside and green when he is not.