• 한국멀티미디어학회논문지
• /
• 제18권4호
• /
• pp.437-442
• /
• 2015

Voice activity detection is very important process that voice activity separated form noisy speech signal for speech enhance. Over the past few years, many studies have been made on voice activity detection, but it has poor performance in low signal to noise ratio environment or fickle noise such as car noise. In this paper, it proposed new voice activity detection algorithm using ensemble variance based on wavelet band entropy and soft thresholding method. We conduct a survey in a lot of signal to noise ratio environment of car noise to evaluate performance of the proposed algorithm and confirmed performance of the proposed algorithm.

• Genomics & Informatics
• /
• 제18권2호
• /
• pp.19.1-19.6
• /
• 2020

In semantic annotation, semantic concepts are linked to natural language. Semantic annotation helps in boosting the ability to search and access resources and can be used in information retrieval systems to augment the queries from the user. In the research described in this paper, we aimed to identify ontological concepts in scientific text contained in spreadsheets. We developed a tool that can handle various types of spreadsheets. Furthermore, we used the NCBO Annotator API provided by BioPortal to enhance the semantic annotation functionality to cover spreadsheet data. Table2Annotation has strengths in certain criteria such as speed, error handling, and complex concept matching.

• 한국정보통신학회논문지
• /
• 제9권4호
• /
• pp.890-895
• /
• 2005

동적(온라인) 서명인증시스템은 내부 처리 과정에서는 불필요한 점들을 제거하는 전처리과정, 서명의 변화폭을 줄여주고 서명자의 고유한 특징 정보를 추출하는 특징추출과정, 두 서명의 특징벡터를 비교하여 유사도를 계산하는 비교과정, 보안수준에 따른 인증 여부를 결정하는 판단과정으로 구성되며, 사용자 관점에서의 화면 구성은 서명을 입력받아 기준서명과 보안수준 값을 만들어 주는 등록화면과 권한 부여를 위하여 진서명인지 모조서명인지를 판단하는 인증화면으로 나누어진다. 본 논문에서는 동적 서명인증시스템의 처리 속도, 서명의 특징벡터의 추출방법과 비교 알고리즘, 사용자 인터페이스 등과 실제 환경에서의 설계 및 구현에 대한 연구이다.

• 한국소성가공학회:학술대회논문집
• /
• 한국소성가공학회 1997년도 추계학술대회논문집
• /
• pp.17-23
• /
• 1997

In the two-dimensional Finite Element Method for forming simulation, mesh generation and remeshing process are very significant. In this paper, using the modified splitting mesh generation algorithm, we can overcome the limitation of existing techniques and acquire mesh, which has optimal mesh density. A modified splitting algorithm for automatically generating quadrilateral mesh within a complex domain is described. Unnecessary meshing process for density representation is removed. Especially, during the mesh generation with high gradient density like as shear band representation, the modified mesh density scheme, which will generate quadrilateral mesh with the minimized error, which takes effect on FEM solver, is introduced.

• 한국지능정보시스템학회:학술대회논문집
• /
• 한국지능정보시스템학회 2004년도 추계학술대회
• /
• pp.158-161
• /
• 2004

In this paper, we propose a level of contexts such as low level and high level contexts, and its criteria of categorization are the existence of interaction and composite process done by the inference mechanism. When there is no matching high level service for the associated high level context, this context is described as meaningless. If the services cannot be provided with the entities by the system, this situational information is of little consequence. To provide services with the entities, we propose "community computing" architectural concept which provide the high level service to a group of agents in a community, and can be managed by the service scenario

• 정보학연구
• /
• 제10권4호
• /
• pp.27-39
• /
• 2007

Object tracking in a real time environment is one of challenging subjects in computer vision area during past couple of years. This paper proposes a method of object detection and tracking using adaptive background estimation in real time environment. To obtain a stable and adaptive background, we combine 3-frame differential method and running average single gaussian background model. Using this background model, we can successfully detect moving objects while minimizing false moving objects caused by noise. In the tracking phase, we propose a matching criteria where the weight of position and inner brightness distribution can be controlled by the size of objects. Also, we adopt a Kalman Filter to overcome the occlusion of tracked objects. By experiments, we can successfully detect and track objects in real time environment.

• 대한원격탐사학회:학술대회논문집
• /
• 대한원격탐사학회 2003년도 Proceedings of ACRS 2003 ISRS
• /
• pp.1115-1117
• /
• 2003

The information about transactions of real estate has tended to be not open. Therefore, it has been difficult for individuals to judge the proper price of each real estate. In the course of time several studies have been conducted on proposing criterions for judging the proper price of real estates. As to office buildings and apartments, it is proved techniques required for making criterions have been achieved to a certain extent. Therefore, this research aims to make methods that propose to consumers reliable criteria for judging the proper price of detached houses. The methods are based on hedonic price method and micro-level spatial elements peculiar to detached houses are considered.

• International journal of advanced smart convergence
• /
• 제8권2호
• /
• pp.227-236
• /
• 2019

In this study, we develop a Korotkoff sound based automatic blood pressure measurement device including sensor, hardware, and analysis algorithm. PVDF-based sensor pattern was developed to function as a vibration sensor to detect of Korotkoff sounds, and the film's output was connected to an impedance-matching circuit. An algorithm for determining starting and ending points of the Korotkoff sounds was established, and clinical data from subjects were acquired and analyzed to find the relationship between the values obtained by the auscultatory method and from the developed device. The results from 86 out of 90 systolic measurements and 84 out of 90 diastolic measurements indicate that the developed device pass the validation criteria of the international protocol. Correlation coefficients for the values obtained by the auscultatory method and from the developed device were 0.982 and 0.980 for systolic and diastolic blood pressure, respectively. Blood pressure measurements based on Korotkoff sound signals obtained by using the developed PVDF film-based sensor module are accurate and highly correlated with measurements obtained by the traditional auscultatory method.

• 전자공학회논문지
• /
• 제50권9호
• /
• pp.119-126
• /
• 2013

본 논문은 영상 획득 장치에서 발생하는 색수차를 영상의 국부적 특성을 고려하여 보정하는 알고리즘을 제안한다. 색수차는 파장에 따른 렌즈의 회절률 차이로 인해 발생하며, 그 결과 영상의 강한 에지 영역에서 RGB 채널이 어긋나 색 결함이 생기게 된다. 제안하는 알고리즘은 이러한 색수차가 발생하는 특성을 이용하여 먼저 영상의 에지 영역과 그 주변부를 색수차가 발생되는 영역으로 검출한다. 이 후, 검출된 에지 영역에서 RGB 채널의 에지를 일치시키는 방법을 이용하여 색수차를 제거한다. 그 결과 기존의 색수차 제거 방법으로 널리 사용되는 전역적인 영상 와핑 방법으로는 제거하지 못하는 축상 색수차와 이미징 센서로부터 발생하는 purple fringing 결함까지도 저감시킬 수 있다. 실험 결과를 통해 제안하는 방법이 기존 방법에 비해 수치적인 면과 시각적인 면에서 뛰어난 결과를 보임을 확인할 수 있다.

• 정보처리학회논문지:소프트웨어 및 데이터공학
• /
• 제5권7호
• /
• pp.327-332
• /
• 2016

본 연구는 비 정형 농산물 중 6년근 수삼의 자동 등급 분류하기 위한 선행연구로, 이를 위해 4방향에서 이미지 취득이 가능한 수삼 영상 측정기를 제작 하였으며 총 245 수삼 개체에 대해서 영상을 취득하였다. 취득된 영상의 각 수삼 개체마다 12개의 파라미터를 추출하였으며, KGC 인삼공사의 수삼등급 분류 기준과 각 등급별 평균 파라미터의 분포를 조사하여 최종 4개 파라미터를 선정하였다. 패턴인식 분류기는 Support Vector Machine을 사용하였으며 공용 소프트웨어인 OpenCV Library를 사용하여 k-Class 분류기를 설계하였다. 각 등급별 학습 데이터 수를 10, 15, 20으로 조정하여 등급별 인식률, 본인 거부율, 타인 인식율을 조사하였으며, 학습데이터 수가 10개일 때 1등급 인식률 94%, 2등급 인식률 98%, 3등급 인식률 90%로 가장 높은 인식 성능을 보였다.