범용디지털카메라를이용한 HDR파노라마환경 맵제작 시스템개발

박은혜,황규현,박상훈 (사)한국컴퓨터그래픽스학회 2012 컴퓨터그래픽스학회논문지 Vol.18 No.2

High dynamic range (HDR) images represent a far wider numerical range of exposures than common digital images. Thus it can accurately store intensity levels of light found in the specific scenes generated by light sources in the real world. Although a kind of professional HDR cameras which support fast accurate capturing has been developed, high costs prevent from employing those in general working environments. The common method to produce a HDR image with lower cost is to take a set of photos of the target scene with a range of exposures by general purpose cameras, and then to transform them into a HDR image by commercial softwares. However, the method needs complicate and accurate camera calibration processes. Furthermore, creating HDR environment maps which are used to produce high quality imaging contents includes delicate time-consuming manual processes. In this paper, we present an automatic HDR panorama environment map generating system which was constructed to make the complicated jobs of taking pictures easier. And we show that our system can be effectively applicable to photo-realistic compositing tasks which combine 3D graphic models with a 2D background scene using image-based lighting techniques. HDR (high dynamic range) 영상은 일반적인 디지털 영상보다 훨씬 더 넓은 수치 범위로 빛의 노출을 저장한다. 따라서실세계에 존재하는 광원들에 의해 표현되는 특정 장면에 내재된 빛의 세기를 매우 정확하게 저장할 수 있다. 이러한HDR 영상을 빠르고 정확하게 촬영할 수 있는 전문가용 HDR 카메라가 개발되었으나 높은 가격으로 인해 아직까지 일반적인작업환경에서이용하기에는어려움이있다.낮은비용으로HDR영상을생성하는일반적인방법은범용디지털카메라를 이용해 동일한 장면을 서로 다른 노출로 반복 촬영하고 상용 소프트웨어에서 이들을 입력받아 하나의 HDR 영상으로 변환하는 것이다. 하지만 이러한 방법은 복잡하고 정확한 카메라 보정을 필요로 하는 작업이다. 더욱이 이방법을 이용해 고급 영상 콘텐츠 제작을 위한 HDR 환경 맵을 생성하는 경우 더 섬세한 수작업과 시간 투자를 필요로한다. 본 논문에서는 이러한 촬영 작업을 자동화 하기위해 개발된 HDR 파노라마 환경 맵 제작 시스템에 대해 자세히설명한다.그리고영상기반라이팅기법을적용하여3D그래픽모델을2D배경영상에삽입하는사실적합성작업에서본시스템이효과적으로이용될수있음을 실제 사례를통해보인다.

모션 블러 및 아웃라이어 처리를 통한 강인한 HDR 영상 복원 방법

조호진(Hojin Cho),이승용(Seungyong Lee) 한국정보과학회 2012 정보과학회논문지 : 소프트웨어 및 응용 Vol.39 No.12

본 논문은 아웃라이어 처리를 통한 강인한 HDR 영상 복원 방법을 제시한다. 기존의 방법들은 LDR 영상들에서 흔히 발생하는 긴 노출시간으로 인한 블러 현상이나 저노출/과노출로 인한 포화 픽셀을 고려하지 않았다. 특히, 포화 픽셀은 영상 획득 모델을 따르지 않고 아웃라이어로 작용하여 복원 결과에 심각한 영향을 미치므로 HDR 영상 복원 시 반드시 고려해야 한다. 본 논문이 제시하는 방법은 MAP(Maximum a posterior)을 기반으로 블러 및 아웃라이어를 동시에 반영하여 카메라 파이프라인 및 HDR 영상 복원 문제를 모델링하고, 블러 추정 및 EM(Expectation-Maximization) 알고리즘 기반의 아웃라이어 추정을 통해 품질 저하가 없는 선명한 HDR 영상을 복원한다. 실험 결과를 통해 본 논문이 제시하는 방법이 기존 방법들과 비교해서 더 효과적으로 고품질의 HDR 영상을 복원할 수 있음을 보인다. We propose a robust method for HDR image reconstruction via outlier handling. Previous methods for HDR image reconstruction have not considered motion blur, which often occurs in LDR images taken with long exposure times, or saturations caused by over-/long-exposures. Especially, saturated pixels play a role of outliers while violating the image acquisition model, thus should be considered explicitly in HDR image reconstruction. In this paper, we formulate the camera pipeline and the HDR image reconstruction problem considering both motion blur and outliers based on MAP (Maximum a posterior), and restore sharp HDR images via blur kernel estimation and EM-based outlier handling. Results show that our method can effectively restore high-quality HDR images compared to the previous methods.

CAM-based HDR image reproduction using CA–TC decoupled JCh decomposition

Kwon, Hyuk-Ju,Lee, Sung-Hak Elsevier 2019 SIGNAL PROCESSING-IMAGE COMMUNICATION - Vol.70 No.-

<P><B>Abstract</B></P> <P>High dynamic range (HDR) imaging has been developed for acquiring scene with a high dynamic range. Tone-mapping operators are used to reproduce the dynamic range of HDR content on a low dynamic range (LDR) devices. iCAM06 is a tone-mapping algorithm based on the color appearance model. iCAM06 contains chromatic adaptation and tone compression; however, the chrominance component of the chromatic adaptation is not preserved after tone compression. This paper describes a novel decoupled processing for the chromatic adaptation and tone compression to preserve the chrominance component and reduce the color distortion of iCAM06. Further, a color appearance model is used to separate and combine the chrominance and luminance components of the decoupled processes. The rendered results confirm that the proposed algorithm is superior to the conventional algorithm.</P> <P><B>Highlights</B></P> <P> <UL> <LI> This method enhances the HDR image quality using tone-color uncoupled processing. </LI> <LI> Color-uncoupled tone processing reduces the overall color distortion of tone mapping. </LI> <LI> Image appearance mapping is conducted through preserved chrominance and hue. </LI> </UL> </P>

다중 노출 영상을 이용한 HDR 영상 복원 방법

조호진(Hojin Cho),이승용(Seungyong Lee) 한국정보과학회 2014 정보과학회논문지 : 소프트웨어 및 응용 Vol.41 No.1

본 논문에서는 부호화된 전자 셔터로부터 획득한 한 장의 다중 노출 영상을 이용한 HDR 이미징 방법을 제안한다. 기존 HDR 이미징 방법들은 정적인 장면을 대상으로 하기 때문에 영상 획득 시 피사체 움직임 및 손떨림 등의 모션을 처리하기 어려웠다. 또한, 노출 마스크와 같은 특수한 H/W를 이용하는 방법들은 구현/비용 문제로 실용성이 떨어진다. 제안된 방법에서는 개별 노출을 갖는 영상들을 분리하지 않기 때문에 계단 현상이 없고 세부사항을 보존할 수 있다. HDR 이미징 결과는 부호화된 전자 셔터가 장착된 프로토타입 카메라로 촬영된 실제 영상들을 이용하였으며, 본 논문의 방법이 기존 방법들에 비해 우수한 품질을 갖는다는 것을 보여준다. We propose a novel HDR imaging method using a coded electronic shutter that allows capturing spatially varying exposures simultaneously in a single image. Previous HDR imaging approaches have difficulties in capturing moving objects and handling camera shakes, since they assume the target scenes to be static. Special imaging devices such as exposure-filtering masks could reduce motion artifacts, but manufacturing costs have limited their practicality. Using the proposed framework, we extend the dynamic range of an input image without extracting the subimages for different exposures while preventing jaggy artifacts and preserving fine details. Experimental results on real images captured using a prototype camera equipped with a coded electronic shutter show that our method can generate higher quality HDR images than previous approaches.

HVS-Aware Single-Shot HDR Imaging Using Deep Convolutional Neural Network

비엔 지아 안,이철,Vien, An Gia,Lee, Chul The Korean Institute of Broadcast and Media Engine 2018 방송공학회논문지 Vol.23 No.3

본 논문은 딥 컨볼루션 뉴럴 네트워크(CNN)를 이용하여 행 별로 서로 다른 노출로 촬영된 단일 영상을 HDR 영상으로 변환하는 기법을 제안한다. 제안하는 알고리즘은 먼저 입력 영상에서 저조도 또는 포화로 인해 발생하는 정보 손실 영역을 CNN을 이용하여 복원하여 휘도맵을 생성한다. 또한, CNN 학습 과정에서 인간의 시각 인지 특성을 고려할 수 있는 손실 함수를 제안한다. 마지막으로 복원된 휘도맵에 디모자이킹 필터를 적용하여 최종 HDR 영상을 획득한다. 컴퓨터 모의실험을 통해 제안하는 알고리즘이 기존의 기법에 비해서 높은 품질의 HDR 영상을 취득하는 것을 확인한다. We propose a single-shot high dynamic range (HDR) imaging algorithm using a deep convolutional neural network (CNN) for row-wise varying exposures in a single image. The proposed algorithm restores missing information resulting from under- and/or over-exposed pixels in an input image and reconstructs the raw radiance map. The main contribution of this work is the development of a loss function for the CNN employing the human visual system (HVS) properties. Then, the HDR image is obtained by applying a demosaicing algorithm. Experimental results demonstrate that the proposed algorithm provides higher-quality HDR images than conventional algorithms.

HVS-Aware Single-Shot HDR Imaging Using Deep Convolutional Neural Network

An Gia Vien(비엔 지아 안),Chul Lee(이철) 한국방송·미디어공학회 2018 방송공학회논문지 Vol.23 No.3

We propose a single-shot high dynamic range (HDR) imaging algorithm using a deep convolutional neural network (CNN) for row-wise varying exposures in a single image. The proposed algorithm restores missing information resulting from under- and/or over-exposed pixels in an input image and reconstructs the raw radiance map. The main contribution of this work is the development of a loss function for the CNN employing the human visual system (HVS) properties. Then, the HDR image is obtained by applying a demosaicing algorithm. Experimental results demonstrate that the proposed algorithm provides higher-quality HDR images than conventional algorithms.

HDR 영상 복원을 위해 대비와 텍스쳐 영역 정보를 고려한 혼합 톤 매핑 기법

강주미(Ju-Mi Kang),박대준(Dae-Jun Park),정제창(Jechang Jeong) 한국방송·미디어공학회 2017 방송공학회논문지 Vol.22 No.4

In this paper, we propose a Tone Mapping Operator (TMO) that preserves global contrast and precisely preserves boundary information. In order to reconstruct a High Dynamic Range (HDR) image to a Low Dynamic Range (LDR) display by using Threshold value vs. Intensity value (TVI) based on Human Visual System (HVS) and contrast value. As a result, the global contrast of the image can be preserved. In addition, by combining the boundary information detected using Guided Image Filtering (GIF) and the detected boundary information using the spatial masking of the Just Noticeable Difference (JND) model, And improved the perceived image quality of the output image. The conventional TMOs are classified into Global Tone Mapping (GTM) and Local Tone Mapping (LTM). GTM preserves global contrast, has the advantages of simple implementation and fast execution time, but it has a disadvantage in that the boundary information of the image is lost and the regional contrast is not preserved. On the other hand, the LTM preserves the local contrast and boundary information of the image well, but some areas are expressed unnatural like the occurrence of the halo artifact phenomenon in the boundary region, and the calculation complexity is higher than that of GTM. In this paper, we propose TMO which preserves global contrast and combines the merits of GTM and LTM to preserve boundary information of images. Experimental results show that the proposed tone mapping technique has superior performance in terms of cognitive quality.

영상기반 조명을 이용한 고화질 애니메이션 제작

박혜정,박은혜,윤성민,윤성의,정영일,황규현,김정호,박상훈 동국대학교 영상문화콘텐츠연구원 2008 영상문화콘텐츠연구 Vol.0 No.-

3D 컴퓨터 그래픽스 분야의 사실적인 렌더링 기법의 발전으로 실사 영상과 그래픽 모델의 합성과 관련된 기술들이 다양한 응용 분야에서 활용되고 있다. HDR로 촬영된 실사 영상에 저장된 라이팅 정보를 3D 그래픽 모델의 렌더링 과정에서 라이팅으로 이용하여 자연스러운 합성 영상을 제작하는 영상기반 라이팅 기법이 개발되어 활용되고 있다. 본 논문에서는 고가의 HDR 전용 카메라 없이도 범용 LDR 디지털 카메라를 이용하여 HDRI 영상을 생성할 수 있는 기능을 제공하는 자동 HDRI 제작 시스템에 대해 설명하고, 이 시스템을 이용해 고화질의 애니메이션 콘텐츠를 효과적으로 제작할 수 있음을 보인다. Recent technological innovations of photo-realistic rendering make it possible for composition methods of photos and graphics models to be applicable to a variety of applications. Image-based lighting, which renders 3D graphics models using real world lighting information stored in photo images with high dynamic range, has been widely used to create photo-realistic compositing image. In this paper, we present an automatic HDR image capturing system based on general purpose LDR(Low Dynamic Range) digital cameras without using very expensive HDR cameras, also show high quality animation contents can be effectively produced by using our system.

Measuring Color Defects in Flat Panel Displays Using HDR Imaging and Appearance Modeling

Giljoo Nam,Haebom Lee,Sungsoo Oh,Kim, Min H. IEEE 2016 IEEE Transactions on Instrumentation and Measureme Vol.65 No.2

<P>Measuring and quantifying color defects in flat panel displays (FPDs) are critical in the FPD industry and related business. Color defects are traditionally investigated by professional human assessors, as color defects are subtle perceptual phenomena that are difficult to detect using a camera system. However, human-based inspection has hindered the quantitative analysis of such color defects. Thus, the industrial automation of color defect measurement in FPDs has been severely limited even by leading manufacturers accordingly. This paper presents a systematic framework for the measurement and numerical evaluation of color defects. Our framework exploits high-dynamic-range imaging to robustly measure physically meaningful quantities of subtle color defects. In addition to the application of advanced imaging technology, an image appearance model is employed to predict the human visual perception of color defects as human assessors do. This proposed automated framework can output quantitative analysis of the color defects. This paper demonstrates the performance of the proposed workflow in investigating subtle color defects in FPDs with a high accuracy.</P>

A Multi-exposure Images Fusion Approach for Very Large Dynamic Range Scenes

Chunmeng Wang,Changhe Tu 보안공학연구지원센터 2014 International Journal of Signal Processing, Image Vol.7 No.5

The exposure fusion technique is very useful for directly fusing multi-exposure images into a high quality image, without the physically based high dynamic range (HDR) production step. In this paper, we present an improved method for multi-exposure images fusion using multi-resolution Laplacian pyramid weighted blending. We propose a Weight Modification Factor (WMF) to modify the original weight map for each image guided by contrast, saturation and well-exposedness of each pixel in the image, and to enhance the weight of the informative pixels in ultra-bright or ultra-dark areas. Compared with previous methods, our approach preserves more useful details for scenes with very large dynamic range and achieves no artifacts. Several objective quality metrics prove the advantages of our method.