公开(公告)号：US06480544B1

公开(公告)日：2002-11-12

申请号：US09513586

申请日：2000-02-25

申请人： Kenji Uehara ,  Yoshihiro Murakami ,  Shigeto Funado ,  Osamu Matsunaga ,  Shigeo Fujishiro

发明人： Kenji Uehara ,  Yoshihiro Murakami ,  Shigeto Funado ,  Osamu Matsunaga ,  Shigeo Fujishiro

IPC分类号： H04B166

CPC分类号： H04N19/59 ,  H04N19/105 ,  H04N19/117 ,  H04N19/124 ,  H04N19/139 ,  H04N19/149 ,  H04N19/152 ,  H04N19/172 ,  H04N19/176 ,  H04N19/182 ,  H04N19/50 ,  H04N19/51 ,  H04N19/61 ,  H04N19/80

摘要： In an encoding apparatus for processing a video signal with a signal processing circuit, producing coefficient data by an orthogonal transform of differential data obtained by motion compensation, generating coded data by quantization of the coefficient data and converting the quantization step size at the time of the quantization according to the amount of generated data of the coded data, the signal processing circuit includes: a motion vector detecting circuit for detecting a motion vector of the video signal; a motion code vector amount detecting circuit for calculating the code amount of the motion vector on the basis of the motion vector and for detecting the amount of code of the coefficient data; and a pixel number converting circuit for converting the number of pixels of the video signal according to the amount of code of the motion vector and the amount of code of the coefficient data.

摘要翻译： 在用信号处理电路处理视频信号的编码装置中，通过运动补偿获得的差分数据的正交变换来生成系数数据，通过量化系数数据来生成编码数据，并且在 信号处理电路根据编码数据的生成数据量进行量化，包括：运动矢量检测电路，用于检测视频信号的运动矢量; 运动码矢量检测电路，用于根据运动矢量计算运动矢量的码量，并检测系数数据的码量; 以及像素数转换电路，用于根据运动矢量的代码量和系数数据的代码量来转换视频信号的像素数。

公开(公告)号：US06269123B1

公开(公告)日：2001-07-31

申请号：US09247432

申请日：1999-02-10

申请人： Shigeo Fujishiro ,  Osamu Matsunaga

发明人： Shigeo Fujishiro ,  Osamu Matsunaga

IPC分类号： H04B166

CPC分类号： H04N19/59 ,  H04N19/117 ,  H04N19/124 ,  H04N19/132 ,  H04N19/139 ,  H04N19/176 ,  H04N19/61 ,  H04N19/80 ,  H04N19/85

摘要： A video encoder for reducing the deterioration of picture quality of a compress coded video signal by implementing a method for deciding the number of picture elements needed to properly encode an inputted video signal. The decision is calculated based on the difficulty of compressing the inputted video signal and the number of picture elements needed to accurately reproduce a compressed inputted video signal. The decision circuitry decides the number of picture elements needed for compression of the inputted video signal allowing the number of codes generated in the encoding process to be decreased and quantization accuracy improved. The main advantage is a reduction in deterioration of the video signal usually found when reproducing a compress coded video signal.

摘要翻译： 一种视频编码器，用于通过实现用于确定对输入的视频信号进行适当编码所需的像素数量的方法来减少压缩编码视频信号的图像质量的劣化。 基于压缩输入的视频信号的难度和精确地再现压缩的输入视频信号所需的像素的数量来计算该决定。 决定电路决定压缩输入视频信号所需的像素数量，从而允许降低编码处理中生成的代码数量，提高量化精度。 主要的优点是在再现压缩编码视频信号时通常发现的视频信号的劣化减少。

公开(公告)号：US07529432B2

公开(公告)日：2009-05-05

申请号：US10970292

申请日：2004-10-21

申请人： Shigeo Fujishiro ,  Ikuo Someya

发明人： Shigeo Fujishiro ,  Ikuo Someya

IPC分类号： G06K9/32 ,  G06K11/00 ,  H04N1/46

CPC分类号： G06T3/4007

摘要： A data processing apparatus for converting discrete original image data into interpolated image data having a different number of sampling data from the original image data, including generating means for generating a plurality of the interpolated image data from the original image data, first virtualizing means for generating first virtual data including a virtual succession of the original image data, second virtualizing means for generating second virtual data including a virtual succession of each of the plural interpolated image data, error calculating means for calculating error between each of the second virtual data with respect to the plural interpolated image data and the first virtual data, and selecting means for selecting either one of the plural interpolated image data based on the error.

摘要翻译： 一种数据处理装置，用于将离散的原始图像数据转换为具有与原始图像数据不同数量的采样数据的内插图像数据，包括用于从原始图像数据生成多个内插图像数据的生成装置，用于产生 第一虚拟数据，包括原始图像数据的虚拟连续;第二虚拟化装置，用于生成包括多个内插图像数据中的每一个的虚拟连续的第二虚拟数据;误差计算装置，用于计算每个第二虚拟数据之间相对于 所述多个内插图像数据和所述第一虚拟数据，以及选择装置，用于基于所述错误来选择所述多个内插图像数据中的一个。

公开(公告)号：US5528313A

公开(公告)日：1996-06-18

申请号：US994659

申请日：1992-12-22

申请人： Yutaka Tanaka ,  Shigeo Fujishiro

发明人： Yutaka Tanaka ,  Shigeo Fujishiro

IPC分类号： H04N19/50 ,  G06T7/20 ,  H04N5/14 ,  H04N19/423 ,  H04N19/51 ,  H04N19/543 ,  H04N19/587 ,  H04N19/59 ,  H04N19/80 ,  H04N19/85 ,  H04N7/24

CPC分类号： H04N5/144

摘要： A motion detection signal delivered from a mixer is output through a series circuit of a maximum value calculation portion and a minimum value calculation portion. In each of the calculation portions, calculation is performed within a block of 3-dot.times.3-line. Missing of motion detection is prevented from occurring in the maximum value calculation portion because the motion detection signal of the pixel under attention is calculated therein with the values within the block and, accordingly, the motion detection signal is expanded in the horizontal and vertical directions. A detection spread to a still picture portion occurring in the calculation portion is suppressed in the minimum value calculation portion because, when surrounding motion detection signals therein are lower than the motion detection signal of the pixel under attention, the motion detection signal of the pixel under attention is replaced with the lower value. With the described arrangement, the missing of motion detection is advantageously prevented without inviting the detection spread to a still picture portion.

摘要翻译： 从混频器发送的运动检测信号通过最大值计算部分和最小值计算部分的串联电路输出。 在每个计算部分中，在3-dotx3行的块内进行计算。 由于在其中计算了被注意的像素的运动检测信号，所以在块中计算出运动检测的缺失，因此在水平和垂直方向上扩大运动检测信号。 在最小值计算部分中抑制扩展到计算部分中出现的静止图像部分的检测，因为当其中周围的运动检测信号低于被注意的像素的运动检测信号时，该像素的运动检测信号 注意力用较低的值代替。 利用所述的布置，有利地防止了运动检测的缺失，而不会将检测扩展到静止图像部分。

公开(公告)号：US07224846B2

公开(公告)日：2007-05-29

申请号：US10625520

申请日：2003-07-24

申请人： Shigeo Fujishiro ,  Takuya Kitamura

发明人： Shigeo Fujishiro ,  Takuya Kitamura

IPC分类号： G06K9/36

CPC分类号： H05K3/361 ,  H05K3/305 ,  H05K3/323 ,  H05K2201/10977

摘要： An bit-plane coding pass generator is provided in which for SP-pass processing quantization coefficient of each code block divided into bit planes, “significant (S)” or “non-significant (N)” data in a predetermined area and those around the area and compared with an S/N matching pattern. The S/N matching pattern has been set when a jump can be made from an arbitrary sample point to a next sample point to be processed by SP pass. The jump is made to the next sample point to be processed by SP pass according to a jump address value obtained from a pattern coincident with a current S/N matching pattern. Thus, by reducing the time for the significant propagation (SP) pass defined in JPEG-2000, a code block can be coded at a higher speed by three coding passes.

摘要翻译： 提供了一种位平面编码遍发生器，其中针对每个代码块的SP-pass处理量化系数分为比特平面，在预定区域中的“有效（S）”或“非有效（N）”数据） 与S / N匹配模式进行比较。 当可以从任意采样点到下一个待通过SP处理的采样点进行跳变时，设置S / N匹配模式。 根据从与当前S / N匹配模式一致的图案获得的跳转地址值，对通过SP pass处理的下一个采样点进行跳转。 因此，通过缩短在JPEG-2000中定义的显着传播（SP）传递的时间，可以通过三次编码通过以更高的速度对代码块进行编码。

公开(公告)号：US20050144540A1

公开(公告)日：2005-06-30

申请号：US10970292

申请日：2004-10-21

申请人： Shigeo Fujishiro ,  Ikuo Someya

发明人： Shigeo Fujishiro ,  Ikuo Someya

IPC分类号： G06T3/40 ,  G06F11/00

CPC分类号： G06T3/4007

摘要： Disclosed herein is a data processing apparatus for converting discrete first data into second data having a different number of sampling data from the first data, including generating means for generating a plurality of the second data from the first data, first virtualizing means for generating first virtual data including a virtual succession of the first data, second virtualizing means for generating second virtual data including a virtual succession of each of the plural second data, error calculating means for calculating an error between each of the second virtual data with respect to the plural second data and the first virtual data, and selecting means for selecting either one of the plural second data based on the error.

摘要翻译： 这里公开了一种数据处理装置，用于将离散的第一数据转换成具有来自第一数据的不同数量的采样数据的第二数据，包括用于从第一数据产生多个第二数据的生成装置，用于产生第一虚拟 数据，包括第一数据的虚拟连续;第二虚拟化装置，用于生成包括多个第二数据中的每一个的虚拟连续的第二虚拟数据;误差计算装置，用于计算每个第二虚拟数据之间相对于多个第二数据的误差 数据和第一虚拟数据，以及选择装置，用于基于该错误来选择多个第二数据中的一个。