상세정보
학술대회
A Comparative Study on Preprocessing Methods for Object Tracking in Sports Events
- 발행일
- 201802
- 출처
- International Conference on Advanced Communications Technology (ICACT) 2018, pp.460-462
- 협약과제
- 17CS1200, 스포츠 영상 콘텐츠의 내용 이해 기반 분석/요약/검색 기술 개발, 남도원
- 초록
- Due to the development of machine learning, interest in automatic recognition of objects in videos is increasing. In the sports field, there is a growing demand for a technology for automatically recognizing and tracking various objects in a sports game video. However, it is very difficult to recognize and track an object without proper preprocessing due to its small size and large sports field. In this paper, we compare background substraction and edge detection, which are typical preprocessing methods for automatic recognition and tracking of objects in sports images, and analyze advantages and disadvantages of each method.
- KSP 제안 키워드
- Automatic recognition, Edge Detection, Recognition of objects, Sports events, advantages and disadvantages, background substraction, comparative study, machine Learning, object Tracking, preprocessing methods, small size