Swiss scientists have created a computer system that tracks the progress of sports matches.
In general, the system created by the Swiss experts from the École Polytechnique Fédérale de Lausanne can be used in many fields, but so far it has been tweaked in sports commentators. This system is able to track many people in the image at the same time. For example, the system easily “leads” at the same time all the players of the basketball mast, without having to cling to something on the player himself.
Tracking is conducted through the analysis of several video streams, plus the analysis itself is carried out through the use of three different algorithms. So, the first breaks a solid image, for example, a basketball court, into equal elements, it turns out the grid. The second algorithm tracks the trajectory of each player based on the results of the first algorithm. The third one prevents confusion that may arise due to the intersection of the trajectories of different players. The third algorithm tracks each player, and “leads” him from the first minute to the last.
The system allows to evaluate the effectiveness of the game of each team and each player, and more thoroughly than an ordinary person can do. Moreover, the technology independently evaluates the tactics of each team. But of course, even if you teach the system to comment on the actions of players or teams, all this will not be as interesting as in the case of human commentators. Well, who will insert jokes and anecdotes in between the throws of the ball in the basket? Who will be emotionally comment on the mistakes and successes? Perhaps this will teach and computer, but very soon.
')
Via gizmodo
Tracking is conducted through the analysis of several video streams, plus the analysis itself is carried out through the use of three different algorithms. So, the first breaks a solid image, for example, a basketball court, into equal elements, it turns out the grid. The second algorithm tracks the trajectory of each player based on the results of the first algorithm. The third one prevents confusion that may arise due to the intersection of the trajectories of different players. The third algorithm tracks each player, and “leads” him from the first minute to the last.
The system allows to evaluate the effectiveness of the game of each team and each player, and more thoroughly than an ordinary person can do. Moreover, the technology independently evaluates the tactics of each team. But of course, even if you teach the system to comment on the actions of players or teams, all this will not be as interesting as in the case of human commentators. Well, who will insert jokes and anecdotes in between the throws of the ball in the basket? Who will be emotionally comment on the mistakes and successes? Perhaps this will teach and computer, but very soon.
')
Via gizmodo
Source: https://habr.com/ru/post/132456/
All Articles