• 한국방송∙미디어공학회:학술대회논문집
• /
• 한국방송공학회 1997년도 Proceedings International Workshop on New Video Media Technology
• /
• pp.131-136
• /
• 1997

We present a method to estimate a dense and sharp depth map using multiple cameras for the application to flexible video production. A key issue for obtaining sharp depth map is how to overcome the harmful influence of occlusion. Thus, we first propose to selectively use the depth information from multiple cameras. With a simple sort and discard technique, we resolve the occlusion problem considerably at a slight sacrifice of noise tolerance. However, boundary overreach of more textured area to less textured area at object boundaries still remains to be solved. We observed that the amount of boundary overreach is less than half the size of the matching window and, unlike usual stereo matching, the boundary overreach with the proposed occlusion-overcoming method shows very abrupt transition. Based on these observations, we propose a hierarchical estimation scheme that attempts to reduce boundary overreach such that edges of the depth map coincide with object boundaries on the one hand, and to reduce noisy estimates due to insufficient size of matching window on the other hand. We show the hierarchical method can produce a sharp depth map for a variety of images.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제17권3호
• /
• pp.740-753
• /
• 2023

Text recognition in natural scene images is a challenging problem in computer vision. The accurate identification of ship number characters can effectively improve the level of ship traffic management. However, due to the blurring caused by motion and text occlusion, the accuracy of ship number recognition is difficult to meet the actual requirements. To solve these problems, this paper proposes a dual-branch network based on the CRNN identification network. The network couples image restoration and character recognition. The CycleGAN module is used for blur restoration branch, and the Pix2pix module is used for character occlusion branch. The two are coupled to reduce the impact of image blur and occlusion. Input the recovered image into the text recognition branch to improve the recognition accuracy. After a lot of experiments, the model is robust and easy to train. Experiments on CTW datasets and real ship maps illustrate that our method can get more accurate results.

• Journal of information and communication convergence engineering
• /
• 제9권2호
• /
• pp.197-200
• /
• 2011

In this paper, we propose a multi-view stereoscopic image synthesis algorithm for 3DTV system using depth information with an RGB texture from a depth camera. The proposed algorithm synthesizes multi-view images which a virtual convergence camera model could generate. Experimental results showed that the performance of the proposed algorithm is better than those of conventional methods.

• 한국측량학회:학술대회논문집
• /
• 한국측량학회 2003년도 춘계학술발표회 논문집
• /
• pp.187-190
• /
• 2003

Due to complexity of diverse features in urban area, accurate feature extraction is laborious task in aerial and satellite imagery. Especially occlusion by buildings, and image distortion of shadow effects make processing more difficult work. In this study, algorithm was presented to correct of shadow effects in aerial color images. This algorithm enables user to accurately interpretate urban information by correction of shadow effects in aerial color images

• Korean Journal of Radiology
• /
• 제24권7호
• /
• pp.647-659
• /
• 2023

Objective: The study was conducted to investigate the effect of correct occlusion of the left atrial appendage (LAA) on intracardiac blood flow and thrombus formation in patients with atrial fibrillation (AF) using four-dimensional (4D) flow magnetic resonance imaging (MRI) and three-dimensional (3D)-printed phantoms. Materials and Methods: Three life-sized 3D-printed left atrium (LA) phantoms, including a pre-occlusion (i.e., before the occlusion procedure) model and correctly and incorrectly occluded post-procedural models, were constructed based on cardiac computed tomography images from an 86-year-old male with long-standing persistent AF. A custom-made closed-loop flow circuit was set up, and pulsatile simulated pulmonary venous flow was delivered by a pump. 4D flow MRI was performed using a 3T scanner, and the images were analyzed using MATLAB-based software (R2020b; Mathworks). Flow metrics associated with blood stasis and thrombogenicity, such as the volume of stasis defined by the velocity threshold ($\left|\vec{V}\right|$ < 3 cm/s), surface-and-time-averaged wall shear stress (WSS), and endothelial cell activation potential (ECAP), were analyzed and compared among the three LA phantom models. Results: Different spatial distributions, orientations, and magnitudes of LA flow were directly visualized within the three LA phantoms using 4D flow MRI. The time-averaged volume and its ratio to the corresponding entire volume of LA flow stasis were consistently reduced in the correctly occluded model (70.82 mL and 39.0%, respectively), followed by the incorrectly occluded (73.17 mL and 39.0%, respectively) and pre-occlusion (79.11 mL and 39.7%, respectively) models. The surfaceand-time-averaged WSS and ECAP were also lowest in the correctly occluded model (0.048 Pa and 4.004 Pa^-1, respectively), followed by the incorrectly occluded (0.059 Pa and 4.792 Pa^-1, respectively) and pre-occlusion (0.072 Pa and 5.861 Pa^-1, respectively) models. Conclusion: These findings suggest that a correctly occluded LAA leads to the greatest reduction in LA flow stasis and thrombogenicity, presenting a tentative procedural goal to maximize clinical benefits in patients with AF.

• 한국컴퓨터그래픽스학회논문지
• /
• 제21권3호
• /
• pp.27-35
• /
• 2015

본 연구는 전역 조명 기법 중 하나인 환경-광 가림(ambient occlusion)을 이용한 볼륨 렌더링 방법을 설명한다. 볼륨 밀도 분포를 정규 분포로 가정하여, 환경-광 가림을 불투명도 전이함수의 변경과 무관하게 실시간 가시화할 수 있다. 전처리 과정에서 각 복셀 주변의 일정 크기 영역의 평균과 표준편차를 계산하여 두고, 가시화 단계에서 근방의 불투명도를 추정하여 밝기를 계산한다. 이 논문은 본 연구자들의 기존 연구를 발전시켜 이론적 모델을 일반화하고 출력 영상의 화질을 향상시킨다. 구체적으로 다양한 형태의 불투명도 전이함수를 사용할 수 있는 계산 모델을 제안한다. 그리고 영역의 크기를 다양하게 통계량을 생성하여 근처의 물체에 더 높은 가중치를 부여할 수 있도록 하였다. 최종적으로 환경-광 가림 효과와 지역 조명 효과를 혼합하여, 더 현실감 있는 화질의 볼륨 가시화 영상을 실시간으로 생성할 수 있다.

• 방송공학회논문지
• /
• 제17권5호
• /
• pp.838-850
• /
• 2012

본 논문에서는 스테레오 영상에서 얻은 다양한 특징들을 이용하여 다시점 영상을 생성하는 방법을 제안한다. 제안된 기법에서는 먼저 주어진 스테레오 영상에서 명암변화 주목도 지도(intensity gradient saliency map)를 생성한다. 다음으로 좌우 영상 간에 블럭 단위의 움직임을 나타내는 광류(optical flow)를 계산하고 scale-invariant feature transform(SIFT) 기법을 통해 사물의 크기와 회전에 변하지 않는 영상의 특징 점을 구하여 이 특징점 간의 변이를 구한 다음, 이 두 변이 정보들을 결합하여 변이 주목도 지도(disparity saliency map)를 생성 한다. 생성된 변이 주목도 지도는 가려짐 영역 검출을 통해 오류 변이가 제거된다. 세 번째로 영상 워핑시에 직선의 왜곡을 최소화하기 위해 직선 세그먼트를 얻는다. 마지막으로 다시점 영상은 이렇게 추출된 영상 특징들을 제한 조건으로 사용하여 그리드 메쉬(grid-mesh) 기반 영상 워핑(warping) 기법에 의해 생성된다. 실험 결과를 통해 제안한 기법으로 생성된 다시점 영상의 화질이 기존 DIBR 기법보다 우수한 것을 확인할 수 있었다.

• 대한원격탐사학회지
• /
• 제38권3호
• /
• pp.283-298
• /
• 2022

Synthetic Aperture Radar (SAR)영상은 날씨와 주야에 관계없이 취득될 수 있어 감시, 정찰 및 국토안보 등의 목적을 위한 자동표적인식(Automatic Target Recognition, ATR)에 활용 가능성이 높다. 그러나, 식별 시스템 개발을 위해 다양하고 방대한 양의 시험영상을 구축하는 것은 비용, 운용측면에서 한계가 있다. 최근 표적 모델을 이용하여 시뮬레이션된 SAR 영상에 기반한 표적 식별 시스템 개발에 대한 관심이 높아지고 있다. SAR-ATR 분야에서 대표적으로 이용되는 산란점 매칭과 템플릿 매칭 기반 알고리즘을 적용하여 표적식별을 수행하였다. 먼저 산란점 매칭 기반의 식별은 점을 World View Vector (WVV)로 재구성 후 Weighted Bipartite Graph Matching (WBGM)을 수행하였고, 템플릿 매칭을 통한 식별은 서로 인접한 산란점으로 재구성한 두 영상간의 상관계수를 사용하였다. 개발한 두 알고리즘의 식별성능시험을 위해 최근 미국 Defense Advanced Research Projects Agency (DARPA)에서 배포한 표적 시뮬레이션 영상인 Synthetic and Measured Paired Labeled Experiment (SAMPLE) 자료를 사용하였다. 표준 환경, 표적의 부분 폐색, 랜덤 폐색 정도에 따른 알고리즘 성능을 분석하였다. 산란점 매칭 알고리즘의 식별 성능이 템플릿 매칭보다 전반적으로 우수하였다. 10개 표적을 대상으로 표준환경에서의 산란점 매칭기반 평균 식별률은 85.1%, 템플릿 매칭기반은 74.4%이며, 표적별 식별성능 편차 또한 산란점 매칭기법이 템플릿 매칭기법보다 작았다. 표적의 부분 폐색정도에 따른 성능은 산란점 매칭기반 알고리즘이 템플릿 매칭보다 약 10% 높고, 표적의 랜덤 폐색 60% 발생에도 식별률이 73.4% 정도로 비교적 높은 식별성능을 보였다.

• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 2004년도 ICEIC The International Conference on Electronics Informations and Communications
• /
• pp.435-437
• /
• 2004

This paper presents a personal identification method using teeth images. The method uses images for teeth expressions of anterior and posterior occlusion state and LDA-based technique. Teeth images give merits for recognition because teeth, rigid objects, cannot be deformed at the moment of image acquisition. In the experiments, personal identification for 12 people was successful. It was shown that our method can contribute to multi-modal authentication systems.

• 한국정보처리학회:학술대회논문집
• /
• 한국정보처리학회 2019년도 춘계학술발표대회
• /
• pp.472-474
• /
• 2019

Because of the difference in network structure and loss function, Verification and identification models have their respective advantages and limitations for person reidentification (re-ID). In this work, we propose a multi-task network simultaneously computes the identification loss and verification loss for person reidentification. Given a pair of images as network input, the multi-task network simultaneously outputs the identities of the two images and whether the images belong to the same identity. In experiments, we analyze the major factors affect the accuracy of person reidentification. To address the occlusion problem and improve the generalization ability of reID models, we use the Random Erasing Augmentation (REA) method to preprocess the images. The method can be easily applied to different pre-trained networks, such as ResNet and VGG. The experimental results on the Market1501 datasets show significant and consistent improvements over the state-of-the-art methods.