![]() ![]() Once set, the camera’s speed is constant, Lighting conditions may change (new light sources, for example),Ĭameras are black and white and stationary,Ĭameras produce a stream of frames with speeds ranging above 150 frames per second, The algorithm should work online and cannot use any future frames, The main purpose of the algorithm is to distinguish flickering from movement so the differential frame between two adjacent frames should show only the movement, Mains AC frequency and frame rate are known, If the flickering was not removed, then the motion detection part of the system would not be able to distinguish flickering from moving objects, and the shuttlecock movement detection would not be possible. The algorithm runs before shuttlecock detection and tracking algorithms and allows the next ones to work properly. The algorithm presented in this paper is part of the challenge system. Images are captured by 14–20 cameras working at 150–200 frames per second. The system consists of servers running object movement detection and fast moving shuttlecock tracking algorithms. If a player does not agree with the referee’s decision, she or he can challenge it by asking for video verification. We have built a challenge system for badminton competitions that take place in sports halls where lights are the source of flickering effects in frame streams recorded by high-speed cameras. The proposed flicker removing algorithm reduces flickering only to the extent that it is sufficient to detect movement and analyze trajectories. The number of features of such shuttlecocks is too small to build a robust tracker, so it is necessary to extract additional features from detected movement and trajectories. At higher speeds, pictures of the shuttlecocks are blurred even when using 200fps cameras. The velocity of a shuttlecock varies between 30 km/h to 450 km/h. A moving shuttlecock is a small object that, depending on the position of the camera in relation to the court, may be represented by as small a number of pixels as around a dozen. On the contrary, we needed a fast solution just accurate enough to properly detect a moving small object as a shuttlecock. In this paper, we focus on the flickering effect in high-speed acquisition cameras caused by artificial lighting.Įxisting flicker removal solutions concentrate mostly on accuracy but not on speed. There are different sources of flickering in movies. Most sports venues where tournaments are organized are equipped with lighting causing flickering effect, thus for most tournaments, a workaround solution that removes flickering must be used. The results of the experiments carried out show the high effectiveness of the method implemented on CPU and GPU, allowing effective tracking of objects of interest in preliminary applications of a commercially offered instant review system for badminton.įor many years, big sports organizations like the Olympic committee and UEFA have required from tournament organizers to use only non-flickering lighting, but such lighting systems are expensive and are not common. Our algorithm is adaptive and works when lighting conditions are changing (new light sources) and performs well with various light sources that are causing flickering. In contrast, our original solution is designed with speed in mind with sufficient accuracy to be used before calculating differential frames to detect movement in streams. Most of the solutions found in literature concentrate on effectiveness and accuracy and not on the speed of operation. In this paper, we propose a fast flicker removal algorithm working as an online filter on frame streams at speeds exceeding 200 frames per second. ![]() A fast and yet effective enough algorithm is necessary to remove flickering so movement detection and object tracking algorithms could be used. Sports arenas in which games are played are often equipped with lights generating flickering effects in captured movies. Collected information is used for statistics, as coaches’ and players’ aids to improve technique and tactics or as referees’ aids to verify their decisions or to enrich television broadcasts. High-speed cameras are used in computer vision systems to track balls, shuttlecocks, or players in many different sports. ![]()
0 Comments
Leave a Reply. |