119 research outputs found
A High Quality Depth Map Upsampling Method Robust to Misalignment of Depth and Color Boundaries
No full textIn recent years, fusion camera systems that consist of color cameras and Time-of-Flight (TOF) depth sensors have been popularly used due to its depth sensing capability at real-time frame rates. However, captured depth maps are limited in low resolution compared to the corresponding color images due to physical limitation of the TOF depth sensor. Most approaches to enhancing the resolution of captured depth maps depend on the implicit assumption that when neighboring pixels in the color image have similar values, they are also similar in depth. Although many algorithms have been proposed, they still yield erroneous results, especially when region boundaries in the depth map and the color image are not aligned. We therefore propose a novel kernel regression framework to generate the high quality depth map. Our proposed filter is based on the vector pointing similar pixels that represents the unit vector toward similar neighbors in the local region. The vectors are used to detect misaligned regions between color edges and depth edges. Unlike conventional kernel regression methods, our method properly handles misaligned regions by introducing the numerical analysis of the local structure into the kernel regression framework. Experimental comparisons with other data fusion techniques prove the superiority of the proposed algorithm
Effect of epichlorohydrin treatment on the coating process and performance of high-barrier paper packaging
No full textThe fabrication of coated papers using hydrophilic and biodegradable polymers is important for developing sustainable packaging materials with high barrier and superior mechanical properties. However, water, which is used as the solvent in the paper coating process using hydrophilic polymers, deforms the shape of the paper and deteriorates performance. Therefore, we propose a new coating process that treats Kraft paper (KP) with epichlorohydrin (ECH) as a binder before the coating process. Crosslinked polyvinyl alcohol is coated on the ECH-treated KP using a solution casting method. ECH maintains the shape of the paper and improves coating uniformity; significantly enhances interfacial interactions, which increases barrier properties and sealing strength; and extends the shelf life of biscuits by reducing oxygen and moisture permeability. An ecotoxicity test using Lolium multiflorum demonstrates an insignificant phytotoxicity level for the as-prepared coated papers. Thus, ECH-treated KP is a potential candidate for high -barrier food packaging.
Poly(p-PhenyleneEthynylene) 필름의 기공 조절을 통한 고감도 화학센서 제조에 관한 연구
Get PDF학위논문 (석사)-- 서울대학교 대학원 : 공과대학 화학생물공학부, 2019. 2. 차국헌.In this paper, we synthesized three poly(p-phenyleneethynylene) (PPE) polymers with different side chains. We figured out the influence of side chain polarity in co-monomer on film formation process. Three different fluorescent polymer films with controlled surface morphology have been fabricated for the application to chemosensor. Since the diffusion of 2,4,6-trinitrotoluene (TNT) is limited at the thick polymer film, we have introduced porous structures in PPE films to facilitate the diffusion in a tens to hundreds of nanometer-thick PPE film. To achieve porous structures in PPE films, non-solvent vapor induced phase separation (NVIPS) was adopted during spin-coating process. Introducing hydrophilic group into the side chain to control the interaction with non-solvent (i.e., water) is the key for the facile morphology control in this work. We, eventually, could control the morphology of different PPE polymers which generate variant structures on the film surface by changing the relative humidity. As expected, TNT could diffuse into the film enough to show high sensitivity more effectively, compared with the one that has no pore on the surface. We also found out that both porous structure and film thickness could effect on the quenching performance.
In conclusion, controlling the morphology of the film such as introducing the porous structure can be a promising strategy for enhancing the efficiency of chemosensor.본 연구에서는 poly(p-phenyleneethynylene)s (PPEs)의 합성과 필름 형성 과정에서의 다양한 곁 사슬에 따른 극성의 영향에 대해 살펴보았다. 표면 형태 조절을 한 세 가지 다른 형광고분자 필름을 화학센서의 응용을 위해 만들었다.
다공성 필름의 제작을 통해 기공에 따른 2,4,6-trinitrotoluene (TNT) 증기에 대한 소광 실험을 진행하였다. TNT은 두꺼운 고분자 필름에서는 제한되기 때문에 필름에 기공성 구조를 도입하여 수십-수백나노미터의 두께를 가진 PPE 필름에서도 확산이 잘 일어나도록 하였다. 비용매 증기로 유도되는 상분리 (NVIPS)를 이용해 스핀코팅 방법으로 필름 내 기공을 만들었으며 PPE 고분자의 농도를 미세 조절하여 고분자 박막의 기공을 조절하였다.
PPE 고분자의 곁 사슬의 친수성 구조가 기공 형성을 조절하는 비용매(물)에 주요한 요인임을 확인하였고, 접촉각 측정을 통한 계면에너지 계산으로 수치화하여 최적의 구조를 찾아내었다. 이를 통해 상대습도 조절을 통한 필름의 다양한 표면 구조를 조절하였다. 이를 통해 기공이 없는 필름 표면보다 기공을 도입한 필름 표면에서 TNT가 잘 확산하는 것을 형광 소광 실험을 통해 확인하였고, 본 연구를 토대로 기공 구조 도입과 같은 필름의 형태조절이 화학센서의 효율을 증대시키는데 효과적인 전략이 될 것으로 기대된다.Chapter 1. Introduction 1
Chapter 2. Experimental 6
2.1. Materials and Measurement 6
2.2. Monomer Synthesis 6
2.3. Sonogashira-Hagihara Cross-Coupling Polymerization 7
2.4. Synthesis of Polymer P3 8
2.5. Film Sample Preparation 10
2.6. Film Sample Characterization 10
2.7. Fluorescence Analyses 11
Chapter 3. Results and Discussion 12
3.1. Polymer Synthesis and Characterization 12
3.2. Interfacial Energy Investigation 20
3.3. Porous PPE Film Preparation through NVIPS 23
3.4. Fluorescence Quenching Studies 33
Chapter 4. Conclusions 45
References 47
Abstract 49Maste
특이 스펙트럼 템플릿 비교를 통한 전류 잔차를 이용한 산업용 로봇 기어박스 고장감지
Get PDF학위논문(석사) -- 서울대학교대학원 : 공과대학 기계공학부, 2022.2. 윤병동.Industrial robots are essential equipment for process automation in a wide range of industrial fields. In manufacturing fields, unexpected faults of robots can severely damage the economy of a company. Fault can occur in various components of the robot and a faulty gearbox can have a significant effect on the robot's driving performance and manufactured product. Therefore, in this paper, gearbox fault detection of an industrial robot is performed using current signals applied to the actuating motor. The proposed method synchronizes normal current signal data to reference phase by resampling through Hilbert Transform. The synchronized signals are then split by singular value decomposition, and the principal components are extracted and averaged to establish normal template. Residual signal is then extracted by subtracting normal template from synchronized unknown signal. Finally, health management feature is calculated from the residual signal to perform fault detection. To quantify the performance of the proposed method, an evaluation metric detection error is derived. The results of detection error show that the uncertainty of fault detection is declined through the proposed method. The distribution of health feature using proposed method is more concentrated than that of health feature using time synchronous averaging without the normal template.산업용 로봇은 넓은 범위의 산업 분야에서 필수적인 장비이다. 제조산업 분야에서, 예상못한 로봇의 고장은 회사에 심각한 경제적 타격을 초래할 수 있다. 고장은 로봇의 다양한 요소에서 일어날 수 있고, 고장난 기어박스는 로봇의 운행 성능과 제조된 물품의 품질에 중대한 영향을 미칠 수 있다. 그러므로 본 논문은 모터에 인가되는 전류 신호를 이용한 산업용 로봇의 기어박스 고장감지를 수행했다. 제안된 방법은 정상 전류 신호 데이터를 힐버트 변환을 통한 리샘플링을 통해 참조 위상에 동기화한다. 동기화된 신호들은 특이 스펙트럼 분석을 통해 분해되고, 주요 성분들이 추출되어 그 평균값이 정상 템플릿을 구축한다. 이후 잔차 신호가 고장 상태를 모르는 동기화된 신호에서 정상 템플릿을 뺌으로써 추출된다.
마지막으로 고장 감지를 위해 잔차 신호에서 건전성 인자가 계산된다. 제안된 방법의 성능을 정량화하기 위해 감지 오차라는 평가 지표를 도입했다. 이 지표를 통해 제안된 방법을 사용하였을 때 정상 템플릿을 사용하지 않은 시간 동기화 평균을 통해 구한 건전성 인자의 분포보다 밀집된 분포를 보였다. 즉, 감지 오차가 감소하여 고장 감지에서의 불확실성이 감소했다는 것을 확인할 수 있었다.Abstract i
Chapter 1 Introduction 1
1.1 Motivation 1
1.2 Dissertation Layout 3
Chapter 2 Theoretical Backgrounds 4
2.1 Characteristic of Current Signal in Fault Diagnosis of Industrial Robot 4
2.2 Hilbert Transform 5
2.3 Singular Spectrum Analysis 6
Chapter 3 Proposed Method 11
3.1 Signal Synchronizing 12
3.2 Establishing Normal Template & Fault Detection using Residual Signal 13
Chapter 4 Experimental Validation 15
4.1 Data Description 15
4.1.1 Testbed Setup 15
4.1.2 Acquired Data 17
4.2 Result & Discussion 18
Chapter 5 Conclusion & Future Work 22
References 23
국문 초록 25석
Fast first arrival picking algorithm for noisy microseismic data
No full textMost microseismic events occur during hydraulic fracturing. Thus microseismic monitoring, by recording seismic waves from microseismic events, is one of the best methods for locating the positions of hydraulic fractures. However, since microseismic events have very low energy, the data often have a low signal-to-noise ratio (S/N ratio) and it is not easy to pick the first arrival time. In this study, we suggest a new fast picking method optimised for noisy data using cross-correlation and stacking. In this method, a reference trace is selected and the time differences between the first arrivals of the reference trace and those of the other traces are computed by cross-correlation. Then, all traces are aligned with the reference trace by time shifting, and the aligned traces are summed together to produce a stacked reference trace that has a considerably improved S/N ratio. After the first arrival time of the stacked reference trace is picked, the first arrival time of each trace is calculated automatically using the time differences obtained in the cross-correlation process. In experiments with noisy synthetic data and field data, this method produces more reliable results than the traditional method, which picks the first arrival time of each noisy trace separately. In addition, the computation time is dramatically reduced.This work was supported by the New & Renewable Energy Core Technology Program of the Korea Institute of Energy Technology Evaluation and Planning (KETEP) granted financial resource from Ministry of Trade, Industry & Energy, Republic of Korea (No. 20133030000220)
Predicting biogas production from organic waste through anaerobic co-digestion
No full textInternational audienc
Regionalized GHG emission factors and amounts based on local electricity consumption and energy source generation
No full textInternational audienceElectricity is often generated in one region and consumed in another. The regions consuming electricity should bear the responsibility for the greenhouse gas (GHG) emissions associated with their consumption. A consumption-based calculation helps address the imbalance between generation regions (typically where fossil fuel power plants such as coal and LNG are located) and consumption regions. This approach can reduce the issue of disproportionately concentrating GHG emission responsibility on specific regions. In this context, we propose a regionalized approach that allocates GHG emissions based on electricity consumption rather than production. Utilizing data from the Korea Electric Power Corporation, the study estimates GHG emissions by energy source and considers electricity transfers between regions. The study results revealed that the responsibility for emissions from consumption is concentrated in cities that rely on coal-fired power generation. For example, Seoul, which supplied electricity from Chungnam where coal-fired power is the main source, has an emission factor of 633 kgCO2eq/kWh. In contrast, Gyeongbuk, which primarily produces electricity through nuclear power, has an emission factor of only 12 kgCO2eq/kWh. The findings reveal significant disparities in GHG emission factors across each region, highlighting the limitations of the current system and the need for more accurate regional emission assessments
Development of predictive and evaluation models for the retro reflectivity performance of pavement lane markings [version 2; peer review: 2 approved]
Get PDFBackground The functions of pavement-markings are to safely and smoothly guide drivers and vehicles along a road-lane, alert drivers not to leave the lane, and provide traffic information. These functions are directly related to the safety and economy of passengers and vehicles on roads. The visibility of pavement-markings is linked to surface damage during the daytime and is significantly affected by luminance at night. Therefore, damage and luminance should be considered key-factors for maintenance and should be measured in detail. In Seoul, mobile equipment such as image-scanners and damage-detectors has been used to measure damage conditions, and a luminance measurement vehicle has been introduced to observe reflected light intensity from road-markings. The mobile measurement methods are respectively utilized due to their objective difference of the measurement. However, measurement discrepancies in temporal, locational, and environmental conditions can lead to considerable errors. Moreover, a standardized maintenance index is currently lacking for planning repair work in Seoul. In that regards, one objective of this research was to develop a retroreflectivity prediction model using a vehicle that detects pavement and marking damage. Method The model enables the prediction of retroreflectivity without additional measurement equipment. Another objective was to develop degradation models based on the age of pavement-markings and to evaluate their performance for maintenance and budget planning. Results As a result, a regression model was determined to be the optimal predictive model for retroreflectivity. A logarithmic model was selected to predict retroreflectivity degradation. Additionally, the Seoul-Road-Marking-Index was developed. Conclusion The developed models are expected to assist in establishing maintenance plans, formulating budgets, and determining repair work, as well as developing relevant regulations and standards in Seoul
Gait characteristics for a lower limb exoskeleton implementing the precedence walking assistance mechanism
No full textIn this paper, we analyse human gait patterns, including knee and hip joint torques and muscle activities, during step initiation phase and continuous walking phase. Additionally, we present a lower limb exoskeleton called the Unmanned Technology Research Centre Exoskeleton (UTRCEXO) implementing a precedence walking assistance mechanism based on the gait characteristics. The operator equipped with the Unmanned Technology Research Centre Exoskeleton (UTRCEXO) walks with a 15 kg load at 3.3 km/h step velocity
The Change of Liquidity Premium Caused by a New Multiplier Regulation in the KOSPI200 Index Options Market
No full text- …
