Automatic composition of broadcast sports video

Abstract

This study examines an automatic broadcast soccer video composition system. The research is important as the ability to automatically compose broadcast sports video will not only improve broadcast video generation efficiency, but also provides the possibility to customize sports video broadcasting. We present a novel approach to the two major issues required in the system's implementation, specifically the camera view selection/switching module and the automatic replay generation module. In our implementation, we use multi-modal framework to perform video content analysis, event and event boundary detection from the raw unedited main/sub-camera captures. This framework explores the possible cues using mid-level representations to bridge the gap between low-level features and high-level semantics. The video content analysis results are utilized for camera view selection/switching in the generated video composition, and the event detection results and mid-level representations are used to generate replays which are automatically inserted into the broadcast soccer video. Our experimental results are promising and found to be comparable to those generated by broadcast professionals.

Visual 3D Model-based Tracking toward Autonomous Live Sports Broadcasting using a VTOL Unmanned Aerial Vehicle in GPS-Impaired Environments

“…Recently, multiple moving colour cameras and image processing techniques allow viewers to see dynamic scenes from a variety of angles [2,3]. As many cameras are used for providing better coverage of the field, especially for soccer and football, technology for choosing the optimal camera is important [4,5]. Some systems detect multiple players and track them with PTZ camera automatically [6].…”

“…These days, there is a growing interest in UAVs and drones for various purposes, such as surveillance, reconnaissance operations, monitoring, and aerial photography. Many broadcasting companies use UAVs and drones with cameras to obtain dynamic scenes from a wide view without complex equipment 5 . Amazon is planning to use UAVs for delivery in the USA and are now free to test its Prime Air service 6 .…”

Real-time selection of video streams for live TV broadcasting based on Query-by-Example using a 3D model

“…For example, the TD in charge of the live broadcasting of a car race could be interested, at the last lap of the race, in a shot that shows the cars fighting for the first position. A system to provide automatically the TD with the cameras that are obtaining these interesting shots (as in [1,2,3,4]), would be very useful for broadcasting corporations. However, a challenge for this kind of systems is how to obtain a precise definition of what an interesting shot for the TD is, to query the available cameras in order to retrieve those that can provide a similar shot.…”

“…However, a challenge for this kind of systems is how to obtain a precise definition of what an interesting shot for the TD is, to query the available cameras in order to retrieve those that can provide a similar shot. Both [1] and [3] are focused on the automatic selection of the best camera view in ball sports without interaction of the TD, and specifically they considered that an interesting shot is the one that obtains the best view of the ball and the one that provides a clear view, respectively. In [2] the use of location-dependent queries [5] was proposed as the basic building block for monitoring sport events and help the TD in the broadcasting task; however, the system had several limitations, as it was not able to manage 3D models of the queries and the scenes, and occlusions among objects were not taken into account.…”

ViComp: composition of user-generated videos

“…This includes camera selection for lecture webcast [9,41] and meetings [27,42], sports video broadcast [8,11,38], home-video summarization [7,15], and multi-camera mashup generation [30,35]. In video editing for summarization, the continuity of the event is disregarded and only key frames are included in the output video.…”