• Annual Conference of KIPS
• /
• 2013.11a
• /
• pp.1439-1441
• /
• 2013

본 연구에서는 비디오 감시 시스템을 기반으로 한우 축사에서 승가 행위 검출을 위한 최적의 카메라 위치를 결정한다. 실외 환경에서는 소들간의 겹침이나 조명 변화 등 다양한 어려움이 발생하기 때문에, 이를 극복하기 위하여 승가 시 소의 몸체가 올라간다는 특성을 이용한다. 즉, 등높이가 1.2m에서 1.3m 사이 크기의 한우를 사육하는 축사에서 축사 측면에 1.55m 높이로 카메라를 설치하고 여기서 획득된 영상으로 실험한 결과, 발정기 탐지를 위한 승가 행위를 자동으로 검출할 수 있음을 확인하였다.

• Annual Conference of KIPS
• /
• 2010.11a
• /
• pp.572-575
• /
• 2010

본 논문은 SNR 이 매우 낮은 저조도 영상의 잡음 제거를 위한 새로운 기술을 제안한다. 제안하는 기술은 입력 영상에서 파라미터를 자동/적응적 방식으로 추정하는 알고리즘을 특징으로 한다. 제안하는 기술의 효율성을 검증하기 위해 실질적인 환경에서 취득한 저조도 동영상들을 가지고 실험을 수행하였다. 실험을 통해 제안하는 기술을 활용하여 적응적으로 추정된 파라미터가 필터링(filtering) 성능을 잘 유지시킴을 검증하였다. 또한 기존 연구들과 비교할 때 저조도 동영상의 명암대비 향상과 잡음 제거에 우수한 결과를 보임을 검증하였다.

• Annual Conference of KIPS
• /
• 2020.05a
• /
• pp.536-539
• /
• 2020

길이 좁고 차도와 인도의 구분이 없는 골목길의 특성상 사각지대가 많고 보행자의 동선을 예측하기 힘들어 교통사고가 많이 발생하고 있다. 따라서 본 논문에서는 AI 를 활용, 영상 내 사물을 추적하여 골목길에서의 사고를 예방하는 시스템을 제안한다. 해당 시스템은 Object - Detection & Tracking 을 사용하여 보행자 및 차량을 식별·추적하여 두 개 이상의 사물이 동시에 교차로에 접근 시 사고 예방 알람을 발생시킨다. 이 시스템을 전국에 설치되어 있는 CCTV 에 활용하면 추가적인 비용과 설치 시간에 제한받지 않고 전국적으로 응용할 수 있을 것으로 기대된다.

• The KIPS Transactions:PartB
• /
• v.10B no.6
• /
• pp.657-664
• /
• 2003

Reliable tracking of moving humans is essential to motion estimation, video surveillance and human-computer interface. This paper presents a new approach to human motion tracking that combines appearance-based and model-based techniques. Monocular color video is processed at both pixel level and object level. At the pixel level, a Gaussian mixture model is used to train and classily individual pixel colors. At the object level, a 3D human body model projected on a 2D image plane is used to fit the image data. Our method does not use inverse kinematics due to the singularity problem. While many others use stochastic sampling for model-based motion tracking, our method is purely dependent on nonlinear programming. We convert the human motion tracking problem into a nonlinear programming problem. A cost function for parameter optimization is used to estimate the degree of the overlapping between the foreground input image silhouette and a projected 3D model body silhouette. The overlapping is computed using computational geometry by converting a set of pixels from the image domain to a polygon in the real projection plane domain. Our method is used to recognize various human motions. Motion tracking results from video sequences are very encouraging.

• Journal of the Institute of Convergence Signal Processing
• /
• v.7 no.3
• /
• pp.95-101
• /
• 2006

A face detection plays an important role in face recognition, video surveillance, and human computer interaction. In this paper, we propose a progressive threshold method to detect human faces in real time. The consecutive face images are acquired from camera and transformed into YCbCr color space images. The skin color of the input images are separated using a skin color filter in the YCbCr color space and some candidated face areas are decided by connected component analysis. The intensity equalization is performed to avoid the effect of many circumstances and an arbitrary threshold value is applied to get binary images. The eye area can be detected because the area is clearly distinguished from others in the binary image progressive threshold method searches for an optimal eye area by progressively increasing threshold from low values. After progressive thresholding, the eye area is normalized and verified by back propagation algorithm to finalize the face detection.

• Journal of Broadcast Engineering
• /
• v.16 no.6
• /
• pp.1018-1025
• /
• 2011

Scalable Video Coding (SVC) extension of H.264/AVC is a new video coding standard for media convergence by providing diverse videos of different spatial-temporal-quality layers with a single bitstream. Recently, real-time SVC codecs are being developed for the application areas of surveillance video and mobile video, etc. This paper presents the design and implementation of a H.264/SVC decoder based on an embedded DSP using Open SVC Decoder (OSD) which is a real-time software decoder designed for the PC environment. The implementation consists of porting C code of the OSD software from PC to DSP environment, profiling the complexity performance of OSD with further optimization, and integrating the optimized decoder into the TI Davinci EVM (Evaluation Module). 50 QCIF/CIF frames or 15 SD frames per second can be decoded with the implemented DSP-based SVC decoder.

• Journal of the Institute of Electronics Engineers of Korea CI
• /
• v.47 no.2
• /
• pp.29-38
• /
• 2010

Network Video Recorder becomes popular as a next generation surveillance system connecting all cameras and video server in network environment because it can provide ease of installation and efficient management and maintenance. But in case of data damage, the storage device in traditional NVR has no recovery scheme and it is disabled in processing real-time requests. In this paper, we propose an Reliable Storage using Parity and Mirroring scheme for improving reliability on storage device and maintaining system on realtime. RSPM uses a Liberation coding to recover damaged multimedia data and dynamic mirroring to repair corrupted system data and to maintain real-time operation. RSPM using the Liberation code is 11.29% lesser than traditional file system and 5.21% less than RSPM using parity code in terms of loss rate of damaged multimedia data.

• Journal of Broadcast Engineering
• /
• v.14 no.4
• /
• pp.409-418
• /
• 2009

This paper proposes a method to estimate the flow speed of pedestrians in surveillance videos. In the proposed method, the average moving speed of pedestrians is measured by estimating the size of real-world motion from the observed motion vectors. For this purpose, a pixel-to-meter conversion factor is introduced which is calculated from camera parameters. Also, the height information, which is missing because of camera projection, is predicted statistically from simulation experiments. Compared to the previous works for flow speed estimation, our method can be applied to various camera views because it separates scene parameters explicitly. Experiments are performed on both simulation image sequences and real video. In the experiments on simulation videos, the proposed method estimated the flow speed with average error of about 0.08m/s. The proposed method also showed promising results for the real video.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.12 no.6
• /
• pp.135-139
• /
• 2012

The wide diffusion of internet and the overall increased reliance on technology for information communication, dissemination and gathering have created an unparalleled mass of data. Sifting through this data is defining and will define in the foreseeable future a big part of contemporary computer science. Within this data, a growing proportion is given by personal information, which represents a unique opportunity to study human activities extensively and live. One important recurring challenge in many disciplines is the problem of people re-identification. In its broadest definition, re-identification is the problem of newly recognizing previously identified people, such as following an unknown person while he walks through many different surveillance cameras in different locations. Our goals is to review how several diverse disciplines define and meet this challenge, from person re-identification in video-surveillance to authorship attribution in text samples to distinguishing users based on their preferences of pictures. We further envision a situation where multidisciplinary solutions might be beneficial.

• Journal of Advanced Marine Engineering and Technology
• /
• v.39 no.9
• /
• pp.953-958
• /
• 2015

Raspberry Pi is a credit card-sized computer with support for a large number of input and output peripherals. This makes it the perfect platform for interaction with many different devices and for usage in a wide range of applications. When combined with Wi-Fi, it can communicate remotely, therefore increasing its suitability for the construction of wireless sensor nodes. In addition, data processing and decision-making can be based on artificial intelligence, what is performed in developed testbed on the example of monitoring and determining the confidence of fire. In this paper, we demonstrated the usage of Raspberry Pi as a sensor web node for fire-safety monitoring in a building. When the UV-flame sensors detect a flame as thin as that of a candle, the Raspberry Pi sends a push-message to notify the assigned smartphone of the on-site situation through the GCM server. A mobile app was developed to provide a real-time video streaming service in order to determine a false alarm. If an emergency occurs, one can immediately call for help.