• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.15 no.1
• /
• pp.418-424
• /
• 2014

In this paper, we propose a boundary input method for video surveillance systems. Since intrusion of a moving object is decided by comparition of its position and the surveillance boundary, the boundary input method is a basic function in video surveillance. Previous methods are difficult to adapt to the change of surveillance environments such as the size of surveillance area, the number of cameras, and the position of cameras because those build up the surveillance boundary using the captured image in the center of each camera. In our approach, the whole surveillance boundary is once defined in the form of polygon based on the satellite map and transformed into each camera environment. Its characteristics is that the boundary input is independent from the surveillance environment. Given the position of a moving object, the time complexity of its intrusion detection shows O(n), where n is the number of polygon vertices. To verify our method, we implemented a 3D simulation and assured that the input boundary can be reused in each camera without any redefinition.

• The KIPS Transactions:PartB
• /
• v.10B no.1
• /
• pp.95-102
• /
• 2003

In surveillance video system, there are many classes of images and some spatial regions are more important than other regions. The conventional compression method in this system have been compressed there full frames without classfying them depend on their important parts. To improve the accuracy of the image coding and deliver effective compression for the surveillance video system, it was necessary to separate the regions according to their importance. In this paper, we propose a new effective surveillance video image compression method. The proposed scheme defines importance based three-level region of interest block in a frame, such as background, motion object block, and the feature object block. A captured video image frame can be separated to these three different levels of block regions. And depends on the priority, each block can be modified and compressed in different resolution, compression ratio and qualify factor. Therefore, in surveillance video system, this algorithm not only reduces the image processing time and space, but also guarantees the Important image data in high quality to acquire the system's goal.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.19 no.5
• /
• pp.1084-1090
• /
• 2015

In this paper, we propose a Tabu search algorithm to efficiently deploy the sensor nodes for maximizing the network sensing coverage in wireless sensor networks. As the number of the sensor nodes in wireless sensor networks increases, the amount of calculation for searching the solution would be too much increased. To obtain the best solution within a reasonable execution time in a high-density network, we propose a Tabu search algorithm to maximize the network sensing coverage. In order to search effectively, we propose some efficient neighborhood generating operations of the Tabu search algorithm. We evaluate those performances through some experiments in terms of the maximum network sensing coverage and the execution time of the proposed algorithm. The comparison results show that the proposed algorithm outperforms other existing algorithms.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2002.11a
• /
• pp.235-239
• /
• 2002

본 논문에서는 감시 시스템의 기록과 검색의 효율성을 높이기 위해 감시 영상이 가지는 특징을 충분히 고려하여 영상 처리를 적용하였으며 이를 위해 배경 영역과 감시할 물체 영역을 분리하여 검색하는 시스템을 설계한다. 제안한 영상 시스템은 CCD 카메라로부터 영상을 입력받아 먼저 감시 대상의 배경화면과 감시할 대상인 물체 영역을 추출하여 카메라에서 이동물체의 생성시 정지 영상으로 기록함으로서 필요시에 즉시 시간별 또는 장소별로 감시 영상을 실시간적으로 효율적인 검색이 가능한 시스템을 구성하였다.

• Journal of Satellite, Information and Communications
• /
• v.7 no.3
• /
• pp.140-144
• /
• 2012

The classic method of recording and retrieving of images for video monitoring was very inefficient since the system required huge storage capacity. It's because it stored images without separating background and monitored objects. The recording and retrieval method of images for the video monitoring system we proposed in this paper makes it possible to retrieve images easily by using database by the hour and the place. Hence, it retrieve images much more efficiently than the existing method.

• Proceedings of the Korea Institute of Convergence Signal Processing
• /
• 2000.12a
• /
• pp.157-160
• /
• 2000

본 논문에서는 교통 감시 시스템의 필수 단계중에 하나인 실시간 자동차 검출 방법을 제안한다. 제안한 방법은 후보 영역 추출 단계와 자동차 인식 단계로 이루어진다. 첫 번째 단계에서는 연속된 두 프레임간의 차영상 분석 방법을 기반으로 하여 움직임이 있는 후보 영역을 추출한다. 두 번째 단계에서는 추출된 후보 영역에 자동차가 포함되어 있는지를 판별하기 위해 웨이블릿 변환 계수들을 입력으로 하는 신경망을 사용한다. 일반 도로에서 획득한 230대의 자동차가 포함된 동영상을 실험한 결과, 자동차 검출율은 97.8%, 프레임당 처리 시간은 0.12ms이다. 본 논문에서 제안한 실시간 자동차 검출 방법은 교통 감시 시스템에 유용하게 적용될 수 있다.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2010.11a
• /
• pp.70-71
• /
• 2010

요즘 감시카메라 시장은 전 세계적으로 이슈가 되고 있다. 감시카메라는 적은 인원으로 많은 장소를 한 눈에 감시할 수 있고, 문제가 발생했을 때, 녹화 저장된 영상을 통해 그 상황을 다시 볼 수 있다. 이렇게 다양한 기능과 편리함으로 우리에게 도움을 주는 감시카메라이지만, 마스크나 선글라스, 또는 여러 가지 잡음에 의해 얼굴 영상의 부분이 훼손되는 상황에서는 신원 확인을 하기가 어렵다. 때문에 가려진 얼굴을 제대로 인식하기 위해 영상 처리 분야에서 얼굴의 가려진 영역을 찾아 그 부분을 재구성하고자 하는 연구가 활발히 진행되고 있다. 이에 본 논문은 기존 PCA 방법을 이용하여 가려진 영역을 찾아내고, PCA를 반복적으로 사용하여 재구성하는 대신 상관관계를 이용하여 얼굴의 가려진 영역을 자동적으로 검출하고 복원하는 방법을 제안한다. 본 논문에서는 두 눈의 중심이 고정되어 있는 BioID 데이터로 상관계수를 구하고 얼굴의 특정 부분을 임의로 가려 실험을 수행하였다. 제안된 방법의 결과는 PCA 방법으로 수행한 결과와 함께 비교되어 원본 영상과의 오류 값이 더 작게 나오는 것을 확인할 수 있었다.

• Journal of Navigation and Port Research
• /
• v.42 no.3
• /
• pp.201-206
• /
• 2018

The purpose of this study is to estimate a proper Domain Watch between anchored vessels in order to propose a method for the efficient management of VTS(Vessel Traffic Service) of the N-anchorage in Busan harbor, which is the largest port in Korea. For this purpose, we proposed the calculation method of Domain Watch and investigated the ship length(L), the distance between anchored vessels ($D_{ij}$), the domain radius(R), and the domain radius vs L(R/L) during the peak time of the vessels in the latest usage of anchorage. As a result of technical analysis for the surveyed data, the minimum R/L for securing the safety distance between anchored vessels was selected based on 2.85 corresponding to the 70th percentile of the total data. This result was applied to the N-anchorage of Busan and compared with the 'Guidelines of Port and Harbor Design(2014)', and we have confirmed that it is reasonable to set the Domain radius with the minimum 2.85L or more in VTS. This study considers the safety management of anchorage for VTS. This study could contribute to the safety of vessels using anchorage and the safety management plan of VTS when it is applied to other ports in operation such as it was in Busan.

• Journal of the Korea Institute of Information Security & Cryptology
• /
• v.21 no.6
• /
• pp.171-181
• /
• 2011

Proposed in this paper is the intelligent video surveillance system to effectively detect multiple loitering objects even that disappear from the out of camera's field of view and later return to a target zone. After the background and foreground are segmented using Gaussian mixture model and shadows are removed, the objects returning to the target zone is recognized using the chromaticity histogram and the duration of loitering is preserved. For more accurate measurement of the loitering behavior, the camera calibration is also applied to map the image plane to the real-world ground. Hence, the loitering behavior can be detected by considering the time duration of the object's existence in the real-world space. The experiment was performed using loitering video and all of the loitering behaviors are accurately detected.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.13 no.4
• /
• pp.787-794
• /
• 2018

The various natural disasters such as floods and localized heavy rains are increasing due to the global warming. If a natural disaster can be detected and analyzed in advance and more effectively, it can prevent enormous damage of natural disasters. Recent development in visual sensor technologies has encouraged various studies on monitoring environments including rivers. In this paper, we propose a method to detect water regions from river images which can be exploited for river surveillance systems using video sensor networks. In the proposed method, we first segment a river image finely using the minimum spanning tree algorithm. Then, the seed regions for the river region and the background region are set by using the preliminary information, and each seed region is expanded by merging similar regions to segment the water region from the image. Experimental results show that the proposed method separates the water region from a river image easier and accurately.