• Journal of the Institute of Convergence Signal Processing
• /
• v.12 no.1
• /
• pp.13-19
• /
• 2011

In this paper, a visual surveillance system using SOPC based NIOS II embedded processor and C2H compiler was implemented. In this system, the IP is constructed by C2H compiler for the output of the camera images, image processing, serial communication and network communication, then, it is implemented to effectively control each IP based on the SOPC and the NIOS II embedded processor. And, an algorithm which updates the background images for high speed and robust detection of the moving objects is proposed using the Adaptive Gaussian Mixture Model(AGMM). In results, it can detecte the moving objects(pedestrians and vehicles) under day-time and night-time. It is confirmed that the proposed AGMM algorithm has better performance than the Adaptive Threshold Method(ATM) and the Gaussian Mixture Model(GMM) from our experiments.

• Journal of Internet Computing and Services
• /
• v.18 no.6
• /
• pp.85-92
• /
• 2017

Recently, in the field of video surveillance, a Deep Learning based learning method has been applied to a method of detecting a moving person in a video and analyzing the behavior of a detected person. The human activity recognition, which is one of the fields this intelligent image analysis technology, detects the object and goes through the process of detecting the body keypoint to recognize the behavior of the detected object. In this paper, we propose a method for Body Keypoint Localization based on Object Detection using RGB-D information. First, the moving object is segmented and detected from the background using color information and depth information generated by the two cameras. The input image generated by rescaling the detected object region using RGB-D information is applied to Convolutional Pose Machines for one person's pose estimation. CPM are used to generate Belief Maps for 14 body parts per person and to detect body keypoints based on Belief Maps. This method provides an accurate region for objects to detect keypoints an can be extended from single Body Keypoint Localization to multiple Body Keypoint Localization through the integration of individual Body Keypoint Localization. In the future, it is possible to generate a model for human pose estimation using the detected keypoints and contribute to the field of human activity recognition.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.2 no.2 s.3
• /
• pp.85-95
• /
• 2003

This paper proposes a multiple moving objects tracking system which is adaptable itself to circumstances. Snake model is sensitive to the start position value because it does not accurately express contours of objects in complex image. It can be improved as the proposed system gets background images by using difference images, segments objects using neighborhood pixels and assesses the position feature values acquired on the start position value to deformable Snake model. And also the system can simplify complex background images and reduce search regions by the constituent points of a Snake laid in Positions of object. It is showed that the proposed system can be appBied to multiple moving vehicle racking systems by the experimental results of 30fps AVI file.

• Annual Conference of KIPS
• /
• 2019.10a
• /
• pp.19-20
• /
• 2019

서버리스 프레임워크(Serverless Framework)는 마이크로서비스 아키텍처의 이론을 클라우드와 컨테이너를 기반으로 구현한 것으로 아마존의 AWS(Amazon Web Service)와 같은 퍼블릭 클라우드 플랫폼이 서비스됨에 따라 활용도 높아지고 있다. 하지만 현재까지의 플랫폼들은 GPU 와 같은 하드웨어의 의존성을 가진 인공지능 모델의 서비스에는 지원이 부족하다. 이에 본 논문에서는 컨테이너 기반의 오픈소스 서버리스 플랫폼을 대상으로 엔비디어-도커와 k8s-device-plugin 을 적용하여 GPU 활용이 가능한 서버리스 플랫폼을 구현하였다. 또한 인공지능 모델이 컨테이너에서 구동될 때 반복되는 가중치 로드를 줄이기 위한 구조를 제안한다. 본 논문에서 구현된 서버리스 플랫폼은 객체 검출 모델인 SSD(Single Shot Multibox Detector) 모델을 이용하여 성능 비교 실험을 진행하였으며, 그 결과 인공지능 모델이 적용된 서버리스 플랫폼의 함수 응답 시간이 개선되었음을 확인하였다.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2021.07a
• /
• pp.125-127
• /
• 2021

본 논문에서는 목조 문화재의 변위 현상 중 하나인 크랙 현상을 감지할 수 있는 EfficientNet 기반 모델을 제안한다. 우선 사전 학습된 EfficientNet모델을 통해 학습 이미지로부터 심층 특징을 추출하고 크랙이 존재하는지 아닌지에 대해 분류하기 위한 완전 연결 신경망을 학습한다. 그런 다음 새로운 목조 문화재 이미지가 들어왔을 때 학습한 모델을 통해서 크랙이 존재하는지에 대해 최종적으로 판별하게 된다. 데이터 셋으로는 충주시 근처의 문화재에 방문해서 수집한 목조 이미지를 가지고 정상 및 비정상으로 구분한 데이터 셋을 사용하였다. 실험 결과 EfficientNet을 사용한 딥 러닝 기반 모델이 다른 사전 학습된 합성 곱 신경망 모델보다 더 좋은 성능을 나타냄을 확인하였다. 이러한 결과로부터 우리가 제안한 방법이 목재 문화재에서의 크랙 검출에 있어서 적합함을 보여준다.

• Annual Conference of KIPS
• /
• 2024.05a
• /
• pp.576-579
• /
• 2024

경제 위기 대비를 위해 인공지능을 활용한 주식시장 변동성 이상을 탐지하는 목적을 가지고 있다. 글로벌 이슈와 경제 위기 대비를 위해 주식시장 변동성 예측의 중요성이 부각되고 있으며, 기존의 주식시장 변동성 지수인 VIX 의 한계로 인해 더 복잡한 모델 및 인공지능을 활용한 연구에 관심이 집중되고 있다. 기존의 주식시장 변동성 예측에 관한 연구들은 통계적인 방법을 사용했으며 인공지능을 이용한 연구 또한 대부분 이상치 구간을 표시하여 예측을 목표로 하고 있으나 이러한 접근법은 라벨이 있는 데이터 수집 어려움, 클래스 불균형 문제가 있다. 본 연구는 인공지능을 활용한 주식시장 변동성 탐지에 기여하고 지도 학습 방식 대신 비지도 학습 기반의 이상탐지모델을 사용하여 주식시장 변동성을 예측하는 새로운 방법론을 제안한다. 본 연구에서 개발한 인공지능 모델은 IsolationForest 모델을 활용하며, 시계열 데이터를 전처리한 후 정상성을 확보하는 등의 과정을 거친다. 실험 결과로 인공지능 모델이 주요 경제이슈를 이상치로 검출하는 성능을 확인하였으며 재현율 약 93.6%, 정밀도 100%로 높은 성능을 달성했다.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.50 no.4
• /
• pp.171-181
• /
• 2013

Recently, ground penetrating radar(GPR) has been widely used in detecting metallic and nonmetallic buried landmines and a number of related researches have been reported. A novel preprocessing method is proposed in this paper to flag potential locations of buried mine-like objects from GPR array measurements. GPR operates by measuring the reflection of an electromagnetic pulse from discontinuities in subsurface dielectric properties. As the GPR pulse propagates in the geologic medium, it suffers nonlinear attenuation as the result of absorption and dispersion, besides spherical divergence. In the proposed algorithm, a logarithmic transformed regression model which successfully represents the time-varying signal amplitude of the GPR data is estimated at first. Then, background signals may be densely distributed near the regression model and candidate signals of targets may be far away from the regression model in the time-amplitude space. Based on the observation, GPR signals are decomposed into candidate signals of targets and background signals using residuals computed from the estimated value by regression and the measurement of GPR. Candidate signals which may contain target signals and noise signals need to be refined. Finally, targets are detected through the refinement of candidate signals based on geometric signatures of mine-like objects. Our algorithm is evaluated using real GPR data obtained from indoor controlled environment and the experimental results demonstrate remarkable performance of our mine-like object detection method.

• Journal of the Korea Computer Graphics Society
• /
• v.23 no.1
• /
• pp.49-55
• /
• 2017

Computed tomographic angiography (CTA) is widely used in the diagnosis and treatment of coronary artery disease because it shows not only the whole anatomical structure of the cardiovascular three-dimensionally but also provides information on the lesion and type of plaque. However, due to the large size of the image, there is a limitation in manually extracting coronary arteries, and related researches are performed to automatically extract coronary arteries accurately. As the coronary artery originate from the ascending aorta, the ascending aorta and ostium should be detected to extract the coronary tree accurately. In this paper, we propose an automatic segmentation for the ostium as a starting structure of coronary artery in CTA. First, the region of the ascending aorta is initially detected by using Hough circle transform based on the relative position and size of the ascending aorta. Second, the volume of interest is defined to reduce the search range based on the initial area. Third, the refined ascending aorta is segmented by using a two-dimensional geodesic active contour. Finally, the two ostia are detected within the region of the refined ascending aorta. For the evaluation of our method, we measured the Euclidean distance between the result and the ground truths annotated manually by medical experts in 20 CTA images. The experimental results showed that the ostia were accurately detected.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.39 no.3
• /
• pp.141-148
• /
• 2021

This paper presents a method to detect zebra-crossing using deep learning which combines SegNet and ResNet. For the blind, a safe crossing system is important to know exactly where the zebra-crossings are. Zebra-crossing detection by deep learning can be a good solution to this problem and robotic vision-based assistive technologies sprung up over the past few years, which focused on specific scene objects using monocular detectors. These traditional methods have achieved significant results with relatively long processing times, and enhanced the zebra-crossing perception to a large extent. However, running all detectors jointly incurs a long latency and becomes computationally prohibitive on wearable embedded systems. In this paper, we propose a model for fast and stable segmentation of zebra-crossing from captured images. The model is improved based on a combination of SegNet and ResNet and consists of three steps. First, the input image is subsampled to extract image features and the convolutional neural network of ResNet is modified to make it the new encoder. Second, through the SegNet original up-sampling network, the abstract features are restored to the original image size. Finally, the method classifies all pixels and calculates the accuracy of each pixel. The experimental results prove the efficiency of the modified semantic segmentation algorithm with a relatively high computing speed.

• Proceedings of the Korean Information Science Society Conference
• /
• 2005.11a
• /
• pp.130-132
• /
• 2005

초고속 네트워크의 폭발적인 확산과 함께 네트워크 침입 사례 또한 증가하고 있다. 이를 검출하기 위한 방안으로 침입 탐지 시스템에 대한 관심과 연구 또한 증가하고 있다. 네트워크 침입을 탐지위한 방법으로 기존의 알려진 공격을 찾는 오용 탐지와 비정상적인 행위를 탐지하는 방법이 존재한다. 본 논문에서는 이를 혼합한 하이브리드 형태의 새로운 침입 탐지 시스템을 제안한다. 기존의 혼합된 방식과는 다르게 네트워크 데이터의 모델링과 탐지를 위해 가우시안 혼합 모델을 사용한다. 가우시안 혼합 모델에 기반한 침입 탐지 시스템의 성능을 평가하기 위해 DARPA'99 데이터에 적용하여 실험하였다. 실험 결과 정상과 공격은 확연히 구분되는 결과를 나타내었으며, 공격 간의 분류도 상당 수 가능하였다.