• Proceedings of the Korea Information Processing Society Conference
• /
• 2012.11a
• /
• pp.510-511
• /
• 2012

지능형 자동차의 차량 안전시스템의 요구가 증가되면서 FCW(Frontal Collision Warning)에 대한 중요도가 높아지고 있다. 이미 몇몇 회사들이 상용화에 성공한 위 기술은 단일 카메라, 복수의 카메라, 또는 라이더와 카메라의 정보 혼합 방식 등을 사용하고 있다. 본 논문은 단일 카메라 기반 방식에서 차량 후보를 필터링하는 방법에 대해 소개한다

• Journal of Broadcast Engineering
• /
• v.22 no.6
• /
• pp.808-816
• /
• 2017

360 VR video systems has become important to provide immersive effect for viewers. The system consists of stitching, projection, compression, inverse projection, viewport extraction. In this paper, an efficient luminance compensation technique for 360 VR video sequences, where feature extraction and histogram equalization algorithms are utilized. The proposed luminance compensation algorithm enhance the performance of stitching in 360 VR system. The simulation results showed that the proposed technique is useful to increase the quality of the displayed image.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• fall
• /
• pp.124-125
• /
• 2021

최근 증강현실(AR), 가상현실(VR), 혼합현실(XR) 분야가 각광받고 있으며, 3차원 공간과 사물을 인식하여 다양한 콘텐츠 서비스를 제공하는 기술이 개발되고 있다[1]. 3차원 공간과 사물을 인식하기 위해 가장 널리 사용되는 방법은 RGB 카메라를 이용하는 것이다[2]. RGB 카메라를 이용하여 촬영한 영상을 분석한 후 분석된 결과를 이용하여 카메라와 환경의 관계를 추정한다. 시차는 사용자가 촬영한 복수의 이미지에서 특징점의 차이를 이용하여 계산된다. 실험적으로 구한 깊이에 대해 계산된 디스패리티에 시차 정보와 스케일링 정보를 더하여 3차원 특징점을 생성한다. 제안하는 알고리즘은 단일 모바일 디바이스에서 획득한 영상을 사용한다. 특징점 매칭을 기반으로한 디스패리티 추정과 시차조정 3D 특징점 생성이다. 실제 깊이 값과 비교했을 때, 생성된 3차원 특징점은 실측값의 10% 이내의 오차가 있음을 실험적으로 증명하였다. 따라서 제안하는 방법을 이용하여 유효한 3차원 특징점을 생성할 수 있다.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.6 no.3
• /
• pp.45-56
• /
• 2007

Automatic tracking of vehicles is important to accurately estimate the traffic information including vehicle speeds in video-based traffic measurement systems. Because of the limited field of view, the range of visual tracking with a single camera is restricted. In order to enlarge the tracking range for better chance of monitoring the vehicle behaviors, a tracking with consecutive multiple video sequences is necessary. This parer proposes a carefully designed rule-based vehicle racking scheme and apply it for the tracking for two well synchronized video sequences. In the scheme, almost all possible cases that can appear in the video-based vehicle tracking are considered to make rules. Also, the rule based scheme is augmented with Kalman filter. The result of tracking can be successfully used to collect data such as temporal variation of vehicle speed and behavior of individual vehicle behaviors in the enlarged tracking region.

• Journal of the Korean Society for Aeronautical & Space Sciences
• /
• v.37 no.12
• /
• pp.1209-1216
• /
• 2009

This paper presents the pose estimation of a small UAV utilizing visual information from low cost cameras installed indoor. To overcome the limitation of the outside flight experiment, the indoor flight test environment based on multi-camera systems is proposed. Computer vision algorithms for the proposed system include camera calibration, color marker detection, and pose estimation. The well-known extended Kalman filter is used to obtain an accurate position and pose estimation for the small UAV. This paper finishes with several experiment results illustrating the performance and properties of the proposed vision-based indoor flight test environment.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.25 no.1
• /
• pp.35-41
• /
• 2015

When a video surveillance system tracks a specific object, it is very important to get quickly the information of the object through fast image processing. Usually one camera surveillance system for tracking the object made results in various problems such like occlusion, image noise during the tracking process. It makes difficulties on image based moving object tracking. Therefore, to overcome the difficulties the multi video surveillance system which installed several camera within interested area and looking the same object from multi angles of view could be considered as a solution. If multi cameras are used for tracking object, it is capable of making a decision having high accuracy in more wide space. This paper proposes a method of recognizing and tracking a specific object like a car using the homography in which multi cameras are installed at the crossroad.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.26 no.12A
• /
• pp.2067-2073
• /
• 2001

Biometric systems are forms of technology that use unique human physical characteristics to automatically identify a person. They have sensors to pick up some physical characteristics, convert them into digital patterns, and compare them with patterns stored for individual identification. However, lip-print recognition has been less developed than recognition of other human physical attributes such as the fingerprint, voice patterns, retinal at blood vessel patterns, or the face. The lip print recognition by a CCD camera has the merit of being linked with other recognition systems such as the retinal/iris eye and the face. A new method using multi-resolution architecture is proposed to recognize a lip print from the pattern kernels. A set of pattern kernels is a function of some local lip print masks. This function converts the information from a lip print into digital data. Recognition in the multi-resolution system is more reliable than recognition in the single-resolution system. The multi-resolution architecture allows us to reduce the false recognition rate from 15% to 4.7%. This paper shows that a lip print is sufficiently used by the measurements of biometric systems.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2013.06a
• /
• pp.40-42
• /
• 2013

본 논문에서는 사각 평면이 아닌 실린더나 구 형태의 비평면 스크린에 고품질 파노라마 영상을 렌더링 하기 위해 스틸 카메라를 이용한 자동 캘리브레이션과 멀티 프로젝터를 이용한 파노라마 렌더링 시스템을 제안한다. 스틸 카메라를 이용한 비평면 스크린 캘리브레이션은 특정 패턴을 스크린에 투사하고 이를 촬영하여 프로젝터의 투사 왜곡을 보정하고 멀티 프로젝터간의 중첩 영역에 대한 보정을 자동 수행한다. 캘리브레이션이 완료된 이후에 멀티 프로젝터를 이용하여 고품질의 파노라마 비디오를 렌더링 하기 위해서 본 논문에서는 복수의 멀티 시스템과 동기화 카드를 이용하여 렌더링 시스템간의 재생 동기화를 수행 하였다. 제안하는 렌더링 시스템을 적용하면 기존의 전문가의 수동작업에 의한 인력 및 보정 시간을 줄일 수 있으며, 시스템 환경이 변하더라도 쉽게 적응적으로 렌더링 환경을 구축하는 것이 가능하고 고품질의 파노라마 비디오를 렌더링 하는 것이 가능하다.

• The Journal of the Convergence on Culture Technology
• /
• v.9 no.5
• /
• pp.605-611
• /
• 2023

Accidents at construction sites have a very high rate of fatalities due to the nature of being prone to severe injury patients. In order to reduce the mortality rate of severely injury patients, quick response is required, and some systems that detect accidents using AI technology and cameras have been devised to respond quickly to accidents. However, since existing accident detection systems use only a single camera, there are blind spots, Thus, they cannot detect all accidents at a construction site. Therefore, in this paper, we present the system that minimizes the detection blind spot by using multiple cameras. Our implemented system extracts feature points from the images of multiple cameras with the YOLO-pose library, and inputs the extracted feature points to a Long Short Term Memory-based recurrent neural network in order to detect accidents. In our experimental result, we confirme that the proposed system shows high accuracy while minimizing detection blind spots by using multiple cameras.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.44 no.1
• /
• pp.1-9
• /
• 2007

In this paper, we propose a new 3D modeling and rendering system using multiple stereo cameras. When target objects are captured by cameras, each capturing PC segments the objects and estimates disparity fields, then they transmit the segmented masks, disparity fields, and color textures of objects to a 3D modeling server. The modeling server generates 3D models of the objects from the gathered masks and disparity fields. Finally, the server generates a video at the designated point of view with the 3D model and texture information from cameras.