• ETRI Journal
• /
• v.35 no.6
• /
• pp.1115-1125
• /
• 2013

This paper presents a scheme for geometric correction of projected content for planar and quadratic projection surfaces. The scheme does not require the projection surface to be perfectly quadratic or planar and is therefore suitable for uneven low-cost commercial and home projection surfaces. An approach based on the recursive subdivision of second-order B$\acute{e}$zier patches is proposed for the estimation of projection distortion owing to surface imperfections. Unlike existing schemes, the proposed scheme is completely automatic, requires no prior knowledge of the projection surface, and uses a single uncalibrated camera without requiring any physical markers on the projection surface. Furthermore, the scheme is scalable for geometric calibration of multi-projector setups. The efficacy of the proposed scheme is demonstrated using simulations and via practical experiments on various surfaces. A relative distortion error metric is also introduced that provides a quantitative measure of the suppression of geometric distortions, which occurs as the result of an imperfect projection surface.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 2004.04a
• /
• pp.334-337
• /
• 2004

차세대 생산 시스템(Next Generation Manufacturing System： NGMS)의 핵심 개념은 분산 생산 시스템과 다품종 소량의 유연 생산 시스템의 지원이다. 이러한 시스템의 구성을 위하여 실시간 데이터에 기반한 예측 모델이 필수적인데, 이러한 예측 기능을 통하여 생산공정의 관리와 운영, 특히 전체 공정관리를 효율적으로 수행할 수 있다. 한편, 공정으로부터 전송된 데이터는 특정한 형태의 지식으로 표현된다. 이러한 지식들은 시스템에 대한 다양한 정보를 가지고 있으므로 정보를 이용하여 시스템 상태를 빠르고 쉽게 진단할 수 있다. 공정 진단은 현재 공정 상태에서 생산되는 제품의 품질을 추정할 수 있는 정보로 활용된다. 본 논문에서는 이러한 개념이 바탕이 되어 공정관리 시스템을 설계하였다. 제안된 시스템의 적용 대상은 반도체 제조 공정의 단위 공정인 에칭 공정이다. 에칭 공정은 공정 중에 연속적인 검사가 수행되지 않고 최종 제품에 대한 검사가 수행되므로 불량 원인을 찾는 것이 쉽지 않다. 따라서 본 논문에서는 공정관리를 위한 의사지원시스템을 통해 공정의 연속적인 간접진단을 수행하고자 하였다. 본 연구에서 사용된 의사지원시스템은 각 공정에서 얻어지는 데이터와 경험적 지식을 토대로 공정시스템의 해석과 진단이 가능한 시스템이다.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2018.07a
• /
• pp.51-52
• /
• 2018

화력발전용 순환유동층 보일러는 환경오염의 주요인인 질소산화물(NOx)과 황산화물(SOx)의 배출량이 적은 친환경 화력발전용 보일러로 화력발전 업계에서 각광받고 있는 추세이다. 그러나 순환유동층 보일러의 연료인 유동매체는 미분탄과 같이 작지만 단단한 고체이므로 유동매체의 타격으로 인해 워터월(waterwall) 튜브의 마모는 물론 누설까지 야기할 수 있다. 순환유동층 보일러 튜브에서 누설된 증기는 보일러 내부에 클링커(Clinker)를 발생시키고 이는 순환유동층 보일러 튜브 표면에 응고되어 열전도율을 감소시킬 뿐만 아니라 보일러 운전정지의 원인이 된다. 따라서 본 논문에서는 음향방출 센서를 이용하여 화력발전용 순환유동층 보일러 튜브의 누설 위치를 추정하는 방법을 제안한다. 제안 방법에서는 매질의 분자단위 이동에 의해 발생되는 탄성파를 감지할 수 있는 음향방출 센서를 이용하고, 보일러 워터월 튜브의 멤브레인 용접부와 비용접부(seamless)의 감쇠율을 고려한 위치별 센서 감도 추정 알고리즘을 통해 워터월 튜브의 위치별 진폭 크기를 히트맵으로 표현할 수 있다.

• Journal of the Semiconductor & Display Technology
• /
• v.15 no.3
• /
• pp.24-29
• /
• 2016

One of the main applications of digital image processing is the estimation of the number of certain types of objects (cells, seeds, peoples etc.) in an image. Difficulties of these counting problems depends on various factors including shape and size variation, degree of object clustering, contrast between object and background, object texture and its variation, and so on. In this paper, a new automatic colony counting algorithm is proposed. We focused on the two applications: counting the bacteria colonies on the agar plate and estimating the number of seeds from images captured by smartphone camera. To overcome the shape and size variations of the colonies, we adopted the distance transformation and peak detection approach. To estimate the reference size of the colony robustly, we also used k-means clustering algorithm. Experimental results show that our method works well in real world applications.

• Journal of Integrative Natural Science
• /
• v.14 no.2
• /
• pp.41-49
• /
• 2021

The causes of climate change are natural and artificial. Natural causes include changes in temperature and sunspot activities caused by changes in solar radiation due to large-scale volcanic activities, while artificial causes include increased greenhouse gas concentrations and land use changes. Studies have shown that excessive carbon use among artificial causes has accelerated global warming. Climate change is rapidly under way because of this. Due to climate change, the frequency and cycle of infectious disease viruses are greater and faster than before. Currently, the world is suffering greatly from coronavirus infection-19 (COVID-19). Korea is no exception. The first confirmed case occurred on January 20, 2020, and the number of infected people has steadily increased due to several waves since then, and many confirmed cases are occurring in 2021. In this study, we conduct a study on climate change before and after COVID-19 using weather data from Korea to determine whether climate change affects infectious disease viruses through logistic regression analysis. Based on this, we want to classify before and after COVID-19 through a logistic regression model to see how much classification rate we have. In addition, we compare monthly classification rates to see if there are seasonal classification differences.

• Journal of the Korean Society of Visualization
• /
• v.20 no.3
• /
• pp.128-135
• /
• 2022

Cavitation occurs inevitably in marine propellers rotating at high speed in the water, which is a major cause of underwater radiated noise. Cavitation-induced noise from propellers rotating at a specific frequency not only reduces the sonar detection capability, but also exposes the ship's location, and it causes very fatal consequences for the survivability of the navy vessels. Therefore cavity inception speed (CIS) is one of the important factors determining the special performance of the ship. In this study, we present a method using computer vision that can detect and quantitatively estimate tip vortex cavitation on a propeller rotating at high speed. Based on the model test results performed in a large cavitation tunnel, the effectiveness of this method was verified.

• ETRI Journal
• /
• v.44 no.6
• /
• pp.1034-1046
• /
• 2022

The learning-based multiview stereo (MVS) methods for three-dimensional (3D) reconstruction generally use 3D volumes for depth inference. The quality of the reconstructed depth maps and the corresponding point clouds is directly influenced by the spatial resolution of the 3D volume. Consequently, these methods produce point clouds with sparse local regions because of the lack of the memory required to encode a high volume of information. Here, we apply the atrous spatial pyramid pooling (ASPP) module in MVS methods to obtain dense feature maps with multiscale, long-range, contextual information using high receptive fields. For a given 3D volume with the same spatial resolution as that in the MVS methods, the dense feature maps from the ASPP module encoded with superior information can produce dense point clouds without a high memory footprint. Furthermore, we propose a 3D loss for training the MVS networks, which improves the predicted depth values by 24.44%. The ASPP module provides state-of-the-art qualitative results by constructing relatively dense point clouds, which improves the DTU MVS dataset benchmarks by 2.25% compared with those achieved in the previous MVS methods.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.9 no.1
• /
• pp.169-189
• /
• 2015

Detection of anomalous events from video streams is a challenging problem in many video surveillance applications. One such application that has received significant attention from the computer vision community is traffic video surveillance. In this paper, a Lossy Count based Sequential Temporal Pattern mining approach (LC-STP) is proposed for detecting spatio-temporal abnormal events (such as a traffic violation at junction) from sequences of video streams. The proposed approach relies mainly on spatial abstractions of each object, mining frequent temporal patterns in a sequence of video frames to form a regular temporal pattern. In order to detect each object in every frame, the input video is first pre-processed by applying Gaussian Mixture Models. After the detection of foreground objects, the tracking is carried out using block motion estimation by the three-step search method. The primitive events of the object are represented by assigning spatial and temporal symbols corresponding to their location and time information. These primitive events are analyzed to form a temporal pattern in a sequence of video frames, representing temporal relation between various object's primitive events. This is repeated for each window of sequences, and the support for temporal sequence is obtained based on LC-STP to discover regular patterns of normal events. Events deviating from these patterns are identified as anomalies. Unlike the traditional frequent item set mining methods, the proposed method generates maximal frequent patterns without candidate generation. Furthermore, experimental results show that the proposed method performs well and can detect video anomalies in real traffic video data.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.29 no.4C
• /
• pp.494-504
• /
• 2004

The researches about gaze detection have been much developed with many applications. Most previous researches only rely on image processing algorithm, so they take much processing time and have many constraints. In our work, we implement it with a computer vision system setting a IR-LED based single camera. To detect the gaze position, we locate facial features, which is effectively performed with IR-LED based camera and SVM(Support Vector Machine). When a user gazes at a position of monitor, we can compute the 3D positions of those features based on 3D rotation and translation estimation and affine transform. Finally, the gaze position by the facial movements is computed from the normal vector of the plane determined by those computed 3D positions of features. In addition, we use a trained neural network to detect the gaze position by eye's movement. As experimental results, we can obtain the facial and eye gaze position on a monitor and the gaze position accuracy between the computed positions and the real ones is about 4.2 cm of RMS error.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.22 no.4
• /
• pp.683-690
• /
• 2018

In this paper, a new signal processing technique is proposed for calibrating gain, phase, delay offsets in array antenna based anti-jamming minimum variance distortionless response (MVDR) global-positioning-system (GPS) receivers. The proposed technique estimates gain, phase and delay offsets across the antennas, and compensates for the offsets based on the estimates. A pilot signal with good correlation characteristics is used for accurate estimation of the gain, phase and delay offsets. Based on the cross-correlation, the delay offset is first estimated and then gain/phase offsets are estimated. For fine delay offset estimation and compensation, an interpolation technique is used, and specifically, the discrete Fourier transform (DFT) is employed for the interpolation technique to reduce the computational complexity. The proposed technique is verified through computer simulation using MATLAB. According to the simulation results, the proposed technique can reduce the gain, phaes and delay offset to 0.01 dB, 0.05 degree, and 0.5 ns, respectively.