-
公开(公告)号：US20180041708A1
公开(公告)日：2018-02-08
申请号：US15783603
申请日：2017-10-13
Applicant: Futurewei Technologies, Inc.
Inventor： Zhenyu Wu , Wei Jiang , John Wus , Hong Heather Yu
CPC classification number: H04N5/23267 , G06T7/246 , G06T7/277 , G06T2207/10016 , G06T2207/20016 , H04N5/23251
Abstract: A method of stabilizing a video in real time using a single pass including receiving consecutive video frames, where the consecutive video frames include a current video frame and previous video frames, storing the consecutive video frames in a buffer, estimating a global motion for the current video frame by describing a camera's relative motion between the current video frame and one of the previous video frames adjacent to the current video frame, estimating a long-term camera motion for the current video frame by determining a geometric mean of an accumulation of the estimate of the global motion for the current video frame and an estimate of global motion for each of the previous video frames, and displaying the current video frame on a display of an electronic device, the current video frame stabilized based on the estimate of the long-term camera motion.
-
12.发明申请公开(公告)号：US20170076499A1
公开(公告)日：2017-03-16
申请号：US14852247
申请日：2015-09-11
Applicant: Futurewei Technologies, Inc.
IPC: G06T19/00
CPC classification number: G06T19/006 , G06T7/00 , G06T2219/024 , G11B27/00 , G11B27/031
Abstract: Methods and devices permit a user to insert multiple virtual objects into a real world video scene. Some inserted objects may be statically tied to the scene, while other objects are designated as moving with certain moving objects in the scene. Markers are not used to insert the virtual objects. Users of separate mobile devices can share their inserted virtual objects to create a multi-user, multi-object augmented reality (AR) experience.
Abstract translation: 方法和设备允许用户将多个虚拟对象插入到真实世界的视频场景中。 一些插入的对象可以静态地绑定到场景，而其他对象被指定为与场景中的某些移动对象一起移动。 标记不用于插入虚拟对象。 单独移动设备的用户可以共享其插入的虚拟对象，以创建多用户，多对象增强现实（AR）体验。
-
公开(公告)号：US11037029B2
公开(公告)日：2021-06-15
申请号：US16157983
申请日：2018-10-11
Applicant: Futurewei Technologies, Inc.
Abstract: Provided are an apparatus and a method of multi-stage image recognition. For the multi-stage image recognition, categorized object data is received from a first deep neural network. A second deep neural network is trained on subcategory customization data that relates to a non-ideal environment when the second deep neural network produces invalid subcategorized object data from the categorized object data, and generates an image recognition result using the second deep neural network as trained.
-
14.发明申请Methods and Systems for Light Field Augmented Reality/Virtual Reality on Mobile Devices 审中-公开移动设备光场增强现实/虚拟现实的方法与系统公开(公告)号：US20170069133A1
公开(公告)日：2017-03-09
申请号：US14848759
申请日：2015-09-09
Applicant: Futurewei Technologies, Inc.
Inventor： Jinwei Gu , Bennett Wilburn , Wei Jiang
CPC classification number: G06T19/006 , G06T2200/21
Abstract: Methods and apparatus are described that enable augmented or virtual reality based on a light field. A geometric proxy of a mobile device such as a smart phone is used during the process of inserting a virtual object from the light field into the real world images being acquired. For example, a mobile device includes a processor and a camera coupled to the processor. The processor is configured to define a view-dependent geometric proxy, record images with the camera to produce recorded frames and, based on the view-dependent geometric proxy, render the recorded frames with an inserted light field virtual object.
Abstract translation: 描述了基于光场实现增强或虚拟现实的方法和装置。 在将虚拟对象从光场插入被获取的真实世界图像的过程中，使用诸如智能电话的移动设备的几何代理。 例如，移动设备包括处理器和耦合到处理器的相机。 处理器被配置为定义依赖于视图的几何代理，用相机记录图像以产生记录的帧，并且基于视图相关的几何代理，使用插入的光场虚拟对象呈现所记录的帧。
-
15.发明授权公开(公告)号：US09286680B1
公开(公告)日：2016-03-15
申请号：US14581788
申请日：2014-12-23
Applicant: Futurewei Technologies, Inc.
CPC classification number: G06T7/0018 , G06T1/0007 , H04N5/222 , H04N5/247 , H04N17/002
Abstract: An embodiment method for computationally adjusting images from a multi-camera system includes receiving calibrated image sequences, with each of the calibrated image sequences corresponding to a camera in a camera array and having one or more image frames. A target camera model is computed for each camera in the camera array and according to target camera poses or target camera intrinsic matrices for the respective camera. The computing generates a transformation matrix for each of the one or more first cameras. The transformation matrix for each of the one or more first cameras is applied to the calibrated image sequence corresponding to the respective camera. The transformation matrix warps each image frame of the calibrated image sequence and generates target image sequences.
Abstract translation: 用于从多摄像机系统计算地调整图像的实施例方法包括接收经校准的图像序列，其中每个校准图像序列对应于相机阵列中的相机并且具有一个或多个图像帧。 针对摄像机阵列中的每个摄像机，并根据目标摄像机姿态或相应摄像机的目标摄像机固有矩阵计算目标摄像机模型。 该计算生成一个或多个第一相机中的每一个的变换矩阵。 用于一个或多个第一相机中的每一个的变换矩阵被应用于对应于各个相机的校准图像序列。 变换矩阵扭曲校准图像序列的每个图像帧并产生目标图像序列。
-
Patent Agency Ranking
公开(公告)号：US10748061B2
公开(公告)日：2020-08-18
申请号：US15383804
申请日：2016-12-19
Applicant: Futurewei Technologies, Inc.
Abstract: A robotic device is disclose as having deep reinforcement learning capability. The device includes non-transitory memory comprising instructions and one or more processors in communication with the memory. The instructions cause the one or more processors to receive a sensing frame, from a sensor, comprising an image. The processors then determine a movement transition based on the sensing frame and the deep reinforcement learning, wherein the deep reinforcement learning uses at least one of a map coverage reward, a map quality reward, or a traversability reward to determine the movement transition. The processors then update an area map based on the sensing frame and the deep reinforcement learning using a visual simultaneous localization and mapping (SLAM) process to determine the map updates.
-
公开(公告)号：US20190217476A1
公开(公告)日：2019-07-18
申请号：US15870626
申请日：2018-01-12
Applicant: Futurewei Technologies, Inc.
Abstract: A system and method of tracking an object and navigating an object tracking robot includes receiving tracking sensor input representing the object and an environment at multiple times, responsive to the tracking sensor input, calculating positions of the robot and the object at the multiple times, and using a computer implemented deep reinforcement learning (DRL) network trained as a function of tracking quality rewards and robot navigation path quality rewards, the DRL network being responsive to the calculated positions of the robot and the object at the multiple times to determine possible actions specifying movement of the object tracking robot from a current position of the robot and target, determine quality values (Q-values) for the possible actions, and select an action as a function of the Q-values. A method of training the DRL network is also included.
-
公开(公告)号：US09928656B2
公开(公告)日：2018-03-27
申请号：US14852247
申请日：2015-09-11
Applicant: Futurewei Technologies, Inc.
CPC classification number: G06T19/006 , G06T7/00 , G06T2219/024 , G11B27/00 , G11B27/031
Abstract: Methods and devices permit a user to insert multiple virtual objects into a real world video scene. Some inserted objects may be statically tied to the scene, while other objects are designated as moving with certain moving objects in the scene. Markers are not used to insert the virtual objects. Users of separate mobile devices can share their inserted virtual objects to create a multi-user, multi-object augmented reality (AR) experience.
-
19.发明申请PARALLAX TOLERANT VIDEO STITCHING WITH SPATIAL-TEMPORAL LOCALIZED WARPING AND SEAM FINDING 有权PARALLAX耐力视频刺绣与空间局部化的局部加温和发现公开(公告)号：US20160142643A1
公开(公告)日：2016-05-19
申请号：US14540960
申请日：2014-11-13
Applicant: Futurewei Technologies Inc.
CPC classification number: H04N5/265 , G06T3/4038 , G06T5/006 , G06T11/60 , H04N5/262
Abstract: An apparatus is configured to perform a method of parallax tolerant video stitching. The method includes determining a plurality of video sequences to be stitched together; performing a spatial-temporal localized warping computation process on the video sequences to determine a plurality of target warping maps; warping a plurality of frames among the video sequences into a plurality of target virtual frames using the target warping maps; performing a spatial-temporal content-based seam finding process on the target virtual frames to determine a plurality of target seam maps; and stitching the video sequences together using the target seam maps.
Abstract translation: 一种装置被配置为执行视差容许视频拼接的方法。 该方法包括确定要缝合在一起的多个视频序列; 对视频序列执行空间 - 时间局部扭曲计算过程以确定多个目标翘曲图; 使用目标变形图将视频序列中的多个帧变形为多个目标虚拟帧; 对所述目标虚拟帧执行基于空间 - 时间的基于内容的接缝查找过程以确定多个目标接缝图; 并使用目标接缝图将视频序列拼接在一起。
-
公开(公告)号：US10404917B2
公开(公告)日：2019-09-03
申请号：US15783603
申请日：2017-10-13
Applicant: Futurewei Technologies, Inc.
Inventor： Zhenyu Wu , Wei Jiang , John Wus , Hong Heather Yu
Abstract: A method of stabilizing a video in real time using a single pass including receiving consecutive video frames, where the consecutive video frames include a current video frame and previous video frames, storing the consecutive video frames in a buffer, estimating a global motion for the current video frame by describing a camera's relative motion between the current video frame and one of the previous video frames adjacent to the current video frame, estimating a long-term camera motion for the current video frame by determining a geometric mean of an accumulation of the estimate of the global motion for the current video frame and an estimate of global motion for each of the previous video frames, and displaying the current video frame on a display of an electronic device, the current video frame stabilized based on the estimate of the long-term camera motion.
-
-
-
-
-
-
-
-
-
Search Results
24
records
(took 0.018 seconds)
