CYOS
Canlı Yayın Otomasyon Sistemi
İzleyicilerin standart bir yayın programı ve içeriğin dışında kendi tercihleri ile istediği yayınları izleyebilmesi Netflix, Youtube gibi platformların milyonlarca kullanıcıya ulaşmasını sağlamıştır. Birçok yayın kuruluşu zaman içerisinde geleneksel yayıncılık faaliyetlerini durdurmayı yol haritalarına eklemişlerdir. Tek bir tablet bilgisayar üzerinden yayın yaparak binlerce izleyiciye ulaşan programlar bulunmaktadır. Bu aşamada özellikle spor müsabaka yayınlarının özel bir durumu vardır. Örnek olarak altı kamera ile çekilen ligimizdeki bir maç yayınını irdelersek yönetmen, teknik yönetmen, ağır çekim operatörü, ses teknisyeni, kameramanlar, grafik operatörü gibi yaklaşık 20 kişinin görev yaptığı bir operasyon söz konusudur. Uluslararası maçlar gibi daha kapsamlı yayınlarda görevli personel sayısı 100'ü geçebilmektedir. Proje ile amacımız video / görüntü işleme teknolojileri, yapay zeka, nesnelerin ağı gibi teknolojileri kullanarak bir canlı yayın otomasyon sistemi geliştirmektir.
It is aimed to automate and disseminate a process that is still predominantly carried out by operators by using information and communication technologies. Artificial intelligence, machine vision, image and video processing that will take place in the live broadcast automation system that we aim to create operation performed by at least ten personnel in many different profiles such as director, picture selector, cameraman, technical director, graphic operator, sound technician, slow motion operator. technology will be able to be fulfilled with only a few personnel. Our long-term goal is to broadcast without an operator, provided that minimum broadcast quality is provided.
Artificial intelligence elements will be created and implemented for different sports branches. If we take a volleyball match as an example, if the team and player information about the match to be published before the match is available, it will be taken from the federation database, if not, it will be taken from the referee's table, including the first six formations. Cover page including team logos, line-up graphics with team rosters and top sixes, referee information will be given one after another. The game will be watched with the cameras to be placed on the field, and the system will produce suggestions or decide with artificial intelligence about which camera will be selected and broadcast at which moment with which angle and proximity value.
After certain positions, slow motions will still be prepared by the system. As soon as the score is given by the referee and entered by the table referee on the scoreboard, the score chart on the air will be updated, and the cameras will adjust themselves according to the team that will use the service. During the game, the active zone will be monitored by monitoring the athlete and ball movements with special filters (Kalman, etc.). As soon as any team takes a time-out, certain cameras will be able to adjust themselves to show the side of the court and the spectators. In this way, it is aimed to create branch-specific infrastructures.
Considering the example scenario, the detection and tracking of the active region, the image selection function, the slow motion display function include machine vision, image and video processing technologies. Artificial intelligence technologies will be applied to select cameras, automate or semi-automatic camera movements and optimize them according to sports branches.
