• Proceedings of the KIEE Conference
• /
• 2006.10a
• /
• pp.145-146
• /
• 2006

본 논문은 부분방전(PD: Partial Discharge)의 패턴인식 확률 극대화를 목적으로 신경망(NN: Neural Network) 파라미터 중에서 은닉층 뉴런의 수, 모멘텀(momentum)의 Step size와 Decay rate 를 최적화하기 위하여 유전 알고리즘(GA: Genetic Algonthm)을 적응하였다. 실험적 연구의 대상으로서, GIS(Gas Insulated Switchgear)사고의 주요 원인으로 보고되어있는 결함들을 인위적으로 모의한 16개 Test cell을 이용하여 부분방전을 발생시켰다. 부분방전 신호는 본 연구팀이 개발한 센서를 이용하여 검출되어 데이터베이스가 구축되어 그로부터 추출된 학습 데이터들의 학습에 다음과 같은 5가지 신경망 모델이 적응되었다: Multilayer Perception (MLP), Jordan-Elman Network (JEN), Recurrent Network (RN), Self-Organizing Feature Map (SOFM), Time-Lag Recurrent Network (TLRN). 유전 알고리즘 적용 효율성을 분석하기 위하여 동일한 데이터를 이용하여 다음과 같은 두 가지 방법을 적용한 결과를 상호 비교하였다. 우선 상기 선택된 모델만 적용하였고 다근 하나는 상기 모델과 Genetic Algorithm이 동시에 적용되었다. 모든 모델에 대하여 학습오차와 패턴 분류 확률을 비교한 결과, 유전 알고리즘 적응 시 부분방전 패턴인식 확률이 향상되었음이 확인되어 향후 신뢰성 있는 GIS 부분방전 진단기술에 활용될 수 있을 것으로 사료된다.

• Journal of Institute of Control, Robotics and Systems
• /
• v.17 no.5
• /
• pp.490-498
• /
• 2011

In this paper, we present a two-stage vision-based approach to detect multi views of pedestrian in road scene images. The first stage is HG (Hypothesis Generation), in which potential pedestrian are hypothesized. During the hypothesis generation step, we use a vertical, horizontal edge map, and different colors between road background and pedestrian's clothes to determine the leg position of pedestrian, then a novel symmetry peaks processing is performed to define how many pedestrians is covered in one potential candidate region. Finally, the real candidate region where pedestrian exists will be constructed. The second stage is HV (Hypothesis Verification). In this stage, all hypotheses are verified by Support Vector Machine for classification, which is robust for multi views of pedestrian detection and recognition problems.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.19 no.2
• /
• pp.109-115
• /
• 2001

Recently, in accordance with the introduction of Digital Photogrammetry Systems the use of Digital ortho-photo images have increased and progressed in the study which extract the features from digital ortho-photo image semi-automatically or automatically. However, there are a limit. It has proved in many studies that recognition of the attribution or the features from panchromatic aerial photo is restricted. In this study, I compared color aerial images with panchromatic aerial images and analyzed the characteristics of color aerial images and feature entities which can be extracted semi-automatically. I analyzed extracted feature entities are compared with digital map at a scale of 1:5,000 have constructed in National Geography Institute. With this result, I analyzed the capability of feature extraction and proposed a plan for the study in the future.

• The Transactions of the Korea Information Processing Society
• /
• v.7 no.10
• /
• pp.3300-3309
• /
• 2000

본 논문에서는 변전소에서 수용가까지의 전력공급설비를 나타내는 도면인 배전설비도면의 주요 기호인 전주와 전선인식 방법에 과하여 기술한다. 제안하는 방법은 원형성에 근거하여 전주후보를 추출한 후 이들 사이의 연결성에 근거하여 전선을 인식한 다음, 전주후보들 중에서 전주를 확인하는 방법으로 다음과 같이 네 개의 단계로 구성된다. 첫 번째는 히스토그램 분석을 통하여 얻어진 임계값을 사용하여 입력영상에서 배전설비영역을 추출하는 단계이고, 두 번째는 추출된 배전설비영역을 세선화 하는 단계이다. 세 번째는 세선영상의 분기점 근처에 정의된 탐색영역에서 원형성을 측정하여 전주후보를 추출하는 단계이다. 네 번째는 전주후보들이 가지는 분기점들 간의 연결성을 측정하여 전선을 인식하는 단계이다. 전선인식이 완료되면 전주후보들 중에서 전선을 가지는 후보들만을 추출하여 전주를 인식한다. 제안된 방법은 한국전력공사의 배전설비도면들 중에서 무작위로 추출한 표본 약 30매를 대상으로 실험하고 그 결과를 제시한다.

• Transactions of the Korean Society of Automotive Engineers
• /
• v.22 no.7
• /
• pp.107-116
• /
• 2014

This paper presents camera-based road surface marking detection methods suited to sensor fusion-based positioning system that consists of low-cost GPS (Global Positioning System), INS (Inertial Navigation System), EDM (Extended Digital Map), and vision system. The proposed vision system consists of two parts: lane marking detection and RSM (Road Surface Marking) detection. The lane marking detection provides ROIs (Region of Interest) that are highly likely to contain RSM. The RSM detection generates candidates in the regions and classifies their types. The proposed system focuses on detecting RSM without false detections and performing real time operation. In order to ensure real time operation, the gating varies for lane marking detection and changes detection methods according to the FSM (Finite State Machine) about the driving situation. Also, a single template matching is used to extract features for both lane marking detection and RSM detection, and it is efficiently implemented by horizontal integral image. Further, multiple step verification is performed to minimize false detections.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.7 no.6
• /
• pp.1167-1174
• /
• 2006

Text region detection from a natural scene is useful in many applications such as vehicle license plate recognition. Therefore, in this paper, we propose a text region extraction method using pattern histogram of character-edge maps. We create 16 kinds of edge maps from the extracted edges and then, we create the 8 kinds of edge maps which compound 16 kinds of edge maps, and have a character feature. We extract a candidate of text regions using the 8 kinds of character-edge maps. The verification about candidate of text region used pattern histogram of character-edge maps and structural features of text region. Experimental results show that the proposed method extracts a text regions composed of complex background, various font sizes and font colors effectively.

• Proceedings of the Korean Society for Noise and Vibration Engineering Conference
• /
• 2000.11a
• /
• pp.796-801
• /
• 2000

The monitoring and diagnostics of the rotating machinery have been received considerable attention for many years. The objectives are to classify the machinery condition and to find out the cause of abnormal condition. This paper describes a signal classification method for diagnosing the rotating machinery using the artificial neural network and the wavelet transform. In order to extract salient features, the wavelet transform are used from primary noise signals. Since the wavelet transform decomposes raw time-waveform signals into two respective parts in the time space and frequency domain, more and better features can be obtained easier than time-waveform analysis. In the training phase for classification, self-organizing feature map(SOFM) and learning vector quantization(LVQ) are applied, and the accuracies of them are compared with each other. This paper is focused on the development of an advanced signal classifier to automatise the vibration signal pattern recognition. This method is verified by small reciprocating compressors, for refrigerator and normal and abnormal conditions are classified with high flexibility and reliability.

• The Journal of Korea Robotics Society
• /
• v.14 no.4
• /
• pp.357-362
• /
• 2019

This paper proposes a model and train method that can real-time detect objects and distances estimation based on a monocular camera by applying deep learning. It used YOLOv2 model which is applied to autonomous or robot due to the fast image processing speed. We have changed and learned the loss function so that the YOLOv2 model can detect objects and distances at the same time. The YOLOv2 loss function added a term for learning bounding box values x, y, w, h, and distance values z as 클래스ification losses. In addition, the learning was carried out by multiplying the distance term with parameters for the balance of learning. we trained the model location, recognition by camera and distance data measured by lidar so that we enable the model to estimate distance and objects from a monocular camera, even when the vehicle is going up or down hill. To evaluate the performance of object detection and distance estimation, MAP (Mean Average Precision) and Adjust R square were used and performance was compared with previous research papers. In addition, we compared the original YOLOv2 model FPS (Frame Per Second) for speed measurement with FPS of our model.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2012.11a
• /
• pp.474-477
• /
• 2012

현대사회는 시각장애인에 대한 복지가 부족하다. 예를 들어 유도블럭의 홰손, 지폐의 점자처리의 모호함 등 시각장애인을 위해 만들어진 복지조차 사용하기 어려운게 현실이다. 그래서 우리는 무선카메라와 Window Phone을 이용하여 상기 불편함을 해소하기 위하여 이 프로젝트를 시작하였다. Guide Line Detection은 앞을 못 보는 시각장애인에게 무선카메라에 보이는 영상에서 유도블럭을 찾아 시각장애인과의 거리를 음성으로 알려준다. Bill Recognition은 지폐를 인식하여 음성으로 알려준다. 길 안내 기능은 길을 찾아가지 못하는 시각장애인에게 특정 지점마다 길 안내정보를 등록하고, 등록된 정보는 시각장애인이 실시간으로 길 안내를 받을 수 있다. 음성인식은 기기를 사용하기 힘든 시각장애인들에 대한 접근성을 높이기 위해 WinPhone Application이 제공하는 모든 기능을 흔들기와 음성만으로 사용할 수 있도록 제공한다. 무선카메라의 화질과 Window Phone의 GPS 불규칙적인 오차 때문에 많은 시행착오가 있었지만 무선카메라는 웹 캠으로, GPS오차는 BingMap API의 GPS 가상 좌표로 대체하여 프로젝트를 마칠 수 있었다.

• Journal of Korea Multimedia Society
• /
• v.22 no.9
• /
• pp.1000-1010
• /
• 2019

Image semantic segmentation and dehazing are key tasks in the computer vision. In recent years, researches in both tasks have achieved substantial improvements in performance with the development of Convolutional Neural Network (CNN). However, most of the previous works for semantic segmentation assume the images are captured in clear weather and show degraded performance under hazy images with low contrast and faded color. Meanwhile, dehazing aims to recover clear image given observed hazy image, which is an ill-posed problem and can be alleviated with additional information about the image. In this work, we propose a deep multi-task network for simultaneous semantic segmentation and dehazing. The proposed network takes single haze image as input and predicts dense semantic segmentation map and clear image. The visual information getting refined during the dehazing process can help the recognition task of semantic segmentation. On the other hand, semantic features obtained during the semantic segmentation process can provide cues for color priors for objects, which can help dehazing process. Experimental results demonstrate the effectiveness of the proposed multi-task approach, showing improved performance compared to the separate networks.