公开(公告)号：US07940993B2

公开(公告)日：2011-05-10

申请号：US11722141

申请日：2005-12-21

申请人： Takahiro Nagano ,  Tetsujiro Kondo ,  Tsutomu Watanabe ,  Junichi Ishibashi ,  Hisakazu Shiraki ,  Naoki Fujiwara ,  Masanori Kanemaru ,  Shinichiro Kaneko ,  Yasuhiro Suto

发明人： Takahiro Nagano ,  Tetsujiro Kondo ,  Tsutomu Watanabe ,  Junichi Ishibashi ,  Hisakazu Shiraki ,  Naoki Fujiwara ,  Masanori Kanemaru ,  Shinichiro Kaneko ,  Yasuhiro Suto

IPC分类号： G06K9/40 ,  G06K9/00 ,  H04N5/228

CPC分类号： H04N5/21 ,  H04N5/142 ,  H04N5/144

摘要： A motion-setting section (61) sets a motion amount and a motion direction for obtaining processing coefficients. A student-image-generating section (62) generates student images obtained by adding a motion blur to a teacher image not only based on the set motion amount and the set motion direction but also by changing at least one of the motion amount and motion direction in a specific ratio and student images obtained by adding no motion blur to the teacher image. A prediction-tap-extracting section (64) extracts, in order to extract a main term that mainly contains component of the target pixel, at least a pixel value of pixel in the student image whose space position roughly agrees with space position of the target pixel in the teacher image. A processing-coefficient-generating section (65) generates processing coefficients for predicting the target pixels in the teacher images from the pixel values of extracted pixels based on a relationship between the pixels thus extracted and the target pixels in the teacher images. The processing coefficients that are suitable for any motion blur removing which is robust against any shift of the motion vector can be generated through learning.

摘要翻译： 运动设定部（61）设定用于获得处理系数的运动量和运动方向。 学生图像生成部（62）生成不仅基于设定的运动量和设定的运动方向而将运动模糊添加到教师图像而获得的学生图像，而且通过改变运动量和运动方向中的至少一个 以特定比例和学生图像通过添加没有运动模糊获得的教师图像。 预测抽头部（64）为了提取主要包含目标像素的成分的主项，提取学生图像中的空间位置大致与目标的空间位置一致的像素的像素值 像素在老师的形象。 处理系数生成部（65）基于所提取的像素与教师图像中的目标像素之间的关系，从提取的像素的像素值生成用于预测教师图像中的目标像素的处理系数。 可以通过学习来产生适用于对运动矢量的任何移动而鲁棒的任何运动模糊去除的处理系数。

公开(公告)号：US20080019447A1

公开(公告)日：2008-01-24

申请号：US11733945

申请日：2007-04-11

申请人： Tetsujiro Kondo ,  Tsutomu Watanabe

发明人： Tetsujiro Kondo ,  Tsutomu Watanabe

IPC分类号： H04N11/02 ,  H04N7/12

CPC分类号： H04N19/53 ,  H04N19/61 ,  H04N19/63

摘要： A motion vector detecting apparatus includes a first generation unit that generates a first evaluation value table from an image signal of a first level, a second generation unit that generates a second evaluation value table from an image signal of a second level higher than the first level, a correction unit that corrects the first evaluation value table on the basis of the second evaluation value table, a first extraction unit that extracts first candidate vectors using the first evaluation value table, a second extraction unit that extracts second candidate vectors using the corrected first evaluation value table, and a detection unit that detects a motion vector from the first and second candidate vectors.

摘要翻译： 运动矢量检测装置包括：第一产生单元，其从第一电平的图像信号生成第一评估值表;第二生成单元，从第二电平高于第一电平的图像信号生成第二评估值表; 基于第二评估值表校正第一评估值表的校正单元，使用第一评估值表提取第一候选向量的第一提取单元，使用校正后的第一候选向量来提取第二候选向量的第二提取单元 评估值表，以及检测单元，其从第一和第二候选向量检测运动矢量。

公开(公告)号：US07002587B2

公开(公告)日：2006-02-21

申请号：US10640380

申请日：2003-08-14

申请人： Tetsujiro Kondo ,  Hideo Nakaya ,  Kazutaka Ando ,  Shizuo Chikaoka ,  Tsutomu Watanabe ,  Tsutomu Ichikawa ,  Yasunobu Node ,  Yoshiaki Nakamura

发明人： Tetsujiro Kondo ,  Hideo Nakaya ,  Kazutaka Ando ,  Shizuo Chikaoka ,  Tsutomu Watanabe ,  Tsutomu Ichikawa ,  Yasunobu Node ,  Yoshiaki Nakamura

IPC分类号： G09G5/39

CPC分类号： H04N19/53 ,  H04N19/433 ,  H04N19/46 ,  H04N19/513 ,  H04N19/56

摘要： A semiconductor device and an image data processing apparatus are capable of easily realizing address control, simplifying a circuit for address control, and easily realizing access to image data while reducing the volumes of pieces of image data of different resolutions in different hierarchies. In the image data processing apparatus, a memory cell array stores pieces of composite data, each containing a piece of pixel data and additional information data, for example, a motion vector associated with the pixel data. A designated pixel data and a motion vector associated with the pixel data are read out and a search region is suitably predicted and set up on the basis of the motion vector, and motion estimation is performed in the search region, for example, by a block matching process. Accordingly, motion estimation can be performed in a small search region equivalent to a large search region having been generally used for motion estimation, to realize high-speed, high-efficient motion estimation and improve the estimation accuracy.

摘要翻译： 半导体器件和图像数据处理装置能够容易地实现地址控制，简化用于地址控制的电路，并且能够容易地实现对图像数据的访问，同时减少不同等级的不同分辨率的图像数据的数量。 在图像数据处理装置中，存储单元阵列存储各自包含像素数据和附加信息数据的复合数据，例如与像素数据相关联的运动矢量。 读出与像素数据相关联的指定像素数据和运动矢量，并且基于运动矢量适当地预测和建立搜索区域，并且在搜索区域例如通过块来执行运动估计 匹配过程。 因此，可以在与通常用于运动估计的大搜索区域相当的小搜索区域中执行运动估计，以实现高速，高效的运动估计并提高估计精度。

公开(公告)号：US06483545B1

公开(公告)日：2002-11-19

申请号：US09358272

申请日：1999-07-21

申请人： Tetsujiro Kondo ,  Yasushi Tatehira ,  Nobuyuki Asakura ,  Masashi Uchida ,  Takuo Morimura ,  Kazutaka Ando ,  Hideo Nakaya ,  Tsutomu Watanabe ,  Satoshi Inoue ,  Wataru Niitsuma

发明人： Tetsujiro Kondo ,  Yasushi Tatehira ,  Nobuyuki Asakura ,  Masashi Uchida ,  Takuo Morimura ,  Kazutaka Ando ,  Hideo Nakaya ,  Tsutomu Watanabe ,  Satoshi Inoue ,  Wataru Niitsuma

IPC分类号： H04N701

CPC分类号： H04N7/012 ,  H04N7/0125 ,  H04N7/0137 ,  H04N7/0145

摘要： A 525i (interlace) signal can be converted into a 1050i signal or a 525p (progressive) signal. A calculating circuit generates pixels of an output picture signal with a linear estimation expression of predictive taps received from a tap selecting circuit and a coefficient received from a coefficient memory. The coefficient memory stores coefficients pre-obtained for individual classes. A class is determined by combining a spatial class corresponding to spatial class taps received from a tap selecting circuit and motion class taps received from a tap selecting circuit. A line sequential converting circuit converts a scanning line structure of an output signal of the calculating circuit 34 and obtains an output picture signal. The output picture signal is designated with a conversion method selection signal. Information corresponding to the selection signal is loaded from an information memory bank to the coefficient memory and registers.

公开(公告)号：US08391626B2

公开(公告)日：2013-03-05

申请号：US12331930

申请日：2008-12-10

申请人： Noriaki Takahashi ,  Hisakazu Shiraki ,  Takahiro Nagano ,  Yasuhiro Suto ,  Kenji Takahashi ,  Tsutomu Watanabe ,  Tetsujiro Kondo

发明人： Noriaki Takahashi ,  Hisakazu Shiraki ,  Takahiro Nagano ,  Yasuhiro Suto ,  Kenji Takahashi ,  Tsutomu Watanabe ,  Tetsujiro Kondo

IPC分类号： G06K9/40 ,  G06K9/62

CPC分类号： G06T5/003 ,  G06T2207/20081

摘要： A coefficient learning apparatus includes: a student-image generation section configured to generate a student image from the teacher image; a class classification section configured to sequentially set each of pixels in the teacher image as a pixel of interest and generate a class for the pixel of interest from the values of a plurality of specific pixels; a weight computation section configured to add up feature quantities; and a processing-coefficient generation section configured to generate a prediction coefficient on the basis of a determinant including said deterioration equation and a weighted constraint condition equation.

摘要翻译： 系数学习装置包括：学生图像生成部，被配置为从教师图像生成学生图像; 类分类部分，被配置为顺序地将教师图像中的每个像素设置为感兴趣的像素，并从多个特定像素的值生成关注像素的类; 权重计算部，被配置为相加特征量; 以及处理系数生成部，被配置为基于包括所述劣化方程和加权约束条件方程的行列式生成预测系数。

公开(公告)号：US08213496B2

公开(公告)日：2012-07-03

申请号：US11722460

申请日：2005-12-21

申请人： Takahiro Nagano ,  Tetsujiro Kondo ,  Tsutomu Watanabe ,  Junichi Ishibashi ,  Hisakazu Shiraki ,  Naoki Fujiwara ,  Masanori Kanemaru ,  Shinichiro Kaneko ,  Yasuhiro Suto

发明人： Takahiro Nagano ,  Tetsujiro Kondo ,  Tsutomu Watanabe ,  Junichi Ishibashi ,  Hisakazu Shiraki ,  Naoki Fujiwara ,  Masanori Kanemaru ,  Shinichiro Kaneko ,  Yasuhiro Suto

IPC分类号： H04N7/26 ,  G06K9/00 ,  G06K9/62 ,  H04N5/14

CPC分类号： G06T5/50 ,  G06T5/003 ,  G06T7/215 ,  G06T2207/20081 ,  G06T2207/20201 ,  H04N5/145 ,  H04N5/21

摘要： A motion-vector-setting section (31) sets a motion vector in units of pixel in a target image. Based on the motion vector, a target-pixel-setting section (35) sets a target pixel for each image in plural images to be processed. A motion-blur-amount-setting section (33) sets a motion blur amount in units of pixel based on the motion vector and the exposure-time ratio set in units of image in the exposure-time-ratio-setting section (32). A processing-region-setting section (36) sets processing regions corresponding to the target pixel for each of the plural images based on the motion blur amount. A processing-coefficient-setting section (37) sets processing coefficients based on the motion blur amount. A pixel-value-generating section (38) generates motion-blur-removed pixel values that correspond to the target pixel by linear combination of pixel values corresponding to pixels in the processing region and the processing coefficients, so that they can be output from an integration section (39) as one pixel value. By utilizing any time-directional information significantly, motion-blur-removing processing can be accurately performed.

摘要翻译： 运动矢量设定部（31）以目标图像中的像素为单位设定运动矢量。 基于运动矢量，目标像素设定部（35）在多个待处理图像中设定各图像的目标像素。 运动模糊量设定部（33）基于在曝光时间比设定部（32）中以图像为单位设定的运动矢量和曝光时间比设定像素单位的运动模糊量， 。 处理区域设置部分（36）基于运动模糊量设置与多个图像中的每一个对应的目标像素的处理区域。 处理系数设定部（37）基于运动模糊量设定处理系数。 像素值生成部（38）通过与处理区域中的像素对应的像素值和处理系数的线性组合，生成与目标像素对应的运动模糊去除像素值，从而可以从 积分部分（39）作为一个像素值。 通过显着地利用任何时间方向信息，可以准确地执行移动模糊去除处理。

公开(公告)号：US20090161948A1

公开(公告)日：2009-06-25

申请号：US12331930

申请日：2008-12-10

申请人： Noriaki TAKAHASHI ,  Hisakazu Shiraki ,  Takahiro Nagano ,  Yasuhiro Suto ,  Kenji Takahashi ,  Tsutomu Watanabe ,  Tetsujiro Kondo

发明人： Noriaki TAKAHASHI ,  Hisakazu Shiraki ,  Takahiro Nagano ,  Yasuhiro Suto ,  Kenji Takahashi ,  Tsutomu Watanabe ,  Tetsujiro Kondo

IPC分类号： G06K9/62

CPC分类号： G06T5/003 ,  G06T2207/20081

摘要： A coefficient learning apparatus includes: a student-image generation section configured to generate a student image from the teacher image; a class classification section configured to sequentially set each of pixels in the teacher image as a pixel of interest and generate a class for the pixel of interest from the values of a plurality of specific pixels; a weight computation section configured to add up feature quantities; and a processing-coefficient generation section configured to generate a prediction coefficient on the basis of a determinant including said deterioration equation and a weighted constraint condition equation.

摘要翻译： 系数学习装置包括：学生图像生成部，被配置为从教师图像生成学生图像; 类分类部分，被配置为顺序地将教师图像中的每个像素设置为感兴趣的像素，并从多个特定像素的值生成关注像素的类; 权重计算部，被配置为相加特征量; 以及处理系数生成部，被配置为基于包括所述劣化方程和加权约束条件方程的行列式生成预测系数。

公开(公告)号：US20080259169A1

公开(公告)日：2008-10-23

申请号：US11722436

申请日：2005-12-21

申请人： Takahiro Nagano ,  Tetsujiro Kondo ,  Tsutomu Watanabe ,  Junichi Ishibashi ,  Hisakazu Shiraki ,  Naoki Fujiwara ,  Masanori Kanemaru ,  Shinichiro Kaneko ,  Yasuhiro Suto

发明人： Takahiro Nagano ,  Tetsujiro Kondo ,  Tsutomu Watanabe ,  Junichi Ishibashi ,  Hisakazu Shiraki ,  Naoki Fujiwara ,  Masanori Kanemaru ,  Shinichiro Kaneko ,  Yasuhiro Suto

IPC分类号： H04N5/228 ,  H04N5/14

CPC分类号： G06T5/002 ,  G06T5/003 ,  G06T2207/20081 ,  G06T2207/20201 ,  H04N5/145 ,  H04N5/213 ,  H04N5/23248 ,  H04N5/23254 ,  H04N5/2327

摘要： A motion-vector-setting section (31) sets a first motion vector in units of pixel in a target image. An exposure-time-ratio-setting section (32) sets in units of image an exposure time ratio that is a ratio between a time interval of the target image and a period of exposure time. A motion-blur-amount-setting section (33) sets a motion blur amount in units of pixel based on the exposure time ratio and the first motion vector. Based on the motion amount, a processing-region-setting section (36) sets processing regions as well as a processing-coefficient-setting section (37) sets processing coefficients. A pixel-value-generating section (38) generates pixel values that correspond to the target pixel from pixel values in the processing region and the processing coefficients. A motion-blur-adding section (41) adds a motion blur to an image containing the pixel value generated based on an input second motion vector and the first motion vector. An image-moving section (42) moves the motion-blur-added image along a counter vector of the second motion vector. Any more real arbitrary viewpoint image can be generated.

摘要翻译： 运动矢量设定部（31）以目标图像中的像素为单位设定第一运动矢量。 曝光时间比设定部（32）以图像为单位设定作为目标图像的时间间隔与曝光时间的周期之间的比率的曝光时间比。 运动模糊量设定部（33）基于曝光时间比和第一运动矢量，以像素为单位设定运动模糊量。 基于运动量，处理区域设置部分（36）设置处理区域以及处理系数设置部分（37）设置处理系数。 像素值生成部（38）根据处理区域和处理系数的像素值生成与目标像素对应的像素值。 运动模糊添加部分（41）将运动模糊添加到包含基于输入的第二运动矢量和第一运动矢量生成的像素值的图像。 图像移动部分（42）沿着第二运动矢量的反向量移动运动模糊添加图像。 可以生成任何更真实的任意视点图像。

公开(公告)号：US07412384B2

公开(公告)日：2008-08-12

申请号：US10089430

申请日：2001-07-31

申请人： Tetsujiro Kondo ,  Tsutomu Watanabe

发明人： Tetsujiro Kondo ,  Tsutomu Watanabe

IPC分类号： G10L15/08

CPC分类号： G10L19/22 ,  G10L19/18 ,  G10L25/06

摘要： A digital signal processing method and learning method and devices therefor, and a program storage medium which are capable of further improving the waveform reproducibility of a digital signal. Self correlation coefficients are calculated by cutting parts out of the digital signal by multiple windows having different sizes, and the parts are classified based on the calculation results of the self correlation coefficients. Then, the digital signal is converted by the prediction method corresponding to the classified class, so that the conversion further suitable for the features of the digital signal can be conducted.

摘要翻译： 一种数字信号处理方法及其学习方法和装置，以及能够进一步提高数字信号的波形再现性的程序存储介质。 通过利用具有不同尺寸的多个窗口从数字信号中切出零件来计算自相关系数，并且根据自相关系数的计算结果对这些部分进行分类。 然后，通过对应于分类的类别的预测方法来转换数字信号，使得可以进行进一步适合于数字信号的特征的转换。

公开(公告)号：US07397512B2

公开(公告)日：2008-07-08

申请号：US11056141

申请日：2005-02-10

申请人： Tetsujiro Kondo ,  Yasushi Tatehira ,  Nobuyuki Asakura ,  Masashi Uchida ,  Takuo Morimura ,  Kazutaka Ando ,  Hideo Nakaya ,  Tsutomu Watanabe ,  Satoshi Inoue ,  Wataru Niitsuma

发明人： Tetsujiro Kondo ,  Yasushi Tatehira ,  Nobuyuki Asakura ,  Masashi Uchida ,  Takuo Morimura ,  Kazutaka Ando ,  Hideo Nakaya ,  Tsutomu Watanabe ,  Satoshi Inoue ,  Wataru Niitsuma

IPC分类号： H04N5/45

CPC分类号： H04N7/012 ,  H04N7/0125 ,  H04N7/0137 ,  H04N7/0145

摘要： A picture information converting apparatus for generating a plurality of output picture signals. The apparatus includes first data selecting means; second data selecting means; memory means for storing a pre-obtained estimation expression coefficient; pixel value generating means for generating the first pixel with the second pixels selected by the first data selecting means and a linear estimation expression of the estimation expression coefficient; class determining means for generating class information corresponding to the third pixels selected by the second data selecting means and supplying the estimation expression coefficient to the pixel value generating means corresponding to the class information; scanning line structure converting means, connected to the pixel value generating means, for converting a conversion picture to a designated scanning line structure; and controlling means for receiving a control signal that designates a scanning line structure of the output picture signal and for switching over at least one of the estimation expression coefficient, the second pixels selected by the first data selecting means and the third pixels selected by the second data selecting means, corresponding to the control signal.

摘要翻译： 一种用于产生多个输出图像信号的图像信息转换装置。 该装置包括第一数据选择装置; 第二数据选择装置; 用于存储预先获得的估计表达系数的存储装置; 像素值产生装置，用于产生具有由第一数据选择装置选择的第二像素的第一像素和估计表达式系数的线性估计表达式; 类别确定装置，用于产生与由第二数据选择装置选择的第三像素相对应的类别信息，并将估计表达式系数提供给与类信息相对应的像素值产生装置; 扫描线结构转换装置，连接到像素值产生装置，用于将转换图像转换成指定的扫描线结构; 以及控制装置，用于接收指定输出图像信号的扫描线结构的控制信号，并且用于切换估计表达式系数中的至少一个，由第一数据选择装置选择的第二像素和由第二数据选择装置选择的第三像素 数据选择装置，对应于控制信号。