公开(公告)号：US08385602B2

公开(公告)日：2013-02-26

申请号：US12066091

申请日：2006-09-07

申请人： Akihiko Kaino ,  Takashi Horishi ,  Yukihiro Nakamura ,  Kunio Kawaguchi ,  Yasuaki Takahashi ,  Norifumi Yoshiwara ,  Takafumi Morifuji ,  Yuta Choki

发明人： Akihiko Kaino ,  Takashi Horishi ,  Yukihiro Nakamura ,  Kunio Kawaguchi ,  Yasuaki Takahashi ,  Norifumi Yoshiwara ,  Takafumi Morifuji ,  Yuta Choki

IPC分类号： G06K9/00 ,  G06K9/46 ,  G06K9/66 ,  G06K9/36 ,  H04N5/228 ,  H04N5/14 ,  H04N9/64

CPC分类号： H04N19/51 ,  G06T7/223

摘要： An image processing device and method, a program, and a recording medium whereby the detection precision of a motion vector by the gradient method is further improved. A counter value computing unit acquires the number of valid pixels, the number of pixels having no gradient in the horizontal direction, and the number of pixels having no gradient in the vertical direction from a valid pixel number counter, a no-horizontal-gradient counter, and a no-vertical-gradient counter, computes the ratio between valid pixels within a computation block and one-sided gradient pixels among the valid pixels, and controls the value of a flag which a flag setting unit sets in accordance with the computation results. The flag setting unit sets the value of a gradient flag, and outputs the gradient flag to the subsequent stage. At the subsequent stage, gradient method computation and vector evaluation processing are executed based on the gradient flag. The above can be applied to a signal processing device for performing frame frequency conversion processing from a 24P signal to a 60P signal.

摘要翻译： 一种图像处理装置和方法，程序和记录介质，由此通过梯度方法的运动矢量的检测精度进一步提高。 计数器值计算单元从有效像素数计数器获取有效像素的数量，在水平方向上没有梯度的像素的数量和在垂直方向上没有梯度的像素的数量，非水平梯度计数器 和无垂直梯度计数器，计算计算块中的有效像素与有效像素之间的单侧梯度像素之间的比率，并且根据计算结果来控制标志设置单元设置的标志的值 。 标志设置单元设置梯度标志的值，并将梯度标志输出到后级。 在后续阶段，基于梯度标志执行梯度法计算和矢量评估处理。 上述可以应用于用于从24P信号到60P信号进行帧频转换处理的信号处理装置。

公开(公告)号：US20130038703A1

公开(公告)日：2013-02-14

申请号：US13635030

申请日：2011-04-06

申请人： Suguru Ushiki ,  Masami Ogata ,  Takafumi Morifuji

发明人： Suguru Ushiki ,  Masami Ogata ,  Takafumi Morifuji

IPC分类号： H04N13/04 ,  H04N13/00

CPC分类号： H04N21/4884 ,  H04N13/128 ,  H04N13/139 ,  H04N13/172 ,  H04N13/178 ,  H04N13/183 ,  H04N13/189 ,  H04N21/4516

摘要： The present invention relates to data structures, image processing apparatuses, image processing methods, and programs for enabling appropriate processing of video data of 3D content.Obtained content data of 3D content contains video data of 3D images, and information about a reference image size, which is a predetermined image size serving as the reference for the 3D images, and the maximum and minimum parallaxes for the reference image size. An image size detecting unit (71) detects the image size of the video data of the 3D images. If the detected image size of the 3D video data is not the same as the “reference image size,” a maximum/minimum parallax calculating unit (72) calculates the maximum and minimum parallaxes corresponding to the image size of the 3D video data. A caption adjusting unit (73) performs predetermined image processing based on the calculated maximum and minimum parallaxes, such as processing to adjust the depth positions of captions. The present invention can be applied to an image processing apparatus that processes obtained image data of 3D images, for example.

摘要翻译： 本发明涉及数据结构，图像处理装置，图像处理方法以及用于实现3D内容的视频数据的适当处理的程序。 获得的3D内容的内容数据包含3D图像的视频数据，以及作为用于3D图像的基准的预定图像大小的参考图像尺寸以及参考图像尺寸的最大和最小视差的信息。 图像尺寸检测单元（71）检测3D图像的视频数据的图像尺寸。 如果检测到的3D视频数据的图像尺寸与参考图像尺寸不相同，则最大/最小视差计算单元（72）计算与3D视频数据的图像尺寸相对应的最大和最小视差。 字幕调整单元（73）基于所计算出的最大和最小视差进行预定的图像处理，例如用于调整字幕的深度位置的处理。 本发明可以应用于例如处理所获得的3D图像的图像数据的图像处理装置。

公开(公告)号：US20110254925A1

公开(公告)日：2011-10-20

申请号：US13082397

申请日：2011-04-07

申请人： Suguru USHIKI ,  Takafumi Morifuji ,  Masami Ogata

发明人： Suguru USHIKI ,  Takafumi Morifuji ,  Masami Ogata

IPC分类号： H04N13/02

CPC分类号： H04N13/398 ,  H04N13/128 ,  H04N13/139 ,  H04N21/431 ,  H04N21/440263 ,  H04N21/4516 ,  H04N21/816 ,  H04N21/84

摘要： An image processing apparatus includes a first acquiring unit, a second acquiring unit, and a correction processor. The first acquiring unit acquires an intended viewing environment parameter, which is a parameter of an intended viewing environment, together with image data of a three-dimensional picture. The second acquiring unit acquires an actual viewing environment parameter, which is a parameter of an actual viewing environment for a user viewing the three-dimensional picture. The correction processor corrects the three-dimensional picture in accordance with a difference between the acquired intended viewing environment parameter and the acquired actual viewing environment parameter.

摘要翻译： 图像处理装置包括第一获取单元，第二获取单元和校正处理器。 第一获取单元与三维图像的图像数据一起获取作为预期观看环境的参数的预期观看环境参数。 第二获取单元获取作为观看三维图像的用户的实际观看环境的参数的实际观看环境参数。 校正处理器根据所获取的预期观看环境参数与所获取的实际观看环境参数之间的差异来校正三维图片。

公开(公告)号：US07974345B2

公开(公告)日：2011-07-05

申请号：US10744043

申请日：2003-12-24

申请人： Tetsujiro Kondo ,  Tsutomu Watanabe ,  Takafumi Morifuji

发明人： Tetsujiro Kondo ,  Tsutomu Watanabe ,  Takafumi Morifuji

IPC分类号： H04N7/12 ,  H04N11/02

CPC分类号： H04N7/014 ,  H04N7/0112 ,  H04N7/0145 ,  H04N19/136 ,  H04N19/159 ,  H04N19/40 ,  H04N19/44 ,  H04N19/46 ,  H04N19/513 ,  H04N19/56 ,  H04N19/61

摘要： An image data processing apparatus is provided and includes a variable length decoding section to decode and extract quantized DCT coefficients from an encoded signal input thereto and extract encoded information from side information added to the encoded signal. Classification adaptation processing sections use the input signal supplied from a decoding section and a creation signal supplied from a signal storage section to determine the reliability of each of motion compensating vectors of the encoded signal supplied from a coded information storage section, selects that one of the motion compensating vectors which has the highest reliability as a selected vector, and produces a creation signal based on the selection vector.

摘要翻译： 提供了一种图像数据处理装置，包括一个可变长度解码部分，用于从输入的编码信号中解码和提取量化的DCT系数，并从附加到编码信号的副信息中提取编码信息。 分类适应处理部分使用从解码部分提供的输入信号和从信号存储部分提供的创建信号来确定从编码信息存储部分提供的编码信号的每个运动补偿矢量的可靠性， 作为所选择的向量具有最高可靠性的运动补偿向量，并且基于选择向量产生创建信号。

公开(公告)号：US07283678B2

公开(公告)日：2007-10-16

申请号：US11584772

申请日：2006-10-20

申请人： Tetsujiro Kondo ,  Toshihiko Hamamatsu ,  Hideo Nakaya ,  Takeharu Nishikata ,  Hideki Ohtsuka ,  Takeshi Kunihiro ,  Takafumi Morifuji ,  Masashi Uchida

发明人： Tetsujiro Kondo ,  Toshihiko Hamamatsu ,  Hideo Nakaya ,  Takeharu Nishikata ,  Hideki Ohtsuka ,  Takeshi Kunihiro ,  Takafumi Morifuji ,  Masashi Uchida

IPC分类号： G06K9/36 ,  G06K9/46 ,  G06K9/34

CPC分类号： H03M7/30 ,  H04N19/117 ,  H04N19/14 ,  H04N19/176 ,  H04N19/44 ,  H04N19/46 ,  H04N19/60 ,  H04N19/61 ,  H04N19/63 ,  H04N19/86 ,  H04N19/98

摘要： JPEG encoded data are entropy-decoded to quantized DCT coefficients which are sent to a prediction tap extraction circuit (41) and to a class tap extraction circuit (42). The prediction tap extraction circuit (41) and the class tap extraction circuit (42) extract what is needed from the quantized DCT coefficients to form prediction taps and class taps. A classification circuit (43) effects classification based on the class taps. A coefficient table storage unit (44) sends tap coefficients corresponding to the classes resulting from the classification to a sum of products circuit (45), which sum of products circuit (45) then effects linear predictive calculations, using the tap coefficients and the class taps, to generate decoded picture data.

公开(公告)号：US07206452B2

公开(公告)日：2007-04-17

申请号：US11348720

申请日：2006-02-07

申请人： Tetsujiro Kondo ,  Toshihiko Hamamatsu ,  Hideo Nakaya ,  Takeharu Nishikata ,  Hideki Ohtsuka ,  Takeshi Kunihiro ,  Takafumi Morifuji ,  Masashi Uchida

发明人： Tetsujiro Kondo ,  Toshihiko Hamamatsu ,  Hideo Nakaya ,  Takeharu Nishikata ,  Hideki Ohtsuka ,  Takeshi Kunihiro ,  Takafumi Morifuji ,  Masashi Uchida

IPC分类号： G06K9/36 ,  G06K9/46

CPC分类号： H03M7/30 ,  H04N19/117 ,  H04N19/14 ,  H04N19/176 ,  H04N19/44 ,  H04N19/46 ,  H04N19/60 ,  H04N19/61 ,  H04N19/63 ,  H04N19/86 ,  H04N19/98

摘要： JPEG encoded data are entropy-decoded to quantized DCT coefficients which are sent to a prediction tap extraction circuit (41) and to a class tap extraction circuit (42). The prediction tap extraction circuit (41) and the class tap extraction circuit (42) extract what is needed from the quantized DCT coefficients to form prediction taps and class taps. A classification circuit (43) effects classification based on the class taps. A coefficient table storage unit (44) sends tap coefficients corresponding to the classes resulting from the classification to a sum of products circuit (45), which sum of products circuit (45) then effects linear predictive calculations, using the tap coefficients and the class taps, to generate decoded picture data.

公开(公告)号：US20070058873A1

公开(公告)日：2007-03-15

申请号：US11584789

申请日：2006-10-20

申请人： Tetsujiro Kondo ,  Toshihiko Hamamatsu ,  Hideo Nakaya ,  Takeharu Nishikata ,  Hideki Ohtsuka ,  Takeshi Kunihiro ,  Takafumi Morifuji ,  Masashi Uchida

发明人： Tetsujiro Kondo ,  Toshihiko Hamamatsu ,  Hideo Nakaya ,  Takeharu Nishikata ,  Hideki Ohtsuka ,  Takeshi Kunihiro ,  Takafumi Morifuji ,  Masashi Uchida

IPC分类号： G06K9/36

CPC分类号： H03M7/30 ,  H04N19/117 ,  H04N19/14 ,  H04N19/176 ,  H04N19/44 ,  H04N19/46 ,  H04N19/60 ,  H04N19/61 ,  H04N19/63 ,  H04N19/86 ,  H04N19/98

摘要： JPEG encoded data are entropy-decoded to quantized DCT coefficients which are sent to a prediction tap extraction circuit (41) and to a class tap extraction circuit (42). The prediction tap extraction circuit (41) and the class tap extraction circuit (42) extract what is needed from the quantized DCT coefficients to form prediction taps and class taps. A classification circuit (43) effects classification based on the class taps. A coefficient table storage unit (44) sends tap coefficients corresponding to the classes resulting from the classification to a sum of products circuit (45), which sum of products circuit (45) then effects linear predictive calculations, using the tap coefficients and the class taps, to generate decoded picture data.

公开(公告)号：US07035471B2

公开(公告)日：2006-04-25

申请号：US10030591

申请日：2001-05-09

申请人： Tetsujiro Kondo ,  Toshihiko Hamamatsu ,  Hideo Nakaya ,  Takeharu Nishikata ,  Hideki Ohtsuka ,  Takeshi Kunihiro ,  Takafumi Morifuji ,  Masashi Uchida

发明人： Tetsujiro Kondo ,  Toshihiko Hamamatsu ,  Hideo Nakaya ,  Takeharu Nishikata ,  Hideki Ohtsuka ,  Takeshi Kunihiro ,  Takafumi Morifuji ,  Masashi Uchida

IPC分类号： G06K9/36 ,  G06K9/46

CPC分类号： H03M7/30 ,  H04N19/117 ,  H04N19/14 ,  H04N19/176 ,  H04N19/44 ,  H04N19/46 ,  H04N19/60 ,  H04N19/61 ,  H04N19/63 ,  H04N19/86 ,  H04N19/98

摘要： JPEG encoded data are entropy-decoded to quantized DCT coefficients which are sent to a prediction tap extraction circuit (41) and to a class tap extraction circuit (42). The prediction tap extraction circuit (41) and the class tap extraction circuit (42) extract what is needed from the quantized DCT coefficients to form prediction taps and class taps. A classification circuit (43) effects classification based on the class taps. A coefficient table storage unit (44) sends tap coefficients corresponding to the classes resulting from the classification to a sum of products circuit (45), which sum of products circuit (45) then effects linear predictive calculations, using the tap coefficients and the class taps, to generate decoded picture data.

摘要翻译： JPEG编码数据被熵解码为被发送到预测抽头提取电路（41）和类抽头提取电路（42）的量化DCT系数。 预测抽头提取电路（41）和分类提取电路（42）从量化的DCT系数中提取需要的数据，以形成预测抽头和类别抽头。 分类电路（43）根据类别抽头进行分类。 系数表存储单元（44）将从分类得到的类别对应的抽头系数发送到产品总和电路（45），乘积电路（45）之和利用抽头系数和类别来实现线性预测计算 抽头，以产生解码的图像数据。

公开(公告)号：US09094659B2

公开(公告)日：2015-07-28

申请号：US13126966

申请日：2010-08-19

申请人： Takafumi Morifuji ,  Suguru Ushiki ,  Masami Ogata

发明人： Takafumi Morifuji ,  Suguru Ushiki ,  Masami Ogata

IPC分类号： H04N15/00 ,  H04N13/00 ,  H04N13/04 ,  G02B27/22 ,  G06T7/00 ,  H04N5/92 ,  H04N5/89

CPC分类号： H04N13/128 ,  G02B27/2214 ,  G06T7/97 ,  H04N13/122 ,  H04N2013/0081

摘要： To reduce burden of a viewer at the time of performing stereoscopic display by a stereoscopic display device and to show a stereoscopic image in a natural and comfortable manner.A disparity estimating unit 210 estimates disparity from the left and right images of an input image to generate a disparity map. A disparity analyzing unit 230 analyzes the disparity map to generate a disparity control parameter for performing suitable disparity control. A disparity control unit 240 controls the content of processing at an image conversion unit 250 in accordance with the disparity control parameter. The image conversion unit 250 performs image conversion on the input image based on the control by the disparity control unit 240 to output an output image. The image conversion unit 250 performs image conversion of two steps. The image conversion unit 250 performs shift processing for shifting the relative positions of the left and right images of the input image in the horizontal direction as first step image conversion, for example. The image conversion unit 250 performs scaling processing for performing enlargement/reduction of the entire screen with each center of the left and right images as a reference, as second step image conversion, for example.

摘要翻译： 为了减轻立体显示装置的立体显示时的观看者的负担，以自然舒适的方式显示立体图像。 视差估计单元210从输入图像的左图像和右图像估计视差以生成视差图。 视差分析单元230分析视差图以生成用于执行适当视差控制的视差控制参数。 视差控制单元240根据视差控制参数控制图像转换单元250处理的内容。 图像转换单元250基于视差控制单元240的控制输出输出图像，对输入图像进行图像转换。 图像转换单元250执行两个步骤的图像转换。 图像转换单元250例如执行用于将输入图像的左右图像在水平方向上的相对位置偏移的移位处理作为第一步骤图像转换。 图像转换单元250执行缩放处理，用于以左右图像的每个中心为基准进行整个屏幕的放大/缩小，例如作为第二步图像转换。

公开(公告)号：US08565516B2

公开(公告)日：2013-10-22

申请号：US12871525

申请日：2010-08-30

申请人： Takafumi Morifuji ,  Kuniaki Takahashi ,  Don Eklund

发明人： Takafumi Morifuji ,  Kuniaki Takahashi ,  Don Eklund

IPC分类号： G06K9/00

CPC分类号： H04N13/128 ,  H04N13/10 ,  H04N13/156

摘要： An image processor includes a main image statistical information generator that detects a parallax of each predetermined unit of a 3D main image from main image data and generates parallax statistical information, a sub-image statistical information generator unit that detects a parallax of each predetermined unit of a 3D sub-image from sub-image data and generates parallax statistical information, a parallax controller that computes, using the statistical information, a correction amount used for correcting at least one of the main image and sub-image parallaxes so that a positional distance between the main image and the sub-image in a depth direction is within a predetermined range, a converter that converts at least one of the main image data and sub-image data so that at least one of the parallaxes of the images is corrected by the correction amount, and a superimposing unit that superimposes the sub-image data on the main image data.

摘要翻译： 图像处理器包括主图像统计信息生成器，其从主图像数据检测3D主图像的每个预定单位的视差，并生成视差统计信息;子图像统计信息生成单元，其检测每个预定单位的视差 来自子图像数据的3D子图像，并生成视差统计信息;视差控制器，其使用统计信息计算用于校正主图像和副图像视差中的至少一个的校正量，使得位置距离 在主图像和子图像之间的深度方向上处于预定范围内的转换器，转换器，其转换主图像数据和子图像数据中的至少一个，使得通过以下方式校正图像的至少一个视差： 校正量，以及将副图像数据叠加在主图像数据上的叠加单元。