ETRI-Knowledge Sharing Plaform

ENGLISH

성과물

논문 검색
구분 SCI
연도 ~ 키워드

상세정보

학술대회 Automatic High-Speed Flying Ball Detection from Multi-Exposure Images under Varying Light Conditions
Cited 8 time in scopus Download 0 time Share share facebook twitter linkedin kakaostory
저자
김종성, 김명규
발행일
201112
출처
International Conference on Virtual Reality Continuum and Its Applications in Industry (VRCAI) 2011, pp.549-552
DOI
https://dx.doi.org/10.1145/2087756.2087861
협약과제
11CT1300, 스포츠 아케이드 게임을 위한 회전궤적인식 기술, 김명규
초록
In this paper, we propose an automatic technique for high-speed flying ball detection from multi-exposure images under varying light conditions. The proposed technique is a two-step process. First, a new automatic image thresholding process is employed to obtain a preliminary ball region from a multi-exposure image, where a background subtraction method based on our spatial background model is combined to eliminate the effect of varying light conditions. Next, a bottom-up image labeling process is applied for obtaining an accurate segmentation of the final ball region. In addition, an efficient multi-grid scheme is introduced to accelerate the processing speed. The effectiveness and efficiency of the proposed technique was tested against previous methods with real experiments under various light conditions. Experimental results show that the proposed technique is capable of automatic and high-speed detection. © 2011 ACM.
KSP 제안 키워드
Background model, Background subtraction(BS), Effectiveness and efficiency, High Speed, Image Thresholding, Light conditions, Multi-grid, Processing speed, Speed detection, Subtraction method, bottom-up