• Proceedings of the Korean Society of Computer Information Conference
• /
• 2021.01a
• /
• pp.287-288
• /
• 2021

본 논문에서는 macOS의 파일시스템인 HFS+의 B-tree구조를 디지털 포렌식의 관점에서 분석할 수 있는 기능을 갖춘 도구의 구현에 대하여 다룬다. HFS+ 파일시스템의 파일과 디렉토리에 대한 메타정보를 카탈로그 B-tree에서 구하여 디지털 포렌식 정보로 활용한다. HFS+파일시스템 포렌식 분석도구는 C/C++언어로 구현된다. 텍스트 기반의 명령행 프로그램으로 구현되며 macOS/Windows에서 터미널/명령프롬프트에서 각각 실행될 수 있도록 제작된다. 타임스탬프/파일크기/위치 등의 메타데이터의 파싱기능, 리프노드에 저장된 데이터를 이용한 파일/디렉토리 트리 구조의 재구성, B-tree구조에 의한 키워드 탐색 기능, 인덱스 노드 없이 B-tree 리프노드의 구성에 의한 파일/디렉토리 파싱/검색 기능 등이 구현된다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2020.05a
• /
• pp.355-357
• /
• 2020

언어 학습 음원 분석 방법 및 언어 학습 음원을 재생하는 전자 디바이스 연구로, 음원을 문장 단위로 분할하여 스크립트화하는 것을 주요 목표로 한다. 분석과정은 크게 세단계로 나눌 수 있다. 무음 구간 분석, 음원 분할 및 STT 구간, 스크립트 재구성이다. 이런 분석 과정을 통해 나온 결과물의 정확도는 90%로서 본 연구의 목표를 달성한다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2020.05a
• /
• pp.472-475
• /
• 2020

실생활의 사례를 바탕으로 생성된 여러 분야의 데이터셋을 기계학습 (Machine Learning) 문제에 적용하고 있다. 정보보안 분야에서도 사이버 공간에서의 공격 트래픽 데이터를 기계학습으로 분석하는 많은 연구들이 진행 되어 왔다. 본 논문에서는 공격 데이터를 유형별로 정확히 분류할 때, 실생활 데이터에서 흔하게 발생하는 데이터 불균형 문제로 인한 분류 성능 저하에 대한 해결방안을 연구했다. 희소 클래스 관점에서 데이터를 재구성하고 기계학습에 악영향을 끼치는 특징들을 제거하고 DNN(Deep Neural Network) 모델을 사용해 분류 성능을 평가했다.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2022.01a
• /
• pp.93-96
• /
• 2022

본 논문에서는 카메라 센서만을 이용하여 손바닥 사진을 촬영하고 추출된 데이터들을 합성하여 멀티 채널 이미지를 생성 및 분류 모델에 입력하여 신원을 확인하는 딥러닝 모델을 제안한다. 이 모델은 손바닥 사진이 입력되면 손바닥 및 손금 세그멘테이션을 이용하여 마스크 이미지를 추출하고 단일 채널로 구성된 이미지들을 멀티 채널 이미지로 합성/재구성하여 신원을 분류하는 딥러닝 모델이다. 이는 카메라 센서 외 다른 센서가 필요 없다는 장점을 가지고 있으며, 비접촉 신원 인식 시스템에 적용할 수 있다.

• Journal of the Korean Society of Radiology
• /
• v.84 no.1
• /
• pp.240-252
• /
• 2023

Purpose To assess the effect of deep learning image reconstruction (DLIR) for head CT in pediatric patients. Materials and Methods We collected 126 pediatric head CT images, which were reconstructed using filtered back projection, iterative reconstruction using adaptive statistical iterative reconstruction (ASiR)-V, and all three levels of DLIR (TrueFidelity; GE Healthcare). Each image set group was divided into four subgroups according to the patients' ages. Clinical and dose-related data were reviewed. Quantitative parameters, including the signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR), and qualitative parameters, including noise, gray matter-white matter (GM-WM) differentiation, sharpness, artifact, acceptability, and unfamiliar texture change were evaluated and compared. Results The SNR and CNR of each level in each age group increased among strength levels of DLIR. High-level DLIR showed a significantly improved SNR and CNR (p < 0.05). Sequential reduction of noise, improvement of GM-WM differentiation, and improvement of sharpness was noted among strength levels of DLIR. Those of high-level DLIR showed a similar value as that with ASiR-V. Artifact and acceptability did not show a significant difference among the adapted levels of DLIR. Conclusion Adaptation of high-level DLIR for the pediatric head CT can significantly reduce image noise. Modification is needed while processing artifacts.

• Journal of the Korea Society of Computer and Information
• /
• v.14 no.1
• /
• pp.57-64
• /
• 2009

Recently virtualization has been one of the most popular research topics in system software area. While there already exist many commercial virtualization products for general-purpose computer system, little efforts are made to virtualize embedded system. In this paper, we design and implement a virtual machine monitor which divides each physical hardware resource of a embedded system into logical ones and reorganizes them into many virtual machines, which result in running several real-time operating systems concurrently on a single embedded system. We measure various performance metrics of the virtual machine monitor developed on a real embedded system. The results of the study show that our virtual machine monitor has enough potentiality of its application to real-world embedded systems.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2011.11a
• /
• pp.5-8
• /
• 2011

최근 영상이나 이미지로부터 사용자가 원하는 정보를 추출해 내고 재구성 하는 영상 인식, 증강 현실 등의 컴퓨터 비전(Computer Vision) 응용들이 각광을 받고 있다. 이러한 컴퓨터 비전 응용들은 그 동안 많은 알고리즘들의 연구를 통해 꾸준히 개선되고 향상되어 왔으나, 많은 계산량을 요구하기 때문에 임베디드 시스템에서는 널리 쓰이기 힘들었다. 하지만 최근 들어, 스마트폰 등의 모바일 기기에서의 계산 처리 능력이 향상 되고, 소비자 수요가 증가하면서, 이러한 컴퓨터 비전 응용은 점점 모바일 기기에서 널리 쓰이게 되고 있다. 하지만, 여전히 이러한 컴퓨터 응용을 수행하기 위한 계산양은 부족하기 때문에, 충분한 연산량을 제공하기 위한 방법론들이 다양하게 제시되고 있다. 본 논문에서는 이러한 컴퓨터 응용을 위한 프로세서 구조로서 재구성형 프로세서(Reconfigurable Architecture)를 제안한다. 컴퓨터 비전 응용 중 사물 인식 분야에서 널리 쓰이는 SIFT(Scale Invariant Feature Transformation)을 분석하고 이를 재구성형 프로세서에 맵핑하여 성능 향상을 꾀하였다. SIFT의 주요 커널들을 재구성형 프로세서 맵핑한 결과 최소 6.5배에서 최대 9.2배의 성능 향상을 이룰 수 있었다.

• KIPS Transactions on Computer and Communication Systems
• /
• v.5 no.12
• /
• pp.447-454
• /
• 2016

The research on multipath routing has been studied to solve the problem of frequent path breakages due to node and link failures and to enhance data delivery reliability in wireless sensor networks. In the multipath routing, mobile sinks such as soldiers in battle fields and rescuers in disaster areas bring about new challenge for handling their mobility. The sink mobility requests new multipath construction from sources to mobile sinks according to their movement path. Since mobile sinks have continuous mobility, the existing multipath can be exploited to efficiently reconstruct to new positions of mobile sinks. However, the previous protocols do not address this issue. Thus, we proposed an efficient multipath reconstruction protocol called LGMR for mobile sinks in wireless sensor networks. The LGMR address three multipath reconstruction methods based on movement types of mobile sinks: a single hop movement-based local multipath reconstruction, a multiple hop movement-based local multipath reconstruction, and a multiple hop movement-based global multipath reconstruction. Simulation results showed that the LGMR has better performance than the previous protocol in terms of energy consumption and data delivery delay.

• Journal of the Korean Society of Radiology
• /
• v.16 no.7
• /
• pp.919-925
• /
• 2022

This study is an image reconstruction method after cervical CT scan. According to the oblique coronal reconstruction angle, i would like to suggest the reconstruction angle that can best express neural foraminal bony stricture. The angle created by drawing a line passing through the center of the spinal cord in the midsagittal plane of the cervical spine and the uncovertebral joint was measured. A line passing through the center of the spinal cord was drawn from the neural foramen, and the angle formed with the midsagittal plane was measured and compared. At the C4-5 level, the average was 9.2°, at the C5-6 level, the average was 9.9°, and at the C6-7 level, the average was 8.4°, the neural foraminal angle was measured to be larger than the uncovertebral joint angle. There was a statistically significant difference in mean (p<0.01). Also, it was found that the angle increased toward the lower cervical vertebrae. The angle between the neural foramen of the lower cervical vertebrae (C5-6, C6-7) and the center of the spinal cord is between 55 and 60°. Here, if the oblique coronal image is reconstructed to be 90°, the degree of neural foramen stenosis can be observed well. Because it is an image reconstruction using a conventional CT scan image, it does not receive additional radiation exposure. It is of great significance in diagnosing cervical neural foramen bony stenosis.

• Journal of Veterinary Clinics
• /
• v.28 no.5
• /
• pp.510-515
• /
• 2011

This study was performed to quantitatively assess the normal lung volume and density according to the position by multi-detector computed tomography (MDCT) in dogs. Helical CT of the thorax was performed on 4 different positions with dorsal, left lateral, right lateral and ventral recumbency in 6 Pekingese and 6 Maltese dogs. During CT scanning, dogs were kept hyperventilated. Through the 3-dimensional reconstruction of CT images, the lung parameters were measured as the volume and density of the left, right including accessory lobe, and total lung. 3D images represented the different lung shape between Pekingese and Maltese dogs. Their difference of total lung volume and total lung density was not significant on the each position in both breeds. Right lung volume was significantly higher than left. The difference of left and right volume was $66.91{\pm}25.1$ ml. Linear relationship was shown between body weight and lung volume of ventral recumbency position. The dependent lung had higher density and lower volume than nondependent lung in both breed dogs. The volume of nondependent lung was not changed compared with the volume on ventral or dorsal recumbency. The total lung volume measured with MDCT is correlated with the lung density, and the lung density is useful to predict the normal total lung volume.