公开(公告)号：WO2013177779A1

公开(公告)日：2013-12-05

申请号：PCT/CN2012/076338

申请日：2012-05-31

Applicant: THOMSON LICENSING ,  ZHANG, Fan ,  CHEN, Zhibo ,  JIANG, Wenfei

Inventor： ZHANG, Fan ,  CHEN, Zhibo ,  JIANG, Wenfei

IPC: H04N17/00 ,  G06T7/00

CPC classification number: G06T7/0002 ,  G06K9/66 ,  G06T3/4084 ,  G06T2207/20048 ,  G06T2207/30168

Abstract: A method and system for determining a quality metric score for image processing are described including accepting a reference image, performing a pyramid transformation on the accepted reference image to produce a predetermined number of scales, applying image division to each scale to produce reference image patches, accepting a distorted image, performing a pyramid transformation on the accepted distorted image to produce the predetermined number of scales, applying image division to each scale to produce distorted image patches, performing a local distortion calculation for corresponding reference and distorted image patches, summing local distortion calculation results for image patch pairs, multiplying results of the summation operation by a positive weight for each scale, summing the results of the multiplication operation and applying a sigmoid function to results of the second summation operation to produce the quality metric score.

Abstract translation: 描述了用于确定用于图像处理的质量度量得分的方法和系统，包括接受参考图像，对所接受的参考图像执行金字塔变换以产生预定数量的刻度，对每个比例应用图像划分以产生参考图像块， 接受失真的图像，对接受的畸变图像执行金字塔变换以产生预定数量的刻度，对每个刻度进行图像划分以产生失真的图像块，对相应的参考和失真的图像块执行局部失真计算，求和局部失真 图像补丁对的计算结果，将求和运算的结果乘以每个比例的正权重，将乘法运算的结果相加，并将S形函数应用于第二加和运算的结果以产生质量度量得分。

公开(公告)号：WO2013078822A1

公开(公告)日：2013-06-06

申请号：PCT/CN2012/074522

申请日：2012-04-23

Applicant: THOMSON LICENSING ,  ZHANG, Fan ,  XIE, Kai ,  JIANG, Wenfei ,  CHEN, Zhibo

Inventor： ZHANG, Fan ,  XIE, Kai ,  JIANG, Wenfei ,  CHEN, Zhibo

IPC: G06T5/00

CPC classification number: G06T7/40 ,  G06T7/0002 ,  G06T7/13 ,  G06T7/44 ,  G06T2207/10016 ,  G06T2207/20028 ,  G06T2207/30168 ,  H04N19/117 ,  H04N19/14 ,  H04N19/154 ,  H04N19/162 ,  H04N19/172 ,  H04N19/865

Abstract: A particular implementation decomposes an image into a structure component and a texture component. An edge strength map is calculated for the structure component, and a texture strength map is calculated for the texture component. Using the edge strength and the texture strength, texture masking weights are calculated. The stronger the texture strength is, or the weaker the edge strength is, the more distortion can be tolerated by human eyes, and thus, the smaller the texture masking weight is. The local distortions are then weighted by the texture masking weights to generate an overall distortion level or an overall quality metric.

Abstract translation: 特定的实现将图像分解成结构组件和纹理组件。 计算结构分量的边缘强度图，并计算纹理分量的纹理强度图。 使用边缘强度和纹理强度，计算纹理掩蔽权重。 纹理强度越强，边缘强度越弱，人眼可以忍受的扭曲越多，纹理遮蔽重量就越小。 然后通过纹理掩蔽权重对局部失真加权，以产生整体失真水平或总体质量度量。

公开(公告)号：WO2013063784A1

公开(公告)日：2013-05-10

申请号：PCT/CN2011/081744

申请日：2011-11-03

Applicant: THOMSON LICENSING ,  JIANG, Wenfei ,  ZHANG, Fan ,  CHEN, Zhibo

Inventor： JIANG, Wenfei ,  ZHANG, Fan ,  CHEN, Zhibo

IPC: H04N7/32

CPC classification number: H04N19/86 ,  H04N19/103 ,  H04N19/117 ,  H04N19/124 ,  H04N19/132 ,  H04N19/14 ,  H04N19/147 ,  H04N19/176 ,  H04N19/192 ,  H04N19/82

Abstract: A particular implementation forms an initial reconstructed image block from inverse quantization and inverse transform, and further refines the reconstructed image block using pixels from neighboring reconstructed blocks. The image block may be refined using a bilateral filter, whose space parameter and range parameter are adaptive to the quantization parameter. The particular implementation can be used in both encoding and decoding when reconstructing an image block. When used in encoding, the particular implementation can be used jointly with coefficient truncation, where some non-zero transform coefficients are set to zero. The number of remaining non-zero transform coefficients after coefficient truncation may be adaptive to the quantization parameter, the variance of the image block, the number of non-zero transform coefficients of the image block, and the index of the last non- zero transform coefficient in a zigzag scanning order.

Abstract translation: 特定的实施方式从反量化和逆变换形成初始重建图像块，并且使用来自相邻重构块的像素进一步精化重构图像块。 图像块可以使用双边滤波器进行精细化，该双边滤波器的空间参数和范围参数适应量化参数。 当重建图像块时，特定的实现可以用于编码和解码两者。 当用于编码时，特定的实现可以与系数截断一起使用，其中一些非零变换系数被设置为零。 系数截断后剩余的非零变换系数的数量可以适应于量化参数，图像块的方差，图像块的非零变换系数的数量以及最后的非零变换的索引 系数为锯齿形扫描顺序。

公开(公告)号：WO2012142731A1

公开(公告)日：2012-10-26

申请号：PCT/CN2011/000705

申请日：2011-04-22

Applicant: TECHNICOLOR (CHINA) TECHNOLOGY CO. LTD. ,  JIANG, Wenfei ,  CHEN, Zhibo ,  ZHANG, Fan

Inventor： JIANG, Wenfei ,  CHEN, Zhibo ,  ZHANG, Fan

IPC: H04N7/46 ,  H04N7/50

CPC classification number: H04N19/50 ,  H04N19/124 ,  H04N19/132 ,  H04N19/147 ,  H04N19/18 ,  H04N19/467 ,  H04N19/60 ,  H04N19/86

Abstract: The invention proposes modification of quantized coefficients for signalling of a post-processing method. Therefore, it is proposed a method for lossy compress- encoding data comprising at least one of image data and audio data. Said method comprises determining quantized coefficients using a quantization of a discrete cosine transformed residual of a prediction of said data. Said method further comprises modifying said quantized coefficients for minimizing rate-distortion cost wherein distortion is determined using a post-processed reconstruction of the data, the post-processed reconstruction being post-processed according to a post¬ processing method, and compress-encoding said modified coefficients. In said proposed method, the post-processing method is that one of n>l different predetermined post processing method candidates whose position in an predetermined order of arrangement of the post processing method candidates equals a remainder of division, by n, of a sum of the modified coefficients. Doing so removes the overhead of flags in the bit stream.

Abstract translation: 本发明提出了对后处理方法的信令的量化系数的修改。 因此，提出了一种包括图像数据和音频数据中的至少一个的有损压缩编码数据的方法。 所述方法包括使用所述数据的预测的离散余弦变换残差的量化来确定量化系数。 所述方法还包括修改所述量化系数以最小化速率失真成本，其中使用所述数据的后处理重建来确定失真，所述后处理重构根据后处理方法进行后处理，并且对所述 修正系数。 在所述提出的方法中，后处理方法是n≥1个不同的预定后处理方法候选中的一个，其处理方法候选的预定排列顺序等于n的余数除以n 修正系数。 这样做会消除位流中标志的开销。

公开(公告)号：WO2014078985A1

公开(公告)日：2014-05-30

申请号：PCT/CN2012/084886

申请日：2012-11-20

Applicant: THOMSON LICENSING ,  JIANG, Wenfei ,  CUI, Hengbin ,  CHEN, Zhibo

Inventor： JIANG, Wenfei ,  CUI, Hengbin ,  CHEN, Zhibo

IPC: G06T5/00

CPC classification number: G06T5/002 ,  G06T5/008 ,  G06T2207/20021

Abstract: Methods and apparatuses for image regularization are described. The image regularization is performed through total variation (TV) lθ regularization, wherein 0 θ lθ regularization is then performed for each image block using the calculated gradient blocks. The processed is performed iteratively to alleviate the problem of non-regularized neighboring blocks affecting the regularization of the current block. The TV lθ regularization is applicable to applications such as image denoising, video compression, and exposure fusion.

Abstract translation: 描述了用于图像正则化的方法和装置。 图像正则化通过总体变化（TV）l＆thetas进行; 正则化，其中0 <＆thetas; <1。 在一个实施例中，图像被分成图像块。 对于每个图像块，通过识别与图像块的内容的取向一致的方向来确定梯度块计算的方向。 电视l＆thetas; 然后使用计算的梯度块对每个图像块执行正则化。 迭代地执行处理以减轻影响当前块的正则化的非正则化相邻块的问题。 电视剧和电视剧 正则化适用于图像去噪，视频压缩和曝光融合等应用。

公开(公告)号：WO2014134802A1

公开(公告)日：2014-09-12

申请号：PCT/CN2013/072253

申请日：2013-03-06

Applicant: THOMSON LICENSING ,  CHEN, Zhibo ,  LIU, Debing ,  GU, Xiaodong ,  ZHANG, Fan

Inventor： CHEN, Zhibo ,  LIU, Debing ,  GU, Xiaodong ,  ZHANG, Fan

IPC: G06Q30/00

CPC classification number: H04N21/8549 ,  G06F17/30843

Abstract: Various implementations relate to providing a pictorial summary, also referred to as a comic book or a narrative abstraction.In one particular implementation, a first portion in a video is accessed, and a second portion in the video is accessed. A weight for the first portion is determined, and a weight for the second portion is determined. A first number and a second number are determined. The first number identifies how many pictures from the first portion are to be used in a pictorial summary of the video. The first number is one or more, and is determined based on the weight for the first portion. The second number identifies how many pictures from the second portion are to be used in the pictorial summary of the video. The second number is one or more, and is determined based on the weight for the second portion.

Abstract translation: 各种实现涉及提供图形概要，也称为漫画书或叙事抽象。在一个特定实现中，访问视频中的第一部分，并且访问视频中的第二部分。 确定第一部分的重量，并确定第二部分的重量。 确定第一个数字和第二个数字。 第一个数字标识了在视频的图形总结中使用第一部分的图片数量。 第一个数字是一个或多个，并且基于第一部分的权重来确定。 第二个数字识别在视频的图片摘要中将使用第二部分的图片数量。 第二个数是一个或多个，并且基于第二部分的权重来确定。

公开(公告)号：WO2013086654A1

公开(公告)日：2013-06-20

申请号：PCT/CN2011/002096

申请日：2011-12-15

Applicant: THOMSON LICENSING ,  ZHANG, Fan ,  LIAO, Ning ,  XIE, Kai ,  CHEN, Zhibo

Inventor： ZHANG, Fan ,  LIAO, Ning ,  XIE, Kai ,  CHEN, Zhibo

IPC: H04N7/26

CPC classification number: H04N17/004 ,  H04N19/124 ,  H04N19/154 ,  H04N19/176

Abstract: Accuracy and efficiency of video quality measurement are major problems to be solved. According to the invention, a method (506) for accurately predicting video quality uses a rational function of the quantization parameter QP, which is corrected by a correction function that depends on content unpredictability CU. Exemplarily, the correction function is a power function of the CU. Both QP and CU can be computed (511) from the video elementary stream, without full decoding the video. This ensures high efficiency.

Abstract translation: 视频质量测量的准确性和效率是需要解决的主要问题。 根据本发明，用于准确地预测视频质量的方法（506）使用量化参数QP的合理函数，该函数由取决于内容不可预测性CU的校正函数来校正。 示例性地，校正功能是CU的功率函数。 可以从视频基本流中计算QP和CU（511），而不对视频进行完全解码。 这确保了高效率。

公开(公告)号：WO2014121572A1

公开(公告)日：2014-08-14

申请号：PCT/CN2013/077085

申请日：2013-06-09

Applicant: THOMSON LICENSING ,  LIAO, Ning ,  CHEN, Zhibo ,  ZHANG, Fan

Inventor： LIAO, Ning ,  CHEN, Zhibo ,  ZHANG, Fan

IPC: H04N17/00

CPC classification number: H04N17/004 ,  G06K9/00536 ,  G06K9/00711 ,  G06K9/036 ,  G06T7/0002 ,  G06T2207/10016 ,  G06T2207/30168 ,  H04N19/103 ,  H04N19/115 ,  H04N19/154

Abstract: Because neighboring frames may affect how a current frame is perceived, we examine different neighborhoods of the current frame and select a neighborhood that impacts the perceived temporal distortion (i.e., when frames are viewed continuously) of the current frame most significantly. Based on spatial distortion (i.e., when a frame is viewed independently of other frames in a video sequence) of frames in the selected neighborhood, we can estimate initial temporal distortion. To refine the initial temporal distortion, we also consider the distribution of distortion in the selected neighborhood, for example, the distance between the current frame and a closest frame with large distortion, or whether distortion occurs in consecutive frames.

Abstract translation: 由于相邻帧可能会影响当前帧的感知方式，因此我们检查当前帧的不同邻域，并选择影响当前帧最明显的感知时间失真（即连续观看帧）的邻域。 基于空间失真（即当独立于视频序列中的其他帧观看帧）时，可以估计初始时间失真。 为了优化初始时间失真，我们还考虑所选邻域中的失真分布，例如当前帧与具有大失真的最近帧之间的距离，或者是否在连续帧中发生失真。

公开(公告)号：WO2014121571A1

公开(公告)日：2014-08-14

申请号：PCT/CN2013/077082

申请日：2013-06-09

Applicant: THOMSON LICENSING ,  LIAO, Ning ,  CHEN, Zhibo ,  ZHANG, Fan

Inventor： LIAO, Ning ,  CHEN, Zhibo ,  ZHANG, Fan

IPC: H04N17/00

CPC classification number: H04N17/004 ,  G06K9/00536 ,  G06K9/00711 ,  G06K9/036 ,  G06T7/0002 ,  G06T2207/10016 ,  H04N19/103 ,  H04N19/114 ,  H04N19/154 ,  H04N19/48 ,  H04N19/89

Abstract: Spatial distortion (i.e., when a frame is viewed independently of other frames in a video sequence) may be quite different from temporal distortion (i.e., when frames are viewed continuously). To estimate temporal distortion, a sliding window approach is used. Specifically, multiple sliding windows around a current frame are considered. Within each sliding window, a large distortion density is calculated and a sliding window with the highest large distortion density is selected. A distance between the current frame and the closest frame with large distortion in the selected window is calculated. Subsequently, the temporal distortion is estimated as a function of the highest large distortion ratio, the spatial distortion for the current frame, and the distance. In another embodiment, a median of spatial distortion values is calculated for each sliding window and the maximum of median spatial distortion values is used to estimate the temporal distortion.

Abstract translation: 空间失真（即，当视频序列中独立于其他帧观看帧）可能与时间失真（即，连续观看帧时）完全不同。 为了估计时间失真，使用滑动窗口方法。 具体地说，考虑当前帧周围的多个滑动窗口。 在每个滑动窗口内，计算出较大的失真密度，并选择具有最大大失真密度的滑动窗口。 计算当前帧与所选窗口中具有大失真的最近帧之间的距离。 随后，将时间失真估计为最大的大失真比，当前帧的空间失真和距离的函数。 在另一个实施例中，为每个滑动窗口计算空间失真值的中值，并且使用中值空间失真值的最大值来估计时间失真。

公开(公告)号：WO2014134801A1

公开(公告)日：2014-09-12

申请号：PCT/CN2013/072248

申请日：2013-03-06

Applicant: THOMSON LICENSING ,  CHEN, Zhibo ,  LIU, Debing ,  GU, Xiaodong ,  ZHANG, Fan

Inventor： CHEN, Zhibo ,  LIU, Debing ,  GU, Xiaodong ,  ZHANG, Fan

IPC: H04N5/445

CPC classification number: H04N21/8549 ,  G06F17/30837 ,  G11B27/031 ,  H04N5/445 ,  H04N21/4307 ,  H04N21/47205 ,  H04N21/485

Abstract: Various implementations relate to providing a pictorial summary, also referred to as a comic book or a narrative abstraction. In one particular implementation, one or more parameters from a configuration guide are accessed. The configuration guide includes one or more parameters for configuring a pictorial summary of a video. The video is accessed. The pictorial summary for the video is generated. The pictorial summary conforms to the one or more accessed parameters from the configuration guide.

Abstract translation: 各种实现涉及提供图形摘要，也称为漫画书或叙事抽象。 在一个特定实现中，访问来自配置指南的一个或多个参数。 配置指南包括用于配置视频的图形概要的一个或多个参数。 视频被访问。 生成视频的图片摘要。 图形摘要符合配置指南中的一个或多个访问参数。