• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 1998.11a
• /
• pp.525-528
• /
• 1998

본 논문은 현재 표준화 작업이 진행되고 있는 MS사의 M-JPEG을 이용하여 녹화 및 재생이 가능한 감시 시스템의 구현에 대해 설명한다. 기존 아날로그 감시 시스템의 단점인 여러번 재생시 화질 저하와 카메라에 비디오를 연결해야 하는 비효율성을 극복하기 위해 개발된 이 시스템은 사용자들에게 최대의 감시 효율을 제공할 수 있다. 이 무인 감시 시스템은 디지털 기술을 활용하여 모든 화상 처리와 화면 전환을 컴퓨터로 처리함으로서 여러 가지 다양한 모드의 감시환경을 제공하고 뛰어난 화면 capture기능으로 선명한 화질을 제공하여 감시 효과를 극대화시킬 수 있고 원하는 장면을 찾기 위한 검색 기능을 제공함으로서 사용자들에게 편리함을 제공할 수 있다. 본 시스템 구현에 사용된 H/W는 펜티엄-II 333과 카메라 16대 그리고 스위치 보드, MPEG capture 보드를 사용했으며 구동 S/W는 Visual C++로 제작하였다.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.25 no.5B
• /
• pp.913-924
• /
• 2000

This paper proposes an efficient descriptor for objects motion trajectory and a video retrieval algorithm based on objects motion trajectory. The algorithm describes parameters with coefficients of 2-order polynomial for objects motion trajectory after segmentation of the object from the scene. The algorithm also identifies types, intervals, and magnitude of global motion caused by camera motion and indexes them with 6-affine parameters. This paper implements content-based video retrieval using similarity-match between indexed parameters and queried ones for objects motion trajectory. The proposed algorithm will support not only faster retrieval for general videos but efficient operation for unmanned video surveillance system.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2010.11a
• /
• pp.254-256
• /
• 2010

최근 감시시스템 구축에서 디지털 비디오 레코더(DVR)를 이용하는 방식이 주를 이루고 있다. 영상압축 기술의 발전에 따라 디지털 비디오 레코더는 H.264/AVC 인코더를 기반으로 저장 공간 대비 고화질의 영상 저장이 가능해졌다. 또한, 다수 채널의 동시 처리가 가능해졌을 뿐만 아니라 영상 압축 및 저장 기능 외에 지능형 감시를 위한 기능이 요구되고 있다. 본 논문에서는 H.264/AVC 압축 비트열에서 움직임 벡터 추출 및 분석을 통한 영상내 이동 물체 검출 방법을 제안한다. 압축 비트열에서 현재 프레임과 이전 프레임의 움직임 벡터들의 크기와 각도를 비교하여 임계값 이하인 움직임 벡터들을 추출하여 이동 물체를 검출하였다. 이동 물체 검출 성능 평가를 위해 H.264/AVC의 참조 프로그램인 JM 15.1을 사용하여 인코딩 및 비트열 분석을 수행 하였고, 움직임이 검출된 시간 구간만을 디코딩 후 저장하는 알고리즘을 구현하였다. 제안된 방식은 디코딩 부분의 연산량 감소뿐만 아니라 저장소의 물리적 공간 절약이 가능하다.

• The KIPS Transactions:PartB
• /
• v.13B no.5 s.108
• /
• pp.515-524
• /
• 2006

Background extraction is an important technique to find the moving objects in video surveillance system. This paper proposes a new on-line background extraction method for color video using vector order statistics. In the proposed method, using the fact that background occurs more frequently than objects, the vector median of color pixels in consecutive frames Is treated as background at the position. Also, the objects of current frame are consisted of the set of pixels whose distance from background pixel is larger than threshold. In the paper, the proposed method is compared with the on-line multiple background extraction based on Gaussian mixture model(GMM) in order to evaluate the performance. As the result, its performance is similar or superior to the method based on GMM.

• Journal of Sensor Science and Technology
• /
• v.11 no.1
• /
• pp.28-38
• /
• 2002

People are important targets for video surveillance and monitoring(VSAM) but difficult to be analyzed. In this paper, a technique to count people in image sequences is dealt as a prerequisite procedure for automatic tracking and behaviour analysis. A group of people is divided at local minima of the line connecting the highest pixels on the binary image of the people extracted from the image taken by a stationary video camera. As the properties of the divided regions vary according to the relative positions of the people in a group, different states are assigned for the completely occluded, partially occluded, completed separated individual, and wrongly divided regions. By analyzing the transition of the states of divided regions, the number of people on the site monitored is estimated. The technique is checked in real experimental situations.

• Journal of the Korea Institute of Information Security & Cryptology
• /
• v.21 no.6
• /
• pp.161-169
• /
• 2011

Recent intelligent video surveillance system asks for development of more advanced technology for analysis and recognition of video data. Especially, machine learning algorithm such as Support Vector Machine (SVM) is used in order to recognize objects in video. Because SVM training demands massive amount of computation, parallel processing technique is necessary to reduce the execution time effectively. In this paper, we propose a parallel processing method of SVM training with a multi-core processor. The results of parallel SVM on a 4-core processor show that our proposed method can reduce the execution time of the sequential training by a factor of 2.5.

• Journal of Advanced Navigation Technology
• /
• v.12 no.1
• /
• pp.68-73
• /
• 2008

Recently, remote observation and control technology has grown gradually. According to use home network system and expand super-highway information network, digital video technology observation system gradually has increased. In order that the user conveniently use better than to use web browser observation system, it uses existing digital observation system, personal digital assistant and control module in the place of internet.

• Smart Media Journal
• /
• v.13 no.4
• /
• pp.65-75
• /
• 2024

This paper proposes a new approach to video surveillance systems, which have become essential components in modern urban management. By utilizing blockchain and IPFS, it enhances data integrity and privacy protection. Additionally, anomaly detection and automatic video storage are enabled through object detection technology, thus improving urban safety and security. This integrated approach serves as an efficient management methodology for surveillance systems, providing city administrators and citizens with a safer and more effective monitoring environment.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.45 no.6
• /
• pp.108-116
• /
• 2008

In this paper, we propose the algorithm for making panoramic background and object tacking using pan-tilt-zoom camera. We draw an analogy relation between images for cylinder projection, rearrange of images, stitching, and blending. We can then make the panoramic background, and can track the object use the panoramic background. After generated the background, the proposed algorithm tracks the moving object. Therefore it can detect the wide area, and it tracks the object continuously. So the proposed algorithm is able to use at wide area to detect and track the object.

• Journal of Korea Multimedia Society
• /
• v.8 no.10
• /
• pp.1314-1321
• /
• 2005

In video surveillance and monitoring (VSAM), extracting foreground by detecting moving regions is the most fundamental step. The foreground extracted, however, includes not only objects in motion but also their shadows, which may cause errors in following video image processing steps. To remove the shadows, this paper presents a new technique to determine shadow pixels in the foreground image of a VSAM camera system. The proposed technique utilizes a fact that the effect of shadowing to each pixel is different defending on its brightness in a background image when determining shadow pixels unlike existing techniques where unified decision criteria are used to all pixels. Such an approach can easily accommodate local features in an image and hold consistent Performance even in changing environment. In real experiments, the proposed technique showed better results compared with an existing technique.