• Journal of Korea Multimedia Society
• /
• v.10 no.3
• /
• pp.316-324
• /
• 2007

Maximum Intensity Projection (MIP) identifies patients' anatomical structures from MR or CT data sets. Recently, it becomes possible to generate MIP images with interactive speed by exploiting Graphics Processing Unit (GPU) even in large volume data sets. Generally, volume boundary plane is obliquely crossed with view-aligned texture plane in hardware-texture based volume rendering. Since the ray sampling distance is not increased at volume boundary in volume rendering, the aliasing problem occurs due to data loss. In this paper, we propose an efficient method to overcome this problem by Re-rendering volume boundary planes. Our method improves image quality to make dense distances between samples near volume boundary which is a high frequency area. Since it is only 6 clipping planes are additionally needed for Re-rendering, high quality rendering can be performed without sacrificing computational efficiency. Furthermore, our method couldbe applied to Minimum Intensity Projection (MinIP) volume rendering.

• Journal of KIISE:Computing Practices and Letters
• /
• v.16 no.11
• /
• pp.1131-1135
• /
• 2010

This paper presents a method to capture the posture of the human lower-limbs on the 3D space by using a single camera and a planar mirror. The system estimates the pose of the camera facing the mirror by using four coplanar IR markers attached on the planar mirror. After that, the training space is set up based on the relationship between the mirror and the camera. When a patient steps on the weight board, the system obtains relative position between patients' feet. The markers are attached on the sides of both legs, so that some markers are invisible from the camera due to the self-occlusion. The reflections of the markers on the mirror can partially resolve the above problem with a single camera system. The 3D positions of the markers are estimated by using the geometric information of the camera on the training space. Finally the system estimates and visualizes the posture and motion of the both legs based on the 3D marker positions.

• Proceedings of the Korean Information Science Society Conference
• /
• 2001.10b
• /
• pp.487-489
• /
• 2001

본 논문은 구형 영상기반 가상환경에서 하나의 시점 영상 내에 포함되어 있는 임의의 물체를 모델링하여 3차원 운동이 가능한 시스템을 소개한다. 본 논문에서는 카메라 보정을 하지 않고, 물체에 대한 최소한의 기하학적 정보만을 이용하여 물체를 모델링하고, 모델링된 물체의 영상 기반 운동(image-based motion)의 가능성을 제시한다. 구현된 시스템은 구 환경에서의 하나의 시점 영상을 사영평면으로 간주하고 사용자에 의해 입력된 선과 점으로 투영된 3차원 물체의 2차원 모양을 모델링한다. 그리고 소실점을 이용해서 모델링된 입방체의 3차원 운동을 다룬다.

• Proceedings of the Korea Multimedia Society Conference
• /
• 2004.05a
• /
• pp.696-699
• /
• 2004

본 논문은 현실감 있는 영상 기반 환경 모델의 생성을 위해 영상 분할 기반 환경 모델링 방법을 제안한다. 입력 영상을 환경 특성에 따라 바닥면, 천정(하늘), 주변 물체들로 분할하고 분할된 바닥면이나 천정을 참조 평면으로 설정하고 주변 물체들의 깊이값을 계산함으로써 상세한 환경 모델을 얻을 수 있다. 영상 분할 환경 모델링 방법은 환경 맵에 적용하기 용이하며 환경의 특성에 따른 깊이값 추출 방법으로 손쉽게 환경 모델링이 가능하다. 또한, 시점이 이동되고 시차를 갖는 환경의 표현이 가능하다.

• Broadcasting and Media Magazine
• /
• v.24 no.2
• /
• pp.63-72
• /
• 2019

디스플레이 소자 평면에서부터 육안으로 관찰 가능한 입체영상 영역까지의 물리적인 거리 조절을 바탕으로 공간상에서 재현될 수 있는 3차원 영상에 대한 기대감은 다양한 형태의 디지털 홀로그래픽 디스플레이를 개발하는 주요한 요인이 되어 왔다. 공간 광 변조기의 픽셀 크기에 의존하는 디지털 홀로그램 영상의 시야각과 픽셀 수에 의존하는 홀로그램 영상 크기로 인한 제약 사항들을 극복하는 방법을 토대로 다수의 시청자가 동일한 3차원 영상을 관찰할 수 있는 홀로그래픽 디스플레이에 대한 연구가 나타나기 시작하였다. 본 고에서는 테이블 형태의 공간상에서 관찰 가능한 디지털 홀로그램 기술에 관한 연구 동향을 소개하고자 한다.

• Journal of Korea Multimedia Society
• /
• v.3 no.4
• /
• pp.365-371
• /
• 2000

To restore image degraded by out-of-focus blur and additive noise, an iterative restoration is used. Acceleration parameter is usually applied equally to all over the image without considering the local characteristics of degraded images. As a result, the conventional methods are not effective in restoring severely degraded edge region and shows slow convergence rate. To solve this problem we propose an adaptive iterative restoration according to local degradation, in which the acceleration parameter has low value in flat region that is less degraded and high value in edge region that is more degraded. Through experiments, we verified that the proposed method showed better results with fast convergence rate, showed Visually better image in edge region and lower MSE than the conventional methods.

• Proceedings of the Korean Institute of Communication Sciences Conference
• /
• 1995.10a
• /
• pp.814-817
• /
• 1995

• Proceedings of the KSMRM Conference
• /
• 1996.10a
• /
• pp.18-18
• /
• 1996

• The Journal of the Acoustical Society of Korea
• /
• v.31 no.3
• /
• pp.151-160
• /
• 2012

In medical ultrasound imaging, elasticity imaging helps to diagnose tumors such as cancer. This paper is concerned with the application of acoustic radiation force to soft tissue of interest to implement elasticity imaging. In order to reduce the data acquisition time, instead of relying on transmit focusing, a plane wave of burst type is transmitted to apply the acoustic radiation force simultaneously to an entire imaging region to be observed. A homogeneous phantom experiment confirms that increasing the transmit excitation duration instead of employing transmit focusing generates a high enough acoustic radiation force to obtain elasticity images. It is found, however, that a different displacement versus time characteristic is observed unlike the case of using a conventional focused acoustic radiation force. Experimental results obtained through the use of an ultrasound phantom and a bovine liver show that lesions can be correctly differentiated.

• Journal of the Korea Computer Graphics Society
• /
• v.23 no.3
• /
• pp.95-103
• /
• 2017

Horizon correction is a crucial stage for image composition enhancement. In this paper, we propose a deep learning based method for estimating the slanted angle of a photograph and correcting it. To estimate and correct the horizon direction, existing methods use hand-crafted low-level features such as lines, planes, and gradient distributions. However, these methods may not work well on the images that contain no lines or planes. To tackle this limitation and robustly estimate the slanted angle, we propose a convolutional neural network (CNN) based method to estimate the slanted angle by learning more generic features using a huge dataset. In addition, we utilize multiple adaptive spatial pooling layers to extract multi-scale image features for better performance. In the experimental results, we show our CNN-based approach robustly and accurately estimates the slanted angle of an image regardless of the image content, even if the image contains no lines or planes at all.