公开(公告)号：US20030016754A1

公开(公告)日：2003-01-23

申请号：US09902438

申请日：2001-07-10

申请人： MOTOROLA, INC

发明人： Bhavan Gandhi ,  Kevin O'Connell ,  David Nicozisin

IPC分类号： H04N007/12

CPC分类号： H04N19/423 ,  H04N19/107 ,  H04N19/17 ,  H04N19/61

摘要： A method and apparatus for reducing error propagation in digital video signals using random forced intra-refresh of macroblocks. One or more predetermined regions are defined for each digital video frame. Within each predetermined region, a number of macroblocks are selected according to a random permutation of the macroblocks within the region. The selected macroblocks are intra-coded, while the remaining macroblocks are coded according to a standard video compression protocol. This approach provides an efficient method for mitigating error propagation in a decoder. Interior regions may be smaller than exterior regions, providing higher quality for the interior regions, where sensitivity to errors is higher.

摘要翻译： 一种使用宏块的随机强制内部刷新来减少数字视频信号中的误差传播的方法和装置。 为每个数字视频帧定义一个或多个预定区域。 在每个预定区域内，根据该区域内的宏块的随机排列选择多个宏块。 所选择的宏块是帧内编码的，而剩余的宏块根据标准视频压缩协议进行编码。 这种方法提供了一种用于减轻解码器中的错误传播的有效方法。 内部区域可以小于外部区域，为内部区域提供更高的质量，其中误差的敏感性更高。

公开(公告)号：US20040234108A1

公开(公告)日：2004-11-25

申请号：US10443637

申请日：2003-05-22

申请人： Motorola, Inc.

发明人： Dongge Li ,  Bhavan Gandhi

IPC分类号： G06K009/00

CPC分类号： G06K9/00288 ,  G06K9/72

摘要： A processor (10) utilizes information regarding one or more physical dimensions of an individual (14) to better inform a personal identification process. In one embodiment, the measured physical dimensions are utilized to influence the conduct of a face recognition process. In one embodiment, a Bayesian Belief Network can be utilized to facilitate such processes.

摘要翻译： 处理器（10）利用关于个人（14）的一个或多个物理维度的信息来更好地通知个人识别过程。 在一个实施例中，所测量的物理尺寸用于影响面部识别过程的进行。 在一个实施例中，贝叶斯信仰网络可以用于促进这种过程。

公开(公告)号：US20040085483A1

公开(公告)日：2004-05-06

申请号：US10286283

申请日：2002-11-01

申请人： Motorola, Inc.

发明人： Zhu Li ,  Bhavan Gandhi ,  Aggelos K. Katsaggelos

IPC分类号： H04N005/14

CPC分类号： G06F17/30802 ,  G06F17/30811 ,  G06F17/30843 ,  G11B27/031 ,  G11B27/032 ,  G11B27/28 ,  H04N5/144 ,  H04N5/147

摘要： At least one visual significance metric is determined (12) for at least some frames belonging to an original series of frames (11). Key frames are identified (13) as a function, at least in part, of the visual significance metric. Cumulative visual significance values are then determined (14) for at least some of the frames that intervene between each pair of key frames. These cumulative visual significance values are then used to identify (15) frames of additional content interest. Various frames are then selected (16) for use in a visual summary. In one embodiment, all of the key frames and frames of additional content interest are selected for inclusion in the visual summary.

摘要翻译： 对于属于原始系列的帧（11）的至少一些帧，确定至少一个视觉重要性度量（12）。 至少部分地根据视觉重要性度量将关键帧（13）识别为（13）。 然后确定（14）对于每对关键帧之间介入的至少一些帧的累积视觉有效值。 然后使用这些累积视觉显着性值来识别附加内容感兴趣的（15）帧。 然后选择各种帧（16）用于视觉总结。 在一个实施例中，选择附加内容感兴趣的所有关键帧和帧以包括在视觉摘要中。

公开(公告)号：US20030012286A1

公开(公告)日：2003-01-16

申请号：US09902124

申请日：2001-07-10

申请人： MOTOROLA, INC.

发明人： Faisal Ishtiaq ,  Bhavan Gandhi ,  Kevin O'Connell

IPC分类号： H04N007/12

CPC分类号： H04N19/89 ,  H04N19/895

摘要： A method and device for detecting errors in a digital video signal comprising a sequence of image frames, each image frame comprising a sequence of image slices, each image slice comprising a sequence of macroblocks and each macroblock comprising a plurality of pixels. A macroblock decoder includes an error detection unit that operates to calculate an error metric between pixel values on at least part of the boundary between a current macroblock and one or more adjoining macroblocks and to label the current macroblock as suspicious if the error metric is greater than a threshold level. The threshold level is adjusted according to a weighted average error metric from one or more previous image frames. Suspicious macroblocks and subsequent inter-coded macroblocks may be regenerated according to a concealment strategy if a syntax error is found within the current image slice.

摘要翻译： 一种用于检测包括图像帧序列的数字视频信号中的错误的方法和装置，每个图像帧包括一系列图像片段，每个图像片段包括宏块序列，每个宏块包括多个像素。 宏块解码器包括错误检测单元，其操作以计算当前宏块和一个或多个相邻宏块之间的边界的至少一部分上的像素值之间的误差度量，并且如果误差度量大于当前宏块，则将当前宏块标记为可疑 阈值水平。 根据来自一个或多个先前图像帧的加权平均误差度量来调整阈值电平。 如果在当前图像切片内发现语法错误，则可以根据隐藏策略重新生成可疑宏块和随后的帧间编码宏块。

公开(公告)号：US20030053546A1

公开(公告)日：2003-03-20

申请号：US09901809

申请日：2001-07-10

申请人： MOTOROLA, INC.

发明人： Bhavan Gandhi ,  Kevin O'Connell ,  Faisal Ishtiaq ,  Raghavan Subramaniyan

IPC分类号： H04N007/12

CPC分类号： H04L1/20

摘要： The disclosed invention is a method to detect candidate errors within the picture start code, picture header, and picture timestamp. Upon detection, these errors may be confirmed and the impacts mitigated. An error within the timestamp bit field is adaptively detected by the use of a threshold comparison, and a mechanism for concealing the timestamp information leaves only a small timestamp anomaly. An error within the picture start code (PSC) is determined by adaptively analyzing the number of bits and the macro-block location of the next slice or GOB. If a PSC is suspected to have been overrun due to an error, this method allows for data beyond the first GOB or slice to be recovered in the frame. Depending on the extent of use of slices and GOBs, the method can recover a majority of the frame that otherwise would have been completely lost. A candidate error within the picture header is evaluated by utilizing the first slice or GOB of the frame and analyzing information from a previous frame to check whether the current frame's header information should be replaced with the previous frame's header information. Replacing the erred frame header can result in a majority of the data within the frame being recovered that would otherwise be discarded if the header was not replaced.

摘要翻译： 所公开的发明是检测图像起始码，图像头部和图像时间戳内的候选误差的方法。 经检测，可能会确认这些错误并减轻影响。 通过使用阈值比较来自适应地检测时间戳位字段内的错误，并且用于隐藏时间戳信息的机制仅留下小的时间戳异常。 通过自适应地分析下一个切片或GOB的位数和宏块位置来确定图像起始码（PSC）内的错误。 如果PSC被怀疑由于错误而被超载，则该方法允许超出第一个GOB或切片的数据在帧中被恢复。 根据切片和GOB的使用范围，该方法可以恢复否则将完全丢失的大部分帧。 通过利用帧的第一片或GOB并分析来自前一帧的信息来评估图片头中的候选错误，以检查当前帧的标题信息是否应该被前一帧的标题信息替换。 更换错误的帧头可能会导致正在恢复的帧内的大部分数据，否则如果未更换标题，则将被丢弃。