公开(公告)号：EP1510857A4

公开(公告)日：2008-06-04

申请号：EP03733291

申请日：2003-06-05

Applicant: SONY CORP

Inventor： KONDO TETSUJIRO ,  OKUMURA YUJI ,  HAGA TSUGIHIKO ,  MURAKAMI TETSUYA ,  TANAKA KENJI

IPC: G03B21/00 ,  G02B27/06 ,  G02B27/18 ,  G02B27/22 ,  G03B21/14 ,  G03B21/60 ,  G03B35/20 ,  H04N5/74 ,  H04N7/18 ,  H04N13/04

CPC classification number: G02B27/06 ,  G02B27/2235 ,  G02B27/2285 ,  G03B21/28 ,  G03B21/562 ,  G03B21/606 ,  G03B21/62 ,  H04N9/3141 ,  H04N13/045

Abstract: The present invention relates to a display apparatus and a display method that enable multiple users to view high resolution moving images from various viewing points of the users. A light-control screen 25 is rotated, and projectors 211 to 21N irradiate the screen 25 from various directions with light rays corresponding to image data captured from those directions. In this case, when the forward direction of the light-control screen 25 is aligned with the projecting direction of one of the projectors 21n, an image corresponding to light rays emitted from the projector 21n is displayed on the light-control screen 25. The present invention may be applied to a display apparatus for displaying images.

公开(公告)号：EP1633139A4

公开(公告)日：2010-01-13

申请号：EP04745590

申请日：2004-06-04

Applicant: SONY CORP

Inventor： KONDO TETSUJIRO ,  NAKANISHI TAKASHI ,  KIKUCHI DAISUKE ,  CHIKAOKA SHIZUO ,  MIYAI TAKESHI ,  NAKAMURA YOSHIAKI ,  HAGA TSUGIHIKO

IPC: H04N7/01 ,  H03M7/30

CPC classification number: H04N7/0125 ,  H04N7/0145

公开(公告)号：EP1511300A4

公开(公告)日：2006-01-04

申请号：EP03733258

申请日：2003-06-03

Applicant: SONY CORP

Inventor： KONDO TETSUJIRO ,  TANAKA KENJI ,  HAGA TSUGIHIKO

IPC: G06T1/00 ,  G06T15/10 ,  H04N5/262 ,  H04N7/18

CPC classification number: G06T3/0062 ,  G06T15/06 ,  G06T15/10 ,  G06T15/50 ,  H04N5/262 ,  H04N7/18

Abstract: An image is easily changed according to the user viewpoint. Original image data on an object imaged from an arbitrary point on a curved surface ohm as a viewpoint is stored. Attention is paid on a straight line Li connecting the viewpoint P and a point Ri on the object, and the intersection point of the straight line Li and the curved surface ohm is expressed as Qi. When the object is viewed from the viewpoint P, the pixel value corresponding to the light ray going from the point Ri on the object to the view point P coincides with the pixel value corresponding to the light ray going from the point Ri on the object to the point Qi on the curved surface ohm, and the pixel value corresponding to the light ray going from this point Ri to the point Qi exists on the original image data. That is, the pixel value when the point Ri on the object is viewed from the viewpoint P can be obtained from the original image data on the object imaged from the point Qi on the curved surface ohm as the viewpoint. The present invention can be applied, for example, to an image processing device for obtaining an image corresponding to a user viewpoint.

公开(公告)号：JP2009110210A

公开(公告)日：2009-05-21

申请号：JP2007281028

申请日：2007-10-29

Applicant: Sony Corp ,  ソニー株式会社

Inventor： HAGA TSUGIHIKO ,  NAKAYA HIDEO ,  HORISHI MASARU ,  ARAKI RYOSUKE ,  INOMATA SEIICHI ,  KAWAI TAKUO ,  HINO TOSHIYUKI ,  KONDO TETSUJIRO

IPC: G06F17/30 ,  G06T1/00 ,  H03M7/30

Abstract: PROBLEM TO BE SOLVED: To reduce a computation complexity and search time.
SOLUTION: A representative value information generating part 32 generates representative value information of information encoded for each block, based on a representative value of each block. A distribution amount sample value computing part 33 computes a distribution amount expressing a distribution of sample values in the block. Representative value information acquiring parts 42-1, 42-2 acquire representative value information of search and comparison objects. A distance computing part of a search part 24 computes a distance between the representative value information of the comparison object with and the representative value information of the search object. A selection part of the search part 24 selects the comparison object similar to the search object, based on the computed distance.
COPYRIGHT: (C)2009,JPO&INPIT

Abstract translation: 要解决的问题：降低计算复杂度和搜索时间。 解决方案：代表值信息生成部分32基于每个块的代表值生成针对每个块编码的信息的代表值信息。 分配量样本值计算部33计算表示块中的样本值分布的分布量。 代表值信息获取部分42-1,42-2获取搜索和比较对象的代表值信息。 搜索部分24的距离计算部分计算比较对象的代表值信息与搜索对象的代表值信息之间的距离。 搜索部分24的选择部分基于所计算的距离来选择与搜索对象类似的比较对象。 版权所有（C）2009，JPO＆INPIT

公开(公告)号：JP2007109205A

公开(公告)日：2007-04-26

申请号：JP2006193326

申请日：2006-07-13

Applicant: Sony Corp ,  ソニー株式会社

Inventor： KONDO TETSUJIRO ,  KOKUBO TETSUSHI ,  TATSUHIRA YASUSHI ,  TANAKA KENJI ,  MUKAI HITOSHI ,  HIBI TAKAFUMI ,  TANAKA KAZUMASA ,  ARIMITSU TETSUHIKO ,  SHIMA JUNICHI ,  HAGA TSUGIHIKO ,  NAKAMURA NORIYUKI

IPC: G03B21/10 ,  G06T3/00 ,  G09B29/00 ,  G09B29/10 ,  G09G5/00 ,  G09G5/36 ,  H04N5/74 ,  H04N7/18 ,  H04N9/31

Abstract: PROBLEM TO BE SOLVED: To allow a user to secure a higher immersion feeling (realistic feeling). SOLUTION: A remote commander 116 comprises a bird's-eye image display part 117 and a touch panel 118. The bird's-eye image display part 116 displays a bird's-eye image by image data D4 of the bird's-eye image. The touch panel 118 notifies a control part 113 about various kinds of operation of the user by the bird's-eye image display part 117. The control part 113 acquires image data of a wide-angle image related to a spot indicated by the user operation in the bird's-eye image from a media database 112, and displays an image of a partial area cut out from the wide-angle image on the basis of the image data on a background image display part 115 as a background image. The partial area cut out from the wide-angle image is varied by a direction indicated by the user operation in the bird's-eye image. The user can actively move inside scenery on the basis of the bird's-eye image. COPYRIGHT: (C)2007,JPO&INPIT

Abstract translation: 要解决的问题：为了使用户能够确保更高的浸入感（真实感）。 解决方案：遥控器116包括鸟瞰图像显示部分117和触摸屏118.鸟瞰图像显示部分116通过鸟瞰图像的图像数据D4显示鸟瞰图像。 触摸面板118通过鸟瞰图像显示部117向控制部113通知用户的各种操作。控制部113获取与用户操作指示的点相关的广角图像的图像数据 来自媒体数据库112的鸟瞰图像，并且基于作为背景图像的背景图像显示部分115上的图像数据，显示从广角图像切出的部分区域的图像。 从广角图像切出的部分区域由鸟瞰图像中的用户操作指示的方向变化。 用户可以根据鸟瞰图形式主动地在风景中移动。 版权所有（C）2007，JPO＆INPIT

公开(公告)号：JP2006350490A

公开(公告)日：2006-12-28

申请号：JP2005173188

申请日：2005-06-14

Applicant: SONY CORP

Inventor： HAGA TSUGIHIKO ,  KONDO TETSUJIRO ,  NAKAMURA NORIYUKI

IPC: G06F3/14 ,  G06F9/44

Abstract: PROBLEM TO BE SOLVED: To easily perform the development of a function relating to a GUI and the change of the addition or deletion of a function. SOLUTION: A GUI display picture is internally hierarchically configured of a plurality of layers 42-1 to 42-3, and they are displayed on a display so as to be superimposed. At least one CUI component configuring the GUI display picture is plotted on the layers 42-1 to 42-3 of display picture size. A GUI event corresponding to the operation of a user to the GUI display picture is supplied to all the layers 42-1 to 42-3, and each GUI image is plotted according to the content of the received event and its own status in each layer 42-1 to 42-3. This invention may be applied to an information processor for plotting the predetermined GUI display picture, and for making the display display it. COPYRIGHT: (C)2007,JPO&INPIT

公开(公告)号：JP2005210343A

公开(公告)日：2005-08-04

申请号：JP2004013891

申请日：2004-01-22

Applicant: Sony Corp ,  ソニー株式会社

Inventor： KONDO TETSUJIRO ,  CHIKAOKA SHIZUO ,  MIYAI TAKASHI ,  NAKAMURA YOSHITERU ,  NAKANISHI TAKASHI ,  HAGA TSUGIHIKO ,  KIKUCHI DAISUKE

IPC: H04N7/01

Abstract: PROBLEM TO BE SOLVED: To generate a higher-definition image signal in real time. SOLUTION: A student image generator 11 generates a CIF image signal as a student image signal by thinning pixels by 1/2 in horizontal and vertical directions from an inputted SD image signal. A learning pair generator 12 generates a learning pair consisting of a signal pair of the SD image signal and the CIF image signal. A coefficient generator 13 generates a normal equation on the basis of the learning pair, and generates a prediction coefficient. An image operator 14 applies the prediction coefficient to the SD image signal to generate an HD image signal. This invention is applicable to an image processor for generating an HD image signal from an SD image signal. COPYRIGHT: (C)2005,JPO&NCIPI

Abstract translation: 要解决的问题：实时产生更高清晰度的图像信号。 解决方案：学生图像生成器11通过从输入的SD图像信号在水平和垂直方向上将像素细化1/2，从而生成CIF图像信号作为学生图像信号。 学习对生成器12生成由SD图像信号和CIF图像信号的信号对构成的学习对。 系数发生器13基于学习对产生正规方程，并产生预测系数。 图像运算器14将预测系数应用于SD图像信号以产生HD图像信号。 本发明可应用于从SD图像信号产生HD图像信号的图像处理器。 版权所有（C）2005，JPO＆NCIPI

公开(公告)号：JP2005192147A

公开(公告)日：2005-07-14

申请号：JP2003434469

申请日：2003-12-26

Applicant: Sony Corp ,  ソニー株式会社

Inventor： KONDO TETSUJIRO ,  HAGA TSUGIHIKO

IPC: H04N7/01 ,  H04N7/24 ,  H04N19/00 ,  H04N19/105 ,  H04N19/136 ,  H04N19/85 ,  H04N19/94 ,  H04N19/98

Abstract: PROBLEM TO BE SOLVED: To appropriately classify pixels in accordance with contents of a plurality of consecutive images. SOLUTION: A DR history calculation circuit 81 calculates a history DR hist representing a temporal transition of a dynamic range DR of an SD pixel at a position corresponding to a pixel of interest based on an ADRC code resulting from applying ADRC processing to an inputted image, the dynamic range DR and data representing the dynamic range DR in data representing a minimum value MIN. Furthermore, a waveform history calculation circuit 82 calculates a history W hist representing a temporal transition of an ADRC code (waveform) of the SD pixel at the position corresponding to the pixel of interest based on the ADRC code in said data. A class code calculation circuit 83 performs classification based on the history DR hist and the history W hist . The present invention may be applied to an image processing apparatus which displays an image on a display device. COPYRIGHT: (C)2005,JPO&NCIPI

Abstract translation: 要解决的问题：根据多个连续图像的内容适当地对像素进行分类。 解决方案：DR历史计算电路81基于ADRC代码计算表示在与感兴趣像素对应的位置处的SD像素的动态范围DR的时间转变的历史DR hist 通过对输入图像应用ADRC处理产生的动态范围DR和表示最小值MIN的数据中的动态范围DR的数据。 此外，波形历史计算电路82基于ADRC代码来计算表示在与感兴趣像素对应的位置处的SD像素的ADRC码（波形）的时间转变的历史W hist 说数据。 类代码计算电路83基于历史DR hist 和历史W hist 进行分类。 本发明可以应用于在显示装置上显示图像的图像处理装置。 版权所有（C）2005，JPO＆NCIPI

公开(公告)号：JP2005123664A

公开(公告)日：2005-05-12

申请号：JP2003353159

申请日：2003-10-14

Applicant: Sony Corp ,  ソニー株式会社

Inventor： KONDO TETSUJIRO ,  KOKUBO TETSUSHI ,  MUKAI HITOSHI ,  HIBI TAKAFUMI ,  HAGA TSUGIHIKO ,  TANAKA KENJI

IPC: G02B27/22 ,  H04N13/02 ,  H04N13/04

Abstract: PROBLEM TO BE SOLVED: To make it possible to display a stereoscopic image for a plurality of viewers without degrading the resolution. SOLUTION: Five projectors 63-1 to 63-5 located at positions forming a similar positional form to those at which five different cameras are placed project images photographed by the cameras to a screen 101. The lights projected onto the screen 101 are refracted by passing through a cylindrical lens array of the screen 101 to provide an image with parallax viewed from users 120 corresponding to the positions of the projectors 63-1 to 63-5, that is, a stereoscopic image. This invention is applicable to television receivers. COPYRIGHT: (C)2005,JPO&NCIPI

Abstract translation: 要解决的问题：使得可以在不降低分辨率的情况下显示多个观看者的立体图像。

解决方案：位于形成与放置五个不同相机的位置相似的位置的五个投影仪63-1至63-5将照相机拍摄的图像投影到屏幕101.投影到屏幕101上的光是 通过穿过屏幕101的柱面透镜阵列折射以提供具有对应于投影仪63-1至63-5的位置（即，立体图像）的用户120观看的视差的图像。 本发明适用于电视接收机。 版权所有（C）2005，JPO＆NCIPI

公开(公告)号：JP2009265567A

公开(公告)日：2009-11-12

申请号：JP2008118339

申请日：2008-04-30

Applicant: Sony Corp ,  ソニー株式会社

Inventor： HINO TOSHIYUKI ,  HAGA TSUGIHIKO ,  TSUJI KOICHIRO ,  MACHIMURA MASANORI ,  KONDO TETSUJIRO

IPC: G10L15/22 ,  G10L15/06 ,  G10L17/00

Abstract: PROBLEM TO BE SOLVED: To perform voice recognition easily. SOLUTION: A feature quantity extraction part 31 extracts the feature of a waveform shape of input voice and the relevance of the feature as a feature quantity. A phoneme discrimination part 34 discriminate a phoneme from the feature quantity. A phoneme discrimination database 32 stores a feature quantity for each phoneme for each unspecified speaker and each speaker. The phoneme discrimination part 34 discriminates a phoneme by comparing the feature quantity extracted by the feature quantity extraction part 31 with the feature quantity of each phoneme for each unspecified speaker stored in the phoneme discrimination database 32 until a speaker is discriminated by a speaker discrimination part 37. When a speaker is discriminated by the speaker discrimination part 37, the phoneme discrimination part 34 discriminates a phoneme by comparing the feature quantity extracted by the feature quantity extraction part 31 with the feature quantity of each phoneme for each discriminated speaker stored in the phoneme discrimination database 32. COPYRIGHT: (C)2010,JPO&INPIT

Abstract translation: 要解决的问题：轻松执行语音识别。 解决方案：特征量提取部分31提取输入语音的波形形状的特征以及特征的相关性作为特征量。 音素识别部分34将音素与特征量进行区分。 音素辨别数据库32存储每个未指定说话者和每个说话者的每个音素的特征量。 音素识别部34通过将由特征量提取部31提取的特征量与存储在音素辨别数据库32中的每个未指定的说话者的每个音素的特征量进行比较来判别音素，直到扬声器被扬声器识别部37区分 当扬声器被扬声器识别部分37区分时，音素识别部分34通过将由特征量提取部分31提取的特征量与存储在音素辨别中的每个识别说话者的每个音素的特征量进行比较来鉴别音素 数据库32.版权所有（C）2010，JPO＆INPIT