• 한국멀티미디어학회논문지
• /
• 제23권8호
• /
• pp.940-952
• /
• 2020

The Alzheimer's disease (AD) is a neurodegenerative disease commonly found in the elderly individuals. It is one of the most common forms of dementia; patients with AD suffer from a degradation of cognitive abilities over time. To correctly diagnose AD, compuated-aided system equipped with automatic classification algorithm is of great importance. In this paper, we propose a novel deep learning based classification algorithm that takes advantage of MRI biomarker images including brain areas of hippocampus and cerebrospinal fluid for the purpose of improving the AD classification performance. In particular, we develop a new approach that effectively applies MRI biomarker patch images as input to 3D Deep Convolution Neural Network. To integrate multiple classification results from multiple biomarker patch images, we proposed the effective confidence score fusion that combine classification scores generated from soft-max layer. Experimental results show that AD classification performance can be considerably enhanced by using our proposed approach. Compared to the conventional AD classification approach relying on entire MRI input, our proposed method can improve AD classification performance of up to 10.57% thanks to using biomarker patch images. Moreover, the proposed method can attain better or comparable AD classification performances, compared to state-of-the-art methods.

• 정보과학회 컴퓨팅의 실제 논문지
• /
• 제23권4호
• /
• pp.244-249
• /
• 2017

스테그아날리시스(Steganalysis)란 이미지 등 일반적인 자료에 암호화된 정보를 은닉하는 스테가노그래피(Steganography)에 대한 검출 및 분석 방법으로, 기계학습 기반 방법론을 포함한다. 기존 기계학습 기반 스테그아날리시스는 영상(Image)의 특징(Feature) 추출 및 모델링에 기반하며, 최근 딥러닝(Deep Learning)의 적용으로 검출 정확도가 큰 폭으로 향상되었다. 하지만 현존하는 스테그아날리시스 모델은 단일 스테가노그래피 기법에 대해 국한되어 있어 학습에 사용되지 않은 스테고(Stego) 이미지의 경우 검출이 불가능한 결정적 한계를 가진다. 본 연구에서는 다양한 스테가노그래피 기법으로 생성된 스테고 이미지에 딥러닝을 적용하여 스테그아날리시스를 학습하는 범용적 모델을 제안한다. 다양한 실험을 통해 제안 기법의 효용성 및 가능성을 확인하고, 범용적 스테그아날리시스 모델이 각각에 특화된 검출 기법과 유사한 정확도로 스테고 이미지를 검출할 수 있음을 보인다.

• 정보처리학회논문지A
• /
• 제9A권2호
• /
• pp.217-224
• /
• 2002

3D 공간정보를 이용하여 3D 이미지를 처리하는 시스템이 많이 상용화되어 있다. 기존에 3D 이미지를 처리하기 위한 방법으로 고성능의 시스템을 이용하거나 이미지 압축 기술을 사용하였다. 하지만 고성능의 시스템을 사용하여 GIS 시스템을 구현할 경우 가격의 부담이 크다는 문제점이 있고 이미지 압축 기술을 사용하여 GIS 시스템을 구현할 경우 원 이미지에 손실이 크다는 문제점이 있다. 또한 일반 시스템에서 3D 이미지를 처리하려면 3D 이미지의 파일의 크기가 크기 때문에 공간 이미지를 처리하는데 시간이 오래 걸린다는 단점이 있다. 따라서 본 논문에서는 3D 이미지를 병렬로 처리하여 디스플레이 시간을 단축하는 병렬 알고리즘을 제안한다. 본 논문에서 제시된 병렬 알고리즘은 3D 이미지를 다수의 노드로 분할하여 각 노드에서 이미지를 화면에 디스플레이 하는 방법을 사용한다. 병렬컴퓨터의 노드의 수가 증가함에 따라 제안된 알고리즘의 성능이 증가함을 실험을 통해 보였다.

• 조형예술학연구
• /
• 제4권
• /
• pp.123-146
• /
• 2002

[ $\lceil$ ]The Life of Pigeons$\rfloor$ consists of seven different canvases without a leading image It contains fragments of disassociated ordinary subjects from a capitalistic and consuming society. In this respect, the text itself attains multiple meanings throughout with inner disharmony, disassociation and relationships of differences. The divided seven images look as if they are connected as one and are connected events that are happening at the same time and in similar places. A liberal interpretation of this work is given to viewers when the seven canvases have both relations and gaps at the same time. $\lceil$The Life of Pigeons$\rfloor$ attempts the viewer's disruption through its middle stratum of meaning structure, which is a device for viewers to rearrange and deeply analyze the seven images. As a result, the artist allows the viewers to get lost in self-contradiction. A fundamental formal structure adopting post-modernism and abandoning modernism is what we can detect with detailed analysis of the work. For instance, changing surface style appears by dividing or putting images in obliquely, furthermore it clearly shows that the main subject is divided in form such as the subject's division into seven spaces. There are three major characteristics. First, the form of the images is divided and composed through oblique and overlapped images. Second, the main content of the subject tends to be scattered. Third, the subjects are interpreted in multiple meanings due to their allegory and symbolism. The inquiry of $\lceil$The Life of Pigeons$\rfloor$ proves that it takes a post artistic spirituality as its basis and its subjects are divided by the differences and surrounding relationships.

• 대한전기학회논문지:시스템및제어부문D
• /
• 제49권12호
• /
• pp.683-689
• /
• 2000

Two-dimensional(2-D) fiber-optic code-division multiple-access(FO-CDMA) system utilizes the optical orthogonal signature pattern code(OOSPC) to encode and decode 2-D data. Encoded 2-D data are spatially multiplexed and transmitted through an image fiber and receiver recovers the intended data by means of thresholding process. OOSPC's construction methods based on expansion of the optical orthogonal code, which is used in one-dimensional(1-D) FO-CDMA system, are introduced. Each OOSPC's performances are compared by using the bit error rate(BER) of interfering OOSPC's of other users. From the results we verify that a balanced incomplete block design(BIBD) construction has the best performance among other mehtods. We also propose a decomposed bit-plane method for parallel transmission and recovery of 256 gray-scale images using OOSPC's constructed by the BIBD method. The simulation result encourages the feasibility of parallel transmission and recovery of multiuser's images.

• 복식문화연구
• /
• 제19권3호
• /
• pp.622-636
• /
• 2011

The purpose of this study was to assess the effects of self-image congruence on attitudes toward purchase intentions of children's clothing via the Behavioral Intention Model. The empirical study was conducted via on-line survey and data were collected from mothers with children aged 6 to 10 years. A total of 593 respondents answered the questionnaire and 574 usable data were statistically analyzed. SPSS 18.0 was used to conduct descriptive statistical analysis, factor analysis, reliability analysis, cluster analysis, Chi-square test, ANOVA, and multiple regressions. A K-means cluster analysis was conducted based on three dimensions brand images of children's wear. Respondents were divided into four groups: elegant image group, multiple image group, ordinary image group, and childlike image group. Characteristics of consumers and clothing evaluative criteria that mothers considered important differed significantly across groups. Moreover, based on these groups, each dimension of self-congruence had different effects on brand attitude. Brand attitude and subjective norms had different effects on purchase intentions. In conclusion, levels of self-congruence and factors influencing purchase intention varied according to brand images of children's wear.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제16권4호
• /
• pp.1166-1186
• /
• 2022

Artistic font design has become an integral part of visual media. However, without prior knowledge of the font domain, it is difficult to create distinct font styles. When the number of characters is limited, this task becomes easier (e.g., only Latin characters). However, designing CJK (Chinese, Japanese, and Korean) characters presents a challenge due to the large number of character sets and complexity of the glyph components in these languages. Numerous studies have been conducted on automating the font design process using generative adversarial networks (GANs). Existing methods rely heavily on reference fonts and perform font style conversions between different fonts. Additionally, rather than capturing style information for a target font via multiple style images, most methods do so via a single font image. In this paper, we propose a network architecture for generating multilingual font sets that makes use of geometric structures as content. Additionally, to acquire sufficient style information, we employ multiple style images belonging to a single font style simultaneously to extract global font style-specific information. By utilizing the geometric structural information of content and a few stylized images, our model can generate an entire font set while maintaining the style. Extensive experiments were conducted to demonstrate the proposed model's superiority over several baseline methods. Additionally, we conducted ablation studies to validate our proposed network architecture.

• 대한임베디드공학회논문지
• /
• 제18권3호
• /
• pp.125-132
• /
• 2023

Roundabouts have strengths in traffic flow and safety but can present difficulties for inexperienced drivers. Demand to acquire and analyze drone images has increased to enhance a traffic environment allowing drivers to deal with roundabouts easily. In this paper, we propose a roundabout traffic analysis system that detects, tracks, and analyzes vehicles using a deep learning-based object detection model (YOLOv7) in drone images. About 3600 images for object detection model learning and testing were extracted and labeled from 1 hour of drone video. Through training diverse conditions and evaluating the performance of object detection models, we achieved an average precision (AP) of up to 97.2%. In addition, we utilized SORT (Simple Online and Realtime Tracking) and OC-SORT (Observation-Centric SORT), a real-time object tracking algorithm, which resulted in an average MOTA (Multiple Object Tracking Accuracy) of up to 89.2%. By implementing a method for measuring roundabout entry speed, we achieved an accuracy of 94.5%.

• Korean Journal of Radiology
• /
• 제22권4호
• /
• pp.502-512
• /
• 2021

Objective: To objectively and subjectively assess and compare the characteristics of monoenergetic images [MEI (+)] and polyenergetic images (PEI) acquired by dual-energy CT (DECT) of patients with breast cancer. Materials and Methods: This retrospective study evaluated the images and data of 42 patients with breast cancer who had undergone dual-phase contrast-enhanced DECT from June to September 2019. One standard PEI, five MEI (+) in 10-kiloelectron volt (keV) intervals (range, 40-80 keV), iodine density (ID) maps, iodine overlay images, and Z effective (Z_eff) maps were reconstructed. The contrast-to-noise ratio (CNR) and the signal-to-noise ratio (SNR) were calculated. Multiple quantitative parameters of the malignant breast lesions were compared between the arterial and the venous phase images. Two readers independently assessed lesion conspicuity and performed a morphology analysis. Results: Low keV MEI (+) at 40-50 keV showed increased CNR and SNR_{breast lesion} compared with PEI, especially in the venous phase ([CNR: 40 keV, 20.10; 50 keV, 14.45; vs. PEI, 7.27; p < 0.001], [SNR_{breast lesion}: 40 keV, 21.01; 50 keV, 16.28; vs. PEI, 10.77; p < 0.001]). Multiple quantitative DECT parameters of malignant breast lesions were higher in the venous phase images than in the arterial phase images (p < 0.001). MEI (+) at 40 keV, ID, and Z_eff reconstructions yielded the highest Likert scores for lesion conspicuity. The conspicuity of the mass margin and the visual enhancement were significantly better in 40-keV MEI (+) than in the PEI (p = 0.022, p = 0.033, respectively). Conclusion: Compared with PEI, MEI (+) reconstructions at low keV in the venous phase acquired by DECT improved the objective and subjective assessment of lesion conspicuity in patients with malignant breast lesions. MEI (+) reconstruction acquired by DECT may be helpful for the preoperative evaluation of breast cancer.

• Journal of Korean Neurosurgical Society
• /
• 제56권5호
• /
• pp.400-404
• /
• 2014

Objective : Clip artifacts limit the visualization of intracranial structures in CT scans from patients after aneurysmal clipping with cobalt alloy clips. This study is to analyze the parameters influencing the degree of clip artifacts. Methods : Postoperative CT scans of 60 patients with straight cobalt alloy-clipped aneurysms were analyzed for the maximal diameter of white artifacts and the angle and number of streak artifacts in axial images, and the maximal diameter of artifacts in three-dimensional (3-D) volume-rendered images. The correlation coefficient (CC) was determined between each clip artifact type and the clip blade length and clip orientation to the CT scan (angle a, lateral clip inclination in axial images; angle b, clip gradient to scan plane in lateral scout images). Results : Angle b correlated negatively with white artifacts (r=-0.589, p<0.001) and positively with the angle (r=0.636, p<0.001) and number (r=0.505, p<0.001) of streak artifacts. Artifacts in 3-D images correlated with clip blade length (r=0.454, p=0.004). Multiple linear regression analysis revealed that angle b was the major parameter influencing white artifacts and the angle and number of streak artifacts in axial images (p<0.001), whereas clip blade length was a major factor in 3-D images (p=0.034). Conclusion : Use of a clip orientation perpendicular to the scan gantry angle decreased the amount of white artifacts and allowed better visualization of the clip site.