• 방송공학회논문지
• /
• 제9권2호
• /
• pp.110-118
• /
• 2004

This paper shows that blind deconvolution has only finite solutions when an original image and a point spread function are nonzero over a restricted domain, in other words, an observed image has a compact support. The key of the proof is to use z-transformations and factorizations of polynomials. Then, we propose an algorithm to find all finite solutions under the boundary condition. Finally, we confirm that we can extract all sets of an original image and a point spread function from a degraded image by using our algorithm in numerical examples.

• International Journal of Internet, Broadcasting and Communication
• /
• 제10권4호
• /
• pp.43-49
• /
• 2018

Modern people communicate in a digital space without being limited by physical space and time. On the internet, people's desire for self-expression is getting bigger, and the most common self-expression tool is characters. Characters are visual materials to express human thoughts and desires. This study is a method to positively express the viewer's self in an interactive Hangeul system. The proposed system produces new formative shapes and meanings of Hangeul through the process of deconstructing and reconstructing the characters. Immediate character transformations that are synchronized with the motion of the viewer act as a tool of self-expression by inducing the viewer's active interaction repeatedly. Therefore, the proposed interactive Hangeul is more than enough to be used as a new graphic tool beyond the text to be read, and it is highly likely to develop into cultural contents using it.

• 소음진동
• /
• 제9권1호
• /
• pp.103-112
• /
• 1999

A new modeling and analysis technique for one-dimensional distributed parameter systems is presented. First. discretized equations of motion in Laplace domain are derived by applying discretization methods for partial differential equations of a one-dimensional structure with respect to spatial coordinate. Secondly. the z and inverse z transformations are applied to the discretized equations of motion for obtaining a dynamic matrix for a uniform element. Four different discretization methods are tested with an example. Finally, taking infinite on the number of step for a uniform element leads to an exact dynamic matrix for the uniform element. A generalized modal analysis procedure for eigenvalue analysis and modal expansion is also presented. The resulting element dynamic matrix is tested with a numerical example. Another application example is provided to demonstrate the applicability of the proposed method.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2000년도 하계학술대회 논문집 D
• /
• pp.2678-2680
• /
• 2000

In this paper, we calculates mathematical model of ex-core neutron flux monitoring system for nuclear reactor and design the digital system using the bilinear transformations. The output pulse shape and resolving time of the system determines from the proposed method.

• 의학교육논단
• /
• 제16권3호
• /
• pp.141-146
• /
• 2014

This study focused on investigating the future medicine and health care industry paradigm shift and suggesting the right direction for the medical education system in order for students to be better prepared in the near future. Here, I will discuss four issues related to the future of medicine: Health 2.0, digital health, personalized medicine, and innovations of the public healthcare system. Every issue has lessons for medical education and teaching for students who major in health professions. However, it is obvious that the present is an important period of time as, currently, we are at an early stage in the future health care environment. Recently, there have been rapid transformations in various fields of medicine. Therefore, if we fail to lead medical education in the right direction, medical students will suffer from major problems in coping with these changes.

• 전자공학회논문지B
• /
• 제32B권9호
• /
• pp.1239-1246
• /
• 1995

A forced commutated cycloconverter (FCC) is a direct ac-ac converter capable of providing simultaneous voltage and frequency transformations. In this paper, vector control of an induction motor controlling stator current with forced commutated cycloconverter is presented. The advantage of current control is that the stator dynamics are eliminated and high performance vector control can be achieved. A novel modulation method based on dq transformation techniques is presented. Proposed modulation strategy generates the low frequency modulation function by the instantaneous value of the desired output voltages not by the steady state values of output magnitude and output frequency. PI control and predictive control algorithm for current control are applied, and the validity of proposed method is confirmed through digital simulations. Simulation results of step response and torque distubance and current control are presented.

• 한국측량학회지
• /
• 제17권3호
• /
• pp.265-272
• /
• 1999

현재 우리나라의 수치지도에서는 동경원점계에 의한 Bessel타원체 좌표계를 사용하고 있으므로 새로운 지구중심좌표계(ITRF)에 의한 GRS80타원체 좌표계와는 상당한 차이가 발생하게 된다. 본 연구에서는 새로운 좌표계의 도입에 따른 지도좌표계의 전환에 대비할 수 있도록 하기 위하여 두 체계간의 좌표차와 도엽체계에 대한 특성과 경향을 파악하고, 두 좌표계간의 지도좌표계 전환모듈을 제시하였다.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제14권12호
• /
• pp.4835-4848
• /
• 2020

The recent expansion of the digital content market is increasing the technical demand for various facial image transformations within the virtual environment. The recent image translation technology enables changes between various domains. However, current image-to-image translation techniques do not provide stable performance through unsupervised learning, especially for shape learning in the face transition field. This is because the face is a highly sensitive feature, and the quality of the resulting image is significantly affected, especially if the transitions in the eyes, nose, and mouth are not effectively performed. We herein propose a new unsupervised method that can transform an in-wild face image into another face style through radical transformation. Specifically, the proposed method applies two face-specific feature loss functions for a generative adversarial network. The proposed technique shows that stable domain conversion to other domains is possible while maintaining the image characteristics in the eyes, nose, and mouth.

• 한국멀티미디어학회논문지
• /
• 제25권9호
• /
• pp.1328-1337
• /
• 2022

Recently, research on various font designs has been actively conducted to deliver effective emotional information in a digital environment. In this study, we propose a Hangul font outline vector modification algorithm that effectively conveys sensitivity according to weather information and can be transformed immediately. The algorithm performs a series of transformations: sets outlines according to design pattern templates, calculates the glyph's position to reflect physical rules, splits outline segments into smaller sizes and deforms the outlines. Through this, we could create several vector font designs such as humidity, cloud, wind, and snow. The usability evaluation was close to good, so it can be used in diverse ways if we improve readability and effective design expression.

• 로봇학회논문지
• /
• 제17권4호
• /
• pp.425-430
• /
• 2022

As robots are considered one of the mainstream digital transformations, robots with machine vision becomes a main area of study providing the ability to check what robots watch and make decisions based on it. However, it is difficult to find a small object in the image mainly due to the flaw of the most of visual recognition networks. Because visual recognition networks are mostly convolution neural network which usually consider local features. So, we make a model considering not only local feature, but also global feature. In this paper, we propose a detection method of a small marker on the object using deep learning and an algorithm that considers global features by combining Transformer's self-attention technique with a convolutional neural network. We suggest a self-attention model with new definition of Query, Key and Value for model to learn global feature and simplified equation by getting rid of position vector and classification token which cause the model to be heavy and slow. Finally, we show that our model achieves higher mAP than state of the art model YOLOr.