• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.27 no.2A
• /
• pp.137-145
• /
• 2002

In the method of 3D modeling, stereo matching method which obtains three dimensional depth information from the two images is taken from the different view points. In general, it is very essential work for the 3D modeling from 2D stereo images to estimate the exact disparity through fading the conjugate pair of pixel from the left and right image. In this paper to solve the problems of the stereo image disparity estimation, we introduce a novel approach method to improve the exactness and efficiency of the disparity. In the first place, we perform a wavelet transformation of the stereo images and set the reference points in the image by the feature-based matching method. This reference points have very high probability over 95 %. In the base of these reference points we can decide the size of the variable block searching windows for estimating dense disparity of area based method and perform the ordering constraint to prevent mismatching. By doing this, we could estimate the disparity in a short time and solve the occlusion caused by applying the fried-sized windows and probable error caused by repeating patterns.

• Journal of the Korea Convergence Society
• /
• v.9 no.11
• /
• pp.137-143
• /
• 2018

DICOM images of patellofemoral bones were converted into a stereolithography file, and a Unigraphics CAD program was used to create a CAD modeling in which there exists point, line and facet information. The modeling extraction of joint facets was performed by linking two adjacent points into lines in the stereolithography file by using the Unigraphics rapid spacing function and then linking the lines into facets to complete the entire modeling. This modeling extraction was performed based on the anatomical knowledge of joint facet directions. As a result, a personalized space modeling and solid modeling were produced for the joint facets of patellofemoral bones. This was followed by a CAM control computing operation of solid modeling on graphite materials and 5-axis machining of patellofemoral bones. That is the description of a method for a personalized implant modeling by using DICOM images of patellofemoral bones.

• Journal of the Korean Physical Society
• /
• v.73 no.11
• /
• pp.1764-1773
• /
• 2018

A-PET is a quad-head PET scanner developed for use in small-animal imaging. The dimensions of its volumetric field of view (FOV) are $46.1{\times}46.1{\times}46.1mm^3$ and the gap between the detector modules has been minimized in order to provide a highly sensitive system. However, such a small FOV together with the quad-head geometry causes image quality degradation. The main factor related to image degradation for the quad-head PET is the mispositioning of events caused by the penetration effect in the detector. In this paper, we propose a precise method for modelling the system at the high spatial resolution of the A-PET using a LOR (line of response) based ML-EM (maximum likelihood expectation maximization) that allows for penetration effects. The proposed system model provides the detection probability of every possible ray-path via crystal sampling methods. For the ray-path sampling, the sub-LORs are defined by connecting the sampling points of the crystal pair. We incorporate the detection probability of each sub-LOR into the model by calculating the penetration effect. For comparison, we used a standard LOR-based model and a Monte Carlo-based modeling approach, and evaluated the reconstructed images using both the National Electrical Manufacturers Association NU 4-2008 standards and the Geant4 Application for Tomographic Emission simulation toolkit (GATE). An average full width at half maximum (FWHM) at different locations of 1.77 mm and 1.79 mm are obtained using the proposed system model and standard LOR system model, which does not include penetration effects, respectively. The standard deviation of the uniform region in the NEMA image quality phantom is 2.14% for the proposed method and 14.3% for the LOR system model, indicating that the proposed model out-performs the standard LOR-based model.

• Journal of the Korean Society of Radiology
• /
• v.16 no.3
• /
• pp.343-349
• /
• 2022

We intend to use topic modeling to identify radiation-themed papers published from 1989 to 2022 and analyze the relevance and weight between topics. This study analyzed topics derived from national subjects for 717 papers published until recently in 2022 to contribute to the revitalization of research in the field of radiation. Through text mining, overall research trends on the subject distribution of the study were analyzed, and five topics were derived through topic modeling. First, among the papers to be analyzed, a total of 1,675 words were frequency-analyzed through the preprocessing process of key words in a total of 717 papers centered on keywords. Second, as a result of analyzing topics based on the association of constituent words for five topics, it was found that studies focused on minimizing dose in the range that does not degrade image quality in the fields of radiation, image, CT clinical. In addition, it was found that various studies were mainly conducted in the MRI, and the study of ultrasound in various areas of disease analysis was actively attempted.

• Journal of KIISE
• /
• v.42 no.4
• /
• pp.451-458
• /
• 2015

Previous multimodal learning methods focus on problem-solving aspects, such as image and video search and tagging, rather than on knowledge acquisition via content modeling. In this paper, we propose the Multimodal Concept Hierarchy (MuCH), which is a content modeling method that uses a cartoon video dataset and a character-based subtitle generation method from the learned model. The MuCH model has a multimodal hypernetwork layer, in which the patterns of the words and image patches are represented, and a concept layer, in which each concept variable is represented by a probability distribution of the words and the image patches. The model can learn the characteristics of the characters as concepts from the video subtitles and scene images by using a Bayesian learning method and can also generate character-based subtitles from the learned model if text queries are provided. As an experiment, the MuCH model learned concepts from 'Pororo' cartoon videos with a total of 268 minutes in length and generated character-based subtitles. Finally, we compare the results with those of other multimodal learning models. The Experimental results indicate that given the same text query, our model generates more accurate and more character-specific subtitles than other models.

• International Journal of Control, Automation, and Systems
• /
• v.1 no.1
• /
• pp.111-118
• /
• 2003

This paper proposes a two dimensional behavior analysis method for fish in a water tank based on the ARX method and the Kalman filter algorithm using image processing technology. In modeling the behavior of fish, the input is denoted as the environmental change and uses M-sequence. The output is expressed by the partnership between fish. The behavior model of individual fish is identified by the ARX method. It is then estimated by the Kalman filter algorithm. Finally, the fish behavior is analyzed by FFT. To prove the effectiveness of the pro-posed algorithm, it is applied to two tilapias in a water tank with dimensions of 100cm$\times$100cm$\times$50cm. The effectiveness of the proposed method is demonstrated through ARX identification, estimation of Kalman filter, and FFT analysis.

• Proceedings of the KIEE Conference
• /
• 2002.11c
• /
• pp.485-488
• /
• 2002

In this paper, fuzzy logic controllers are designed for compensation of distance errors. Because we can't know information of the depth in a mono camera, these errors are occurred. Also, they are increased as a target object is to keep away from a center of image. Therefore, the errors for each position of joints of an arm robot should be modeled, but accurate models can't be obtained because of no information of the depth, uncertain feature points of image, parameter uncertainties, and illumination. Hence, fuzzy logic controllers for each error are designed for compensation. This paper consists of color image processing, error modeling, and the controller design. Experimental results are given to verify the effectiveness of our proposed method.

• Proceedings of the Korea Society for Industrial Systems Conference
• /
• 2002.06a
• /
• pp.316-322
• /
• 2002

본 연구에서는 고정된 카메라로부터 입력되는 영상열에서 이동 물체를 신뢰성있게 분리하기 위해 형태 정보를 이용한 확장 방법을 제안한다. 영역 분리의 핵심은 배경으로부터 주위 잡음 영역과 무관하게 이동 물체 영역을 분리하는 기술이라고 볼 수 있다. 제안된 방법은 초기 이동 물체가 존재하지 않는 영상을 참고 영상(reference image)으로 하여 입력 영상(input image)과의 차영상(subtraction image)을 구하고, 차영상의 히스토그램(histogram)에서 배경잡음 모델링(modeling)을 통해 배경잡음을 제거한다. 그리고 배경잡음이 제거된 차영상에서 국부 최대값들(local maxima)을 이용해 후보 초기 영역을 선정한 후, 이 영역을 기반으로 영역의 형태정보를 이용하여 영역을 선별적으로 확장하면서 결합하는 방법을 사용하였다. 제안된 방법을 실제 상황에서 얻은 다양한 영상열에 적용한 결과, 기존의 영역 분리 방법보다 주위 잡음과 무관하게 이동 물체를 분리할 수 있음을 확인할 수 있었다.

• Proceedings of the IEEK Conference
• /
• 2000.07a
• /
• pp.58-61
• /
• 2000

In this paper, we describe an effective method to enhance the color night images with spatio-temporal multi-scale retinex focused to the Intelligent Transportation System (ITS) applications such as in the single CCD based Electronic Toll Collection System (ETCS). The basic spatial retinex is known to provide color constancy while effectively removing local shades. However, it is relatively ineffective in night vision enhancement. Our proposed method, STMSR, exploits the iterative time averaging of image sequences to suppress the noise in consideration of the moving vehicles in image frame. In the STMSR method, the spatial term makes the dark images distinguishable and preserves the color information day and night while the temporal term reduces the noise effect for sharper and clearer reconstruction of the contents in each image frame. We show through representative simulations that incorporating both terms in the modeling produces the output sequential images visually more pleasing than the original dim images.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2001.10a
• /
• pp.405-408
• /
• 2001

지리 정보와 속성 정보를 바탕으로 사용자에게 다양한 분석 기능을 제공하는 인터넷 지리정보시스템은 지리정보서비스 사용자 층을 더욱 확대 시키고 있다. 그러나 이들 시스템은 개방성, 상호운용성을 지원하지 않으며 기능 및 데이터 지원 확장, 다양한 데이터 소스로의 접근 등의 어려움을 가지고 있다. 본 논문에서는, OGC(Open GIS Consortium)의 웹 지리정보서비스 아키텍쳐에 기반을 둔 상호운용성과 XML 기반 데이터서비스 등 개방형 웹 지리정보서비스 시스템을 설계, 구현한다. 또한, 컴포넌트 기반의 아키텍쳐를 바탕으로 단위 기능을 컴포넌트 형태로 제공함으로써 구축 비용 및 다양한 사양한 요구사항을 효율적으로 충족시킬 수 있다. 개발된 서비스 컴포넌트는 객체 지향형 모델링 툴인 UML(Unified Modeling Language)로 설계되었고, ATL/COM 과 C++ 환경에서 구축되었다. 이는 재사용성 뿐만 아니라 상호 운용성, 언어 독립성, 쉬운 개발환경 등을 제공한다.