• Proceedings of the Korean Society of Precision Engineering Conference
• /
• 1996.11a
• /
• pp.928-932
• /
• 1996

Single overload fatigue tests with overload sizes ranging from 50% and 100% have been performed to investing ate the fatigue crack growth retardation behavior. A modified and experimental method of Willenborg's model for prediction of crack growth retardation behavior has been developed, based on evaluations of equivalent plastic zone size (EPZS) changing its size along the overload plastic zone boundary. The minimum crack growth rates of each overload size are linearly decreased with overload size increasing, but fatigue lives extended by single overload are increasing much more unlike the crack growth rates. Comparisons of crack growth behavior predicted by EPZS model and Willenborg model have shown that the EPZS model accounts for overload effects better than Willenborg model. These effects include delayed retardation, large retardation region, minimum crack growth rate, and the increase rate of crack growth rate in the region crack growth rate recovered.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2011.10a
• /
• pp.157-160
• /
• 2011

In this paper, We proposed the method of recognition and analysis of human activites using Motion vector in real-time surveillance system. We employs subtraction image techniques to detect blob(human) in the foreground. When MPEG-4 video recording EPZS(Enhanced Predicted Zonal Search) is detected the values of motion vectors were used. In this paper, the activities of human recognize and classified such as meta-classes like this {Active, Inactive}, {Moving, Non-moving}, {Walking, Running}. Each step was separated using a step-by-step threshold values. We created approximately 150 conditions for the simulation. As a result, We showed a high success rate about 86~98% to distinguish each steps in simulation image.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.11 no.4
• /
• pp.779-785
• /
• 2007

Moving object detection is very important for video surveillance in modern days. In special case, we can categorize motions into two types-salient and non-salient motion. In this paper, we first calculate temporal difference image for extract moving objects and adapt to dynamic environments and next, we also propose a new algorithm to detect salient motion information in complex environment by combining temporal difference image and binary block image which is calculated by motion vector using the newest MPEG-4 and EPZS, and it is very effective to detect objects in a complex environment that many various motions are mixed.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• v.9 no.2
• /
• pp.336-339
• /
• 2005

동영상에서의 움직임 검출에 대한 기술은 다양한 응용분야를 가지고 있다. 특히 고정된 카메라에서 움직임 검출과 움직이는 카메라에 대응하여 움직임 검출하고 추적하는 기법들은 현재까지 많은 연구들이 진행되고 있고, 또 응용 제품들의 출시되어지고 있다. 그렇지만 일반적이고 보다 완전한 제품의 개발은 아직 한계가 있다. 본 논문에서는 움직임 벡터를 이용하여 카메라가 패닝(Panning) 여부를 판단하여 패닝 시에는 움직임 검출을 하지 않는 기법을 제안한다. 이 기법은 카메라 패닝 시 즉, 카메라 팬/틸트 드라이버의 고장이나 바람등의 상황에는 움직임 검출을 하지 않음으로써 보다 효과적인 움직임 검출을 할 수 있다. 또한 두 개의 임계치를 적응적으로 조절할 수 있어 기존의 고정된 카메라에서 사용하는 움직임 검출 기법들보다 효율적이라고 할 수 있다.