公开(公告)号：US20130229484A1

公开(公告)日：2013-09-05

申请号：US13804106

申请日：2013-03-14

Applicant: SONY COMPUTER ENTERTAINMENT INC.

Inventor： Kyoko FURUMURA ,  Shinichi HIRATA ,  Takeshi YAMAGISHI

IPC: H04N13/00

CPC classification number: H04N5/272 ,  H04N5/232 ,  H04N5/23238 ,  H04N13/156 ,  H04N13/189

Abstract: A control point detector detects controls points, with which moving images are associated with a region of a panoramic image, by detecting feature points between the panoramic image and a frame of part of the moving images. An alignment processing unit adjusts the alignment of the moving images relative to the panoramic image, based on the control points. A mapping processing unit maps the panoramic image and a still image frame of the moving images the alignment of which adjusted by the alignment processing unit, onto a three-dimensional (3D) panoramic space as textures. A 3D image generator generates a 3D panoramic image, when the three-dimensional panoramic image is viewed in a specified line of sight. When a moving image reproduction unit receives instructions to reproduce the moving images in the 3D panoramic image, the moving image reproduction unit reproduces the moving images.

Abstract translation: 控制点检测器通过检测全景图像与运动图像的一部分的帧之间的特征点来检测运动图像与全景图像的区域相关联的控制点。 对准处理单元基于控制点来调整运动图像相对于全景图像的对准。 映射处理单元将通过对准处理单元调整的对齐的全景图像和静止图像帧映射到作为纹理的三维（3D）全景空间。 当在指定视线中观看三维全景图像时，3D图像生成器生成3D全景图像。 当运动图像再现单元接收到用于再现3D全景图像中的运动图像的指令时，运动图像再现单元再现运动图像。

公开(公告)号：US20150249794A1

公开(公告)日：2015-09-03

申请号：US14710873

申请日：2015-05-13

Applicant: Sony Computer Entertainment Inc.

Inventor： Kyoko FURUMURA ,  Shinichi HIRATA ,  Takeshi YAMAGISHI

IPC: H04N5/272 ,  H04N13/00 ,  H04N5/232

CPC classification number: H04N5/272 ,  H04N5/232 ,  H04N5/23238 ,  H04N13/156 ,  H04N13/189

Abstract: Methods and apparatus provide for: acquiring a still image of a scene at a location; acquiring a moving image of at least one object at the scene taken at a different time of day, a different date, and/or a different season; selecting an image frame from among a plurality of image frames of the moving image as a still image frame; generating a three-dimensional (3D) image from the still image in a three-dimensional (3D) object space; superimposing the still image frame from the moving image onto the three-dimensional (3D) image; displaying the three-dimensional (3D) image on a display screen for a user; and generating the moving image superimposed on the three-dimensional (3D) image, starting at the selected image frame, in response to receiving a command from the user.

Abstract translation: 方法和装置提供：在一个位置获取场景的静止图像; 获取在一天的不同时间，不同日期和/或不同季节拍摄的场景中的至少一个物体的运动图像; 从所述运动图像的多个图像帧中选择图像帧作为静止图像帧; 在三维（3D）对象空间中从静止图像生成三维（3D）图像; 将静止图像帧从运动图像叠加到三维（3D）图像上; 在用户的显示屏上显示三维（3D）图像; 以及响应于接收到来自用户的命令，在所选择的图像帧上生成叠加在三维（3D）图像上的运动图像。

公开(公告)号：US20130229546A1

公开(公告)日：2013-09-05

申请号：US13796038

申请日：2013-03-12

Applicant: SONY COMPUTER ENTERTAINMENT INC.

Inventor： Kyoko FURUMURA ,  Shinichi HIRATA

IPC: H04N5/235

CPC classification number: H04N5/2355 ,  G06T3/4038 ,  G06T5/007 ,  G06T5/50 ,  G06T2207/10004 ,  G06T2207/10144 ,  G06T2207/20208 ,  G06T2207/20221 ,  H04N5/23238

Abstract: A low dynamic range (LDR) image data storage stores LDR image sets each of which includes a plurality of LDR segmented images captured with a camera panned. Here each LDR image set is of a different exposure level, respectively. A high-dynamic-range (HDR) synthesizing unit combines and merges a plurality of LDR segmented images of different exposure levels so as to generate an HDR segmented image set including a plurality of HDR segmented images of different pan angles or tilt angles. A panoramic image synthesizing unit generates an HDR panoramic image by stitching together adjacent HDR segmented images.

Abstract translation: 低动态范围（LDR）图像数据存储存储LDR图像集合，每个LDR图像集包括用相机平移拍摄的多个LDR分割图像。 这里，每个LDR图像组分别具有不同的曝光水平。 高动态范围（HDR）合成单元组合并合并不同曝光水平的多个LDR分割图像，以便产生包括不同摇摄角度或倾斜角度的多个HDR分割图像的HDR分割图像组。 全景图像合成单元通过将相邻的HDR分割图像拼接在一起来生成HDR全景图像。

公开(公告)号：US20160100104A1

公开(公告)日：2016-04-07

申请号：US14966726

申请日：2015-12-11

Applicant: Sony Computer Entertainment Inc.

Inventor： Kyoko FURUMURA ,  Takeshi YAMAGISHI ,  Shinichi HIRATA

IPC: H04N5/232 ,  H04N13/04

CPC classification number: H04N5/23229 ,  G06T3/0062 ,  G06T3/4038 ,  H04N5/23238 ,  H04N5/23293 ,  H04N13/30

Abstract: When a place where a different panoramic image was captured is present in a direction in which a panoramic image to be displayed is captured, a marker setting unit associates a marker indicating that the different panoramic image is present in said image-captured direction with the panoramic image to be displayed. A mapping processing unit maps, as a texture, the panoramic image to be displayed, with which the marker of the different panoramic image is associated, in a three-dimensional panoramic space. A three-dimensional image generation unit generates a three-dimensional panoramic image when the three-dimensional panoramic space is viewed in a designated visual line direction, with a place where the panoramic image to be displayed is captured as a visual point position. A user interface unit accepts a user's instruction regarding the displayed three-dimensional panoramic image.

Abstract translation: 当拍摄不同全景图像的地方存在于拍摄要显示的全景图像的方向时，标记设置单元将指示所述图像拍摄方向上的不同全景图像的标记与全景图像相关联 要显示的图像 映射处理单元在三维全景空间中映射与不同全景图像的标记相关联的要显示的全景图像作为纹理。 当在指定的视线方向上观看三维全景空间时，三维图像生成单元产生三维全景图像，其中拍摄要显示的全景图像的位置作为视点位置。 用户界面单元接受用户关于显示的三维全景图像的指令。

公开(公告)号：US20160086306A1

公开(公告)日：2016-03-24

申请号：US14813605

申请日：2015-07-30

Applicant: Sony Computer Entertainment Inc.

Inventor： Yoichi NISHIMAKI ,  Takeshi YAMAGISHI ,  Shinichi HIRATA ,  Kyoko FURUMURA

IPC: G06T3/40 ,  H04N5/232 ,  G02B27/01 ,  G06T11/60

CPC classification number: G06T3/40 ,  G02B27/0093 ,  G02B27/017 ,  G02B2027/014 ,  G02B2027/0141 ,  G06T11/00 ,  G06T2210/22 ,  H04N5/23238 ,  H04N5/23261 ,  H04N5/23293

Abstract: An image generating device includes: a storage section that stores images of surrounding spaces centered at plural different fixed points; a detecting section that detects translational movement on the basis of the location of the point of view; an image processor that acquires an image of a displaying target by clipping out part of the image of the surrounding space centered at the fixed point, stored in the storage section, on the basis of the location of the point of view and the direction of the line of sight; and a switching section that makes switching to the image of the surrounding space centered at another fixed point closest to the point of view after translational movement if the plural different fixed points are so disposed that the surrounding spaces centered at the fixed points overlap with each other in a world coordinate system and translational movement is detected.

Abstract translation: 图像产生装置包括：存储部，其存储以多个不同固定点为中心的周围空间的图像; 检测部，其基于所述视点的位置检测平移运动; 图像处理器，其基于视点的位置和视点的方向，通过削减存储在存储部中的固定点的中心的周围空间的图像的一部分来获取显示对象的图像; 视线; 以及如果多个不同的固定点被布置为使得以固定点为中心的周围空间彼此重叠，则在平移运动之后切换到最靠近视点的另一固定点的周围空间的图像的切换部分 在世界坐标系中，检测到平移运动。

公开(公告)号：US20130314402A1

公开(公告)日：2013-11-28

申请号：US13804186

申请日：2013-03-14

Applicant: Sony Computer Entertainment Inc.

Inventor： Kyoko FURUMURA ,  Shinichi HIRATA ,  Takeshi YAMAGISHI

IPC: G09G5/14 ,  G06T19/00

CPC classification number: G09G5/14 ,  G03B35/02 ,  G03B37/00 ,  G06T3/0062 ,  G06T3/4038 ,  G06T15/205 ,  G06T19/006 ,  H04N13/279 ,  H04N21/23412 ,  H04N21/4318 ,  H04N21/44012 ,  H04N21/45455 ,  H04N21/8146 ,  H04N21/84

Abstract: A mask region extraction unit extracts a region that is to be masked in a panoramic image. A mask processing unit generates the object image where the region to be masked in the panoramic image is masked. A positioning unit adjusts the direction of a spherical image to the shooting direction of the object image. A mapping processing unit maps the mask-processed object image and the spherical image onto a three-dimensional (3D) object space as textures. A 3D image generator generates a 3D panoramic image, when the 3D panoramic image is viewed in a specified line of sight in such a manner so as to regard the shooting location of the panoramic as a viewpoint position.

Abstract translation: 掩模区域提取单元提取在全景图像中要被掩蔽的区域。 掩模处理单元产生将全景图像中要被掩蔽的区域被遮蔽的对象图像。 定位单元根据对象图像的拍摄方向调整球面图像的方向。 映射处理单元将掩模处理的对象图像和球面图像作为纹理映射到三维（3D）对象空间。 当以指定的视线以这样的方式观看3D全景图像以便将全景的拍摄位置视为视点位置时，3D图像生成器生成3D全景图像。