时空联合视差优化的立体视频重定向

目的 智能适配显示的图像/视频重定向技术近年受到广泛关注。与图像重定向以及2D视频重定向相比,3D视频重定向需要同时考虑视差保持和时域保持。现有的3D视频重定向方法虽然考虑了视差保持却忽略了对视差舒适度的调整,针对因视差过大和视差突变造成视觉不舒适度这一问题,提出了一种基于时空联合视差优化的立体视频重定向方法,将视频视差范围控制在舒适区间。方法 在原始视频上建立均匀网格,并提取显著信息和视差,进而得到每个网格的平均显著值;根据相似性变化原理构建形状保持能量项,利用目标轨迹以及原始视频的视差变化构建时域保持能量项,并结合人眼辐辏调节原理构建视差舒适度调整能量项;结合各个网格的显著性,联合求解所有能量项得到优化后的网格顶点坐标,将其用于确定网格形变,从而生成指定宽高比的视频。结果 实验结果表明,与基于细缝裁剪的立体视频重定向方法对比,本文方法在形状保持、时域保持及视差舒适度方面均具有更好的性能。另外,使用现有的客观质量评价方法对重定向结果进行评价,本文方法客观质量评价指标性能优于均匀缩放和细缝裁剪的视频重定向方法,时间复杂度较低,每帧的时间复杂度至少比细缝裁剪方法降低了98%。结论 提出的时空联合的视差优化方法同时在时域和舒适度上对视差进行优化,并考虑了时域保持,具有良好的视差优化与时域保持效果,展现了较高的稳定性和鲁棒性。本文方法能够用于3D视频的重定向,在保持立体视觉舒适性的同时适配不同尺寸的3D显示屏幕。     

3维立体动画制作的视觉效果评价

目的 视觉效果评价是3维立体动画制作过程中不可忽略的一环。评价过程主要依靠专业人员的行业经验,受人员知识水平、测试环境等因素的影响。针对该问题,提出了一个主客观结合的评价模型。方法 首先建立了一个面向前期制作的三维动画场景数据集用以训练和测试,针对视觉效果评价的两个重要指标:立体感和视觉舒适度,进行主观实验得到相应的分数;提取全局视觉舒适度特征和感兴趣区域立体感特征,使用支持向量回归(SVR)方法,经过训练和测试得到舒适度评价模型和立体感评价模型。结果 通过将性能验证实验得到验证场景的主观分数与评价模型给出的结果进行比对,结果表明,运用评价模型得到的预测分数与观众主观分数基本一致,该模型可以对影响视觉效果的视觉舒适度和立体感予以5级量化评分。结论 本文所提出的视觉舒适度和立体感评价方法,能建立影响视觉效果的特征与主观评分间的关系,用得到的模型预测分数给制作人员一个及时直观的调节依据标准。     

利用多核增强学习的立体图像舒适度评价模型

目的 传统的立体视觉舒适度评价模型,在学习阶段一般采用回归算法,且需要大量的包含主观测试数据的训练样本,针对这个问题,提出一种利用多核增强学习分类算法的立体图像舒适度评价模型。方法 首先,考虑人们在实际观测图像时,对于先后观测到的不同图像进行相互比较的情况,将评价模型看成是偏好分类器,构造包含偏好标签的偏好立体图像对（PSIP）,构成PSIP训练集;其次,提取多个视差统计特征和神经学模型响应特征;然后,利用基于AdaBoost的多核学习算法来建立偏好标签与特征之间的关系模型,并分析偏好分类概率（即相对舒适度概率）与最终的视觉舒适度之间的映射关系。结果 在独立立体图像库上,与现有代表性回归算法相比较,本文算法的Pearson线性相关系数（PLCC）在0.84以上,Spearman等级相关系数（SRCC）在0.80以上,均优于其他模型的各评价指标;而在跨库测试中,本文算法的PLCC、SRCC指标均优于传统的支持向量回归算法。结论 相比于传统的回归算法,本文算法具有更好的评价性能,能够更为准确地预测立体图像视觉舒适度。     

用户多维感知的3D图像体验质量评价

目的 符合用户视觉特性的3维图像体验质量评价方法有助于准确、客观地体现用户观看3D图像或视频时的视觉感知体验,从而给优化3维内容提供一定的思路。现有的评价方法仅从图像失真、深度感知和视觉舒适度中的一个维度或两个维度出发对立体图像进行评价,评价结果的准确性有待进一步提升。为了更加全面和准确地评价3D图像的视觉感知体验,提出了一种用户多维感知的3D图像体验质量评价算法。方法 首先对左右图像的差异图像和融合图像提取自然场景统计参数表示失真特征;然后对深度图像提取敏感区域,对敏感区域绘制失真前后深度变换直方图,统计深度变化情况以及利用尺度不变特征变换（SIFT）关键点匹配算法计算匹配点数目,两者共同表示深度感知特征;接下来对视觉显著区域提取视差均值、幅值表示舒适度特征;最后综合考虑图像失真、深度感知和视觉舒适度3个维度特征,将3个维度特征归一化后联合成体验质量特征向量,采用支持向量回归（SVR）训练评价模型,并得到最终的体验质量得分。结果 在LIVE和Waterloo IVC数据库上的实验结果表明,所提出的方法与人们的主观感知的相关性达到了0.942和0.858。结论 该方法充分利用了立体图像的特性,评价结果优于比较的几种经典算法,所构建模型的评价结果与用户的主观体验有更好的一致性。     

实时立体视觉系统中的深度映射

目的 为减少立体图像中由于水平视差过大引起的视觉疲劳。针对实时渲染的立体视觉系统,给出了一种非均匀深度压缩方法。方法 该方法在单一相机空间内,通过不同的投影变换矩阵生成双眼图像,水平视差由投影变换来控制。为减少深度压缩造成的模型变形而带来的瑕疵,将不同深度区域内物体施以不同的压缩比例;将相机轴距表示为深度的连续函数,通过相机轴距推导出在单一相机空间内获取双眼图像的坐标变换,将深度压缩转换为模型的坐标变换,从而保证压缩比例的连续变化。结果 实验结果表明,该方法能有效提高立体图像的质量。结论 该方法简单、高效,可应用于游戏、虚拟现实等实时立体视觉系统。     

融合边缘保持与改进代价聚合的立体匹配算法

目的 立体匹配是计算机双目视觉的重要研究方向,主要分为全局匹配算法与局部匹配算法两类。传统的局部立体匹配算法计算复杂度低,可以满足实时性的需要,但是未能充分利用图像的边缘纹理信息,因此在非遮挡、视差不连续区域的匹配精度欠佳。为此,提出了融合边缘保持与改进代价聚合的立体匹配。方法 首先利用图像的边缘空间信息构建权重矩阵,与灰度差绝对值和梯度代价进行加权融合,形成新的代价计算方式,同时将边缘区域像素点的权重信息与引导滤波的正则化项相结合,并在多分辨率尺度的框架下进行代价聚合。所得结果经过视差计算,得到初始视差图,再通过左右一致性检测、加权中值滤波等视差优化步骤获得最终的视差图。结果 在Middlebury立体匹配平台上进行实验,结果表明,融合边缘权重信息对边缘处像素点的代价量进行了更加有效地区分,能够提升算法在各区域的匹配精度。其中,未加入视差优化步骤的21组扩展图像对的平均误匹配率较改进前减少3.48%,峰值信噪比提升3.57 dB,在标准4幅图中venus上经过视差优化后非遮挡区域的误匹配率仅为0.18%。结论 融合边缘保持的多尺度立体匹配算法有效提升了图像在边缘纹理处的匹配精度,进一步降低了非遮挡区域与视差不连续区域的误匹配率。     

融合视觉风格和标签约束的少数民族服装图像解析

目的 少数民族服装款式结构复杂,视觉风格各异。由于缺少民族服装语义标签、局部特征繁杂以及语义标签之间存在相互干扰等因素导致少数民族服装图像解析准确率和精度较低。因此,本文提出了一种融合视觉风格和标签约束的少数民族服装图像解析方法。方法 首先基于本文构建的包含55个少数民族的服装图像数据集,按照基本款式结构、着装区域、配饰和不同视觉风格自定义少数民族服装的通用语义标签和民族语义标签,同时设置4组标注对,共8个标注点;然后,结合自定义语义标签和带有标注对的训练图像,在深度完全卷积神经网络SegNet中加入视觉风格以融合局部特征和全局特征,并引入属性预测、风格预测和三元组损失函数对输入的待解析图像进行初步解析;最后,通过构建的标签约束网络进一步优化初步解析结果,避免标签相互干扰,得到优化后的最终解析结果。结果 在构建的少数民族服装图像数据集上进行验证,实验结果表明,标注对有效提升了局部特征的检测准确率,构建的视觉风格网络能够有效融合少数民族服装的全局特征和局部特征,标签约束网络解决了标签之间相互干扰的问题,在结合视觉风格网络和标签约束网络后,能够明显提升少数民族服装解析的平均精度,像素准确度达到了90.54%。结论 本文提出的融合视觉风格和标签约束的少数民族服装图像解析方法,能够提高少数民族服装图像解析的准确率和精度,对传承祖国文化、保护非物质文化遗产具有很好的意义。     

GPU近实时线性双目立体代价聚合

目的 近年来双目视觉领域的研究重点逐步转而关注其“实时化”策略的研究,而立体代价聚合是双目视觉中最为复杂且最为耗时的步骤,为此,提出一种基于GPU通用计算(GPGPU)技术的近实时双目立体代价聚合算法。方法 选用一种匹配精度接近于全局匹配算法的局部算法——线性立体匹配算法(linear stereo matching)作为代价聚合策略;结合线性代价聚合的原理,对其主要步骤(代价计算、均值滤波及系数求解等)的计算流程进行有针对性地并行优化。结果 对于相同的实验样本,用本文方法在NVIDA GTX780 实验平台上能在更短的时间计算出代价矩阵,与原有的CPU实现方法相比,代价聚合的效率平均有了数十倍的提升。结论 实时双目立体代价聚合方法,为在个人通用PC平台上实时获取高质量双目视觉深度信息提供了一个高效可靠的途径。     

利用条件生成对抗网络的光场图像重聚焦

目的 传统的基于子视点叠加的重聚焦算法混叠现象严重,基于光场图像重构的重聚焦方法计算量太大,性能提升困难。为此,本文借助深度神经网络设计和实现了一种基于条件生成对抗网络的新颖高效的端到端光场图像重聚焦算法。方法 首先以光场图像为输入计算视差图,并从视差图中计算出所需的弥散圆(circle of confusion,COC)图像,然后根据COC图像对光场中心子视点图像进行散焦渲染,最终生成对焦平面和景深与COC图像相对应的重聚焦图像。结果 所提算法在提出的仿真数据集和真实数据集上与相关算法进行评价比较,证明了所提算法能够生成高质量的重聚焦图像。使用峰值信噪比(peak signal to noise ratio,PSNR)和结构相似性(structural similarity,SSIM)进行定量分析的结果显示,本文算法比传统重聚焦算法平均PSNR提升了1.82 dB,平均SSIM提升了0.02,比同样使用COC图像并借助各向异性滤波的算法平均PSNR提升了7.92 dB,平均SSIM提升了0.08。结论 本文算法能够依据图像重聚焦和景深控制要求,生成输入光场图像的视差图,进而生成对应的COC图像。所提条件生成对抗神经网络模型能够依据得到的不同COC图像对输入的中心子视点进行散焦渲染,得到与之对应的重聚焦图像,与之前的算法相比,本文算法解决了混叠问题,优化了散焦效果,并显著降低了计算成本。     

视觉信息损失先验的图像分层去雾方法

目的 针对自然场景下含雾图像呈现出低对比度和色彩失真的问题,提出一种基于视觉信息损失先验的图像去雾算法,将透射图预估转化成求解信息损失函数最小值的目标规划问题。方法 首先通过输入图像的视觉特性将图像划分成含雾浓度不同的3个视觉区域。然后根据含雾图像的视觉先验知识构造视觉信息损失函数,通过像素值溢出映射规律对透射率取值范围进行约束,采用随机梯度下降法求解局部最小透射率图。最后将细化后的全局透射率图代入大气散射模型求解去雾结果。结果 结合现有的典型去雾算法进行仿真实验,本文算法能够有效地复原退化场景的对比度和清晰度,相比于传统算法,本文算法在算法实时性方面提升约20%。结论 本文算法在改善中、浓雾区域去雾效果的同时,提升了透射图预估的效率,对改善雾霾天气下视觉成像系统的能见度和鲁棒性具有重要意义。     

Visual comfort enhancement study based on visual attention detection for stereoscopic displays

Although numerous potential causes may lead to visual discomfort when viewing content on three‐dimensional (3D) displays, vergence–accommodation conflict is a particular cause of binocular parallax‐based stereoscopic displays, and it is unavoidable. Based on the study of 3D content visual attention, we proposed a novel stereoscopic depth adjustment method to improve the visual comfort and enhance perceived naturalness. The proposed method combined the 3D image saliency and specific viewing condition to establish a novel model for computing the optimum zero‐disparity plane of stereoscopic image. The results of perception experiments, focused on visual comfort and stereoscopic sensation, supported that the proposed method can significantly enhance stereoscopic viewing comfort and even can improve the stereoscopic sensation by insuring the 3D image fusion.     

Visual comfort enhancement in stereoscopic 3D images using saliency-adaptive nonlinear disparity mapping

Perceptually salient regions have a significant effect on visual comfort in stereoscopic 3D (S3D) images. The conventional method of obtaining saliency maps is linear combination, which often weakens the saliency influence and distorts the original disparity range significantly. In this paper, we propose visual comfort enhancement in S3D images using saliency-adaptive nonlinear disparity mapping. First, we obtain saliency-adaptive disparity maps with visual sensitivity to maintain the disparity-based saliency influence. Then, we perform nonlinear disparity mapping based on a sigmoid function to minimize disparity distortions. Finally, we generate visually comfortable S3D images based on depth-image-based-rendering (DIBR). Experimental results demonstrate that the proposed method successfully improves visual comfort in S3D images by producing comfortable S3D images with high mean opinion score (MOS) while keeping the overall viewing image quality.     

Visual comfort in 2D/3D modes of glasses‐type stereoscopic displays

In this study, we compared visual comfort in 2D/3D modes of the pattern retarder (PR) and shutter glasses (SG) stereoscopic displays by changing viewing factors and image contents. The viewing factors include ambient illuminance/monitor luminance/background luminance and image contents mainly are determined with different disparity limits. The degrees of 2D/3D visual comfort were investigated by using various combinations of ambient illuminance, monitor luminance, background luminance, and disparity limit. A series of psychological experiments were also performed to compare 2D and 3D viewing experiences for the passive PR and active SG stereoscopic displays and to discover more comfortable conditions under various variable combinations. The experiment results show that the various variable combinations affecting visual comfort in the passive PR and active SG stereoscopic displays were significantly different. Finally, we suggest more comfortable conditions of viewing 2D and 3D images for the PR and SG stereoscopic displays.     

Visual discomfort relaxation for stereoscopic 3D images by adjusting zero-disparity plane for projection

It is a challenging task to improve the visual comfort of a stereoscopic 3D (S3D) image with satisfactory viewing experience. In this paper, we propose a visual comfort improvement scheme by adjusting zero-disparity plane (ZDP) for projection. The degree of visual discomfort is predicted by considering three factors: spatial frequency, disparity response, and visual attention. Then, the selection of an optimal ZDP is guided by the predicted visual discomfort map. Finally, the disparity ranges of the crossed and uncrossed disparities are automatically adjusted according to the ZDP as requirements. Experiment results show that the proposed scheme is effective in improving visual comfort while preserving the unchanged depth sensation.     

Visual comfort in autostereoscopic display

Autostereoscopic displays are likely to become widely used products in the future. However, certain physiological factors, especially visual comfort, limit their development. In this study, four observational parameters – ambient illuminance, image content, scaling ratio, and horizontal distance between major and minor objects – were evaluated to determine the degree of visual comfort offered by 3D computer‐generated images on an autostereoscopic display. Visual comfort score with the range of 0–1 is designed to represent the degree of visual comfort for the 3D images with different manipulations of ambient illuminance, image content, scaling ratio, and horizontal distance between major and minor objects in this study. Subjects were asked to indicate images that produced discomfort. The proportion of images for each condition where participants indicated that viewing the image was comfortable was computed. Images receiving a proportion of 0.5 or greater were classified as acceptable. The disparity ranges over which acceptable images were attained for each participant and for each condition were analyzed with analysis of variance. The analytical results indicate that ambient illuminance and image content have a significant effect on the acceptable disparity range, while scaling ratio and horizontal distance between major and minor objects did not.     

Eye movements and visual discomfort when viewing stereoscopic 3D content

The visual brain fuses the left and right images projected onto the two eyes from a stereoscopic 3D (S3D) display, perceives parallax, and rebuilds a sense of depth. In this process, the eyes adjust vergence and accommodation to adapt to the depths and parallax of the points they gazed at. Conflicts between accommodation and vergence when viewing S3D content potentially lead to visual discomfort. A variety of approaches have been taken towards understanding the perceptual bases of discomfort felt when viewing S3D, including extreme disparities or disparity gradients, negative disparities, dichoptic presentations, and so on. However less effort has been applied towards understanding the role of eye movements as they relate to visual discomfort when viewing S3D. To study eye movements in the context of S3D viewing discomfort, a Shifted-S3D-Image-Database (SSID) is constructed using 11 original nature scene S3D images and their 6 shifted versions. We conducted eye-tracking experiments on humans viewing S3D images in SSID while simultaneously collecting their judgments of experienced visual discomfort. From the collected eye-tracking data, regions of interest (ROIs) were extracted by kernel density estimation using the fixation data, and an empirical formula fitted between the disparities of salient objects marked by the ROIs and the mean opinion scores (MOS). Finally, eye-tracking data was used to analyze the eye movement characteristics related to S3D image quality. Fifteen eye movement features were extracted, and a visual discomfort predication model learned using a support vector regressor (SVR). By analyzing the correlations between features and MOS, we conclude that angular disparity features have a strong correlation with human judgments of discomfort.     

Leveraging visual attention and neural activity for stereoscopic 3D visual comfort assessment

Visual comfort assessment (VCA) for stereoscopic three-dimensional (S3D) images is a challenging problem in the community of 3D quality of experience (3D-QoE). The goal of VCA is to automatically predict the degree of perceived visual discomfort in line with subjective judgment. The challenges of VCA typically lie in the following two aspects: 1) formulating effective visual comfort-aware features, and 2) finding an appropriate way to pool them into an overall visual comfort score. In this paper, a novel two-stage framework is proposed to address these problems. In the first stage, primary predictive feature (PPF) and advanced predictive feature (APF) are separately extracted and then integrated to reflect the perceived visual discomfort for 3D viewing. Specifically, we compute the S3D visual attention-weighted disparity statistics and neural activities of the middle temporal (MT) area in human brain to construct the PPF and APF, respectively. Followed by the first stage, the integrated visual comfort-aware features are fused with a single visual comfort score by using random forest (RF) regression, mapping from a high-dimensional feature space into a low-dimensional quality (visual comfort) space. Comparison results with five state-of-the-art relevant models on a standard benchmark database confirm the superior performance of our proposed method.