公开(公告)号：US20130336402A1

公开(公告)日：2013-12-19

申请号：US13895939

申请日：2013-05-16

Applicant: Lidong Xu ,  Yi-Jen Chiu

Inventor： Lidong Xu ,  Yi-Jen Chiu

IPC: H04N7/36

CPC classification number: H04N19/139 ,  H04N5/144 ,  H04N19/513 ,  H04N19/521 ,  H04N19/56 ,  H04N19/57

Abstract: Systems, methods, and computer program products that can be used to determine a search range (SR) when performing motion estimation at, for example, a video encoder or decoder. Determining a motion vector for a current block during motion estimation may involve searching within a search window that may reside in a reference frame, or in a previously decoded block that spatially or temporally neighbors the current block. Such a search seeks a motion vector that minimizes a metric, such as a sum of absolute differences between corresponding blocks of reference frames. A motion vector that minimizes such a metric may be a good candidate for use in motion estimation. The search may become more efficient if a search range is determined such that the extent of the search is bounded. A search range may be determined at the block level or at the picture level.

Abstract translation: 当在例如视频编码器或解码器上执行运动估计时，可用于确定搜索范围（SR）的系统，方法和计算机程序产品。 在运动估计期间确定当前块的运动矢量可以涉及在可能驻留在参考帧中的搜索窗口内或在空间上或时间上邻近当前块的先前解码的块中进行搜索。 这样的搜索寻找使度量最小化的运动矢量，诸如对应的参考帧块之间的绝对差的和。 使这种度量最小化的运动矢量可能是用于运动估计的良好候选。 如果确定搜索范围使得搜索范围有界，则搜索可能变得更有效。 可以在块级别或图片级别确定搜索范围。

公开(公告)号：US20120257104A1

公开(公告)日：2012-10-11

申请号：US13083681

申请日：2011-04-11

Applicant: Jong Dae Oh ,  Yi-Jen Chiu ,  Sang-Hee Lee ,  Ning Lu

Inventor： Jong Dae Oh ,  Yi-Jen Chiu ,  Sang-Hee Lee ,  Ning Lu

IPC: H04N7/01

CPC classification number: H04N7/0115

Abstract: The format of telecined video may be determined including a bottom field first cadence. In addition, video using 2:3:3:2 top field first can be identified. Moreover, mixed cadence videos can also be detected. In some embodiments, mixed cadence videos may be detected by calculating variances of different areas within a frame.

Abstract translation: 可以确定电视视频的格式，包括底场第一节奏。 另外，可以首先识别使用2：3：3：2顶场的视频。 此外，还可以检测到混合的节奏视频。 在一些实施例中，可以通过计算帧内不同区域的方差来检测混合节奏视频。

公开(公告)号：US08184693B2

公开(公告)日：2012-05-22

申请号：US12082561

申请日：2008-04-11

Applicant: Yi-Jen Chiu ,  Lidong Xu

Inventor： Yi-Jen Chiu ,  Lidong Xu

IPC: H04N7/12

CPC classification number: H04N19/36 ,  H04N19/117 ,  H04N19/30 ,  H04N19/46 ,  H04N19/61 ,  H04N19/80

Abstract: Adaptive filtering may be used to increase the quality of tone mapped, baseline layer encoded information. As a result, scalable video codecs may be implemented with improved picture quality in some embodiments.

Abstract translation: 自适应滤波可用于增加色调映射，基线层编码信息的质量。 结果，在一些实施例中，可升级的视频编解码器可以用改进的图像质量来实现。

公开(公告)号：US20100008427A1

公开(公告)日：2010-01-14

申请号：US12283305

申请日：2008-09-11

Applicant: Yi-Jen Chiu ,  Hong Jiang ,  Lidong Xu ,  Ya-Ti Peng

Inventor： Yi-Jen Chiu ,  Hong Jiang ,  Lidong Xu ,  Ya-Ti Peng

IPC: H04N11/02

CPC classification number: H04N9/64 ,  H04N1/64 ,  H04N19/186 ,  H04N19/30 ,  H04N19/60

Abstract: Techniques are described that can be used to either compress or expand video. Color compression techniques are described that can be used to compress the wide color gamut content into lower color gamut for inclusion in a baseline layer. Color expansion techniques are described that convert lower color gamut data into wider color gamut format for inclusion in an enhancement layer. Both of the baseline video stream and enhancement layer video streams may be transmitted through a channel or stored in a memory device to be viewed later. Accordingly, both baseline and enhancement video layers are available so that either lower or higher quality displays can be used to display video.

Abstract translation: 描述了可用于压缩或扩展视频的技术。 描述了可用于将宽色域内容压缩到较低色域以包含在基线层中的色彩压缩技术。 描述了将较低色域数据转换成更宽的色域格式以包含在增强层中的颜色展开技术。 基线视频流和增强层视频流都可以通过信道发送或存储在存储设备中，以便稍后观看。 因此，基线和增强视频层都可用，以便可以使用较低或较高质量的显示器来显示视频。

公开(公告)号：US20090167951A1

公开(公告)日：2009-07-02

申请号：US11967754

申请日：2007-12-31

Applicant: Yi-Jen Chiu

Inventor： Yi-Jen Chiu

IPC: H04N5/00

CPC classification number: H04N5/21 ,  G06T5/002 ,  G06T2207/10016 ,  G06T2207/20012 ,  G06T2207/20182 ,  G06T2207/20192 ,  H04N5/144 ,  H04N19/117 ,  H04N19/86

Abstract: Implementations of a history-based temporal motion noise filtering technique that considers the temporal smooth among multiple pictures as well as the block-based technique to estimate the noise/motion history to better reflect the spatial/temporal smoothness in the local neighborhood are provided. In particular, implementations of the per-pixel blending technique in the block-based noise measurement may be used to better manipulate pixels in both the spatial and temporal domains. A global noise detection technique may be used to estimate the occurrence and/or strength of the noise. A content adaptive spatial filtering content adaptive spatial filter based on a local edge measurement may be used to reduce picture noise as well as preserve edge sharpness. Implementations may be configured for various applications. In particular, programmability options allow users to specify the filter parameters for singularity detection, spatial-only, temporal-only and spatial-temporal filters to achieve user desirable viewing experience.

Abstract translation: 提供了基于历史的时间运动噪声滤波技术的实现，其考虑多个图像之间的时间平滑以及基于块的技术来估计噪声/运动历史以更好地反映局部邻域中的空间/时间平滑度。 特别地，基于块的噪声测量中的每像素混合技术的实现可以用于更好地操纵空间和时域的像素。 可以使用全局噪声检测技术来估计噪声的发生和/或强度。 可以使用基于局部边缘测量的内容自适应空间滤波内容自适应空间滤波器来降低图像噪声以及保持边缘清晰度。 实现可以被配置用于各种应用。 特别地，可编程选项允许用户指定用于奇异性检测的过滤器参数，仅空间，仅时间和空间 - 时间过滤器，以实现用户期望的观看体验。

公开(公告)号：US07548660B2

公开(公告)日：2009-06-16

申请号：US11239965

申请日：2005-09-30

Applicant: Yi-Jen Chiu

Inventor： Yi-Jen Chiu

IPC: G06K9/40

CPC classification number: H04N19/86 ,  H04N19/117 ,  H04N19/14 ,  H04N19/61

Abstract: A method can include selecting a block of pixels. It may be determined whether the block of pixels contains an edge pixel. If the block of pixels contains an edge pixel, a first pixel may be selected among the block of pixels. If it is determined that the first pixel is a ringing noise pixel, a ringing filter may be applied. An edge-preserved filter may be applied if the first pixel is not a ringing noise pixel.

Abstract translation: 一种方法可以包括选择一个像素块。 可以确定像素块是否包含边缘像素。 如果像素块包含边缘像素，则可以在像素块中选择第一像素。 如果确定第一像素是振铃噪声像素，则可以施加振铃滤波器。 如果第一像素不是振铃噪声像素，则可以应用边缘保留滤光器。

公开(公告)号：US20090097763A1

公开(公告)日：2009-04-16

申请号：US11974688

申请日：2007-10-15

Applicant: Yi-Jen Chiu

Inventor： Yi-Jen Chiu

IPC: G06K9/36

CPC classification number: H04N5/142 ,  H04N1/64 ,  H04N5/20 ,  H04N7/0135 ,  H04N7/125 ,  H04N19/105 ,  H04N19/14 ,  H04N19/184 ,  H04N19/593 ,  H04N19/85

Abstract: A technique includes converting a first value for a pixel that is associated with a lower bit depth into a second value for the pixel, which is associated with a higher bit depth based at least in part on a neighborhood of the pixel.

Abstract translation: 一种技术包括至少部分地基于像素的邻域将与较低位深度相关联的像素的第一值转换为像素的第二值，该第二值与较高位深度相关联。

公开(公告)号：US20080239153A1

公开(公告)日：2008-10-02

申请号：US11693453

申请日：2007-03-29

Applicant: Yi-Jen Chiu

Inventor： Yi-Jen Chiu

IPC: H04N5/00

CPC classification number: H04N17/00 ,  G06T5/002 ,  G06T5/20 ,  G06T7/13 ,  G06T2207/10016 ,  G06T2207/20008 ,  G06T2207/20032 ,  H04N5/142 ,  H04N5/21

Abstract: A method of filtering noise from a picture may include determining a set of pixel noise metrics for a set of selected pixels in the picture, based solely on information from the set of selected pixels in the picture. The method may also designate as valid a subset of pixel noise metrics in the set of pixel noise metrics those associated pixels are not located at an edge or are not located in a complicated area of the picture. A set of block noise metrics may be calculated from the valid subset of pixel noise metrics, and a global noise metric for the picture may be ascertained from the set of block noise metrics. The picture may be filtered using the global noise metric to generate a filtered picture.

Abstract translation: 一种滤除图像噪声的方法可以包括基于图像中所选择的像素集合中的信息确定图像中的一组选定像素的像素噪声度量集合。 该方法还可以将像素噪声度量集合中的像素噪声度量的子集指定为相关像素不位于边缘处或者不位于图像的复杂区域中的子集。 可以从像素噪声度量的有效子集计算一组块噪声度量，并且可以从块噪声度量的集合中确定图像的全局噪声度量。 可以使用全局噪声度量来对图像进行滤波以产生滤波图像。

公开(公告)号：US07043088B2

公开(公告)日：2006-05-09

申请号：US10910027

申请日：2004-08-03

Applicant: Yi-Jen Chiu ,  Rajesh Hingorani ,  Obed Duardo ,  Paul Albin Wilford

Inventor： Yi-Jen Chiu ,  Rajesh Hingorani ,  Obed Duardo ,  Paul Albin Wilford

IPC: G06K9/36

CPC classification number: H03M7/42 ,  H04N19/91

Abstract: A method is disclosed for decoding multiple-coded symbols from a coded input symbol stream in a single clock cycle. The method constructs an original Huffman look-up table by extending the associated Huffman tree to decode multiple symbols in one clock cycle in a first embodiment and decodes multiple DCT coefficient symbols in an alternate embodiment. An advantage of the method is that the depth of the new Huffman tree is adjustable thereby making the method easily adaptable to various hardware architectures. A further advantage of the present invention is that the decoding process speed is significantly increased while the size of the lookup table is nominally increased.

Abstract translation: 公开了一种用于在单个时钟周期中从编码输入符号流解码多重编码符号的方法。 该方法通过在第一实施例中在一个时钟周期内扩展相关联的霍夫曼树来解码多个符号来构造原始霍夫曼查找表，并且在替代实施例中解码多个DCT系数符号。 该方法的一个优点是新的霍夫曼树的深度是可调的，从而使该方法容易适应各种硬件架构。 本发明的另一个优点是，在查询表的大小名义上增加时，解码处理速度显着增加。

公开(公告)号：US06771824B1

公开(公告)日：2004-08-03

申请号：US09473809

申请日：1999-12-28

Applicant: Yi-Jen Chiu ,  Rajesh Hingorani ,  Obed Duardo ,  Paul Albin Wilford

Inventor： Yi-Jen Chiu ,  Rajesh Hingorani ,  Obed Duardo ,  Paul Albin Wilford

IPC: G06K936

CPC classification number: H03M7/42 ,  H04N19/91

Abstract: A method is disclosed for decoding multiple-coded symbols from a coded input symbol stream in a single clock cycle. The method constructs an original Huffman look-up table by extending the associated Huffman tree to decode multiple symbols in one clock cycle in a first embodiment and decodes multiple DCT coefficient symbols in an alternate embodiment. An advantage of the method is that the depth of the new Huffman tree is adjustable thereby making the method easily adaptable to various hardware architectures. A further advantage of the present invention is that the decoding process speed is significantly increased while the size of the look-up table is nominally increased.

Abstract translation: 公开了一种用于在单个时钟周期中从编码输入符号流解码多重编码符号的方法。 该方法通过在第一实施例中在一个时钟周期内扩展相关联的霍夫曼树来解码多个符号来构造原始霍夫曼查找表，并且在替代实施例中解码多个DCT系数符号。 该方法的一个优点是新的霍夫曼树的深度是可调的，从而使该方法容易适应各种硬件架构。 本发明的另一个优点是，在查询表的大小名义上增加时，解码处理速度显着增加。