公开(公告)号：US20140195560A1

公开(公告)日：2014-07-10

申请号：US14060314

申请日：2013-10-22

Applicant: Samsung Electronics Co., LTD

Inventor： Xin Xin ,  Zhu Li ,  Abhishek Nagar ,  Gaurav Srivastava ,  Zhan Ma ,  Kong Posh Bhat ,  Felix Carlos Fernandes

IPC: G06F17/30

CPC classification number: G06F16/583

Abstract: To improve precision of visual search processing, SIFT points within a query image are forward matched to features in each of a plurality of repository images and SIFT points within each repository image are backward matched to features within the query image. Forward-only, backward-only and forward-and-backward matches may be weighted differently in determining an image match. Two way matching may be triggered by query image bit rate in excess of a threshold or by a sum of weighted distances between matching points exceeding a threshold. Significant performance gains in eliminating false positive matches are achieved.

Abstract translation: 为了提高视觉搜索处理的精度，查询图像内的SIFT点与多个存储库图像中的每一个中的特征向前匹配，并且每个储存库图像内的SIFT点与查询图像内的特征反向匹配。 在确定图像匹配时，仅向前，向后和向前和向后的匹配可以被不同地加权。 可以通过超过阈值的查询图像比特率或超过阈值的匹配点之间的加权距离之和来触发双向匹配。 实现消除假阳性匹配的显着性能提升。

公开(公告)号：US20140201200A1

公开(公告)日：2014-07-17

申请号：US14153907

申请日：2014-01-13

Applicant: Samsung Electronics Co., Ltd.

Inventor： Zhu Li ,  Abhishek Nagar ,  Kong Posh Bhat ,  Xin Xin ,  Gaurav Srivastava ,  Felix Carlos Fernandes

IPC: G06F17/30

CPC classification number: G06F16/532 ,  G06F16/583 ,  G06K9/4676 ,  G06K9/6212 ,  G06K9/6215

Abstract: Global descriptors for images within an image repository accessible to a visual search server are compared based on order statistics processing including sorting (which is a non-linear transform) and heat kernel matching. Affinity scores are computed for Hamming distances between Fisher vector components corresponding to different clusters of global descriptors from a pair of images and normalized to [0, 1], with zero affinity scores assigned to non-active cluster pairs. Linear Discriminant Analysis is employed to determine a sorted vector of affinity scores to obtain a new global descriptor. The resulting global descriptors produce significantly more accurate matching.

Abstract translation: 基于包括排序（其是非线性变换）和热核心匹配的顺序统计处理来比较可视搜索服务器可访问的图像存储库内的图像的全局描述符。 计算对应于来自一对图像的全局描述符的不同簇的Fisher向量分量之间的汉明距离的归一化分数，并归一化为[0,1]，其中零亲和度分数分配给非活动簇对。 线性判别分析用于确定亲和度分数的排序向量以获得新的全局描述符。 所产生的全局描述符产生明显更准确的匹配。

公开(公告)号：US10068147B2

公开(公告)日：2018-09-04

申请号：US14701312

申请日：2015-04-30

Applicant: Samsung Electronics Co., Ltd.

Inventor： Gaurav Srivastava

IPC: G06K9/00 ,  G06K9/34 ,  G06T5/00 ,  G06T11/60 ,  G06K9/46 ,  G06T5/50 ,  H04N13/00 ,  H04N13/02 ,  G06T7/50

Abstract: An electronic device is provided. The electronic device includes processing circuitry. The processing circuitry is configured to receive a first image comprising an image of a first object and a second image comprising an image of a second object. The processing circuitry is also configured to identify a depth of the second object in the second image. The processing circuitry is further configured to insert the image of the first object into the second image at a depth position based on the depth of the second object. The processing circuitry is configured to generate to display the image of the first object and the image of the second object in the second image.

公开(公告)号：US20160321515A1

公开(公告)日：2016-11-03

申请号：US14701312

申请日：2015-04-30

Applicant: Samsung Electronics Co., Ltd.

Inventor： Gaurav Srivastava

IPC: G06K9/34 ,  G06K9/46 ,  H04N13/02 ,  H04N13/00 ,  G06T5/50 ,  G06T7/00

CPC classification number: G06K9/342 ,  G06K9/4642 ,  G06T5/005 ,  G06T5/50 ,  G06T7/50 ,  G06T11/60 ,  G06T2207/10021 ,  G06T2207/30196 ,  H04N13/106 ,  H04N13/239

Abstract: An electronic device is provided. The electronic device includes processing circuitry. The processing circuitry is configured to receive a first image comprising an image of a first object and a second image comprising an image of a second object. The processing circuitry is also configured to identify a depth of the second object in the second image. The processing circuitry is further configured to insert the image of the first object into the second image at a depth position based on the depth of the second object. The processing circuitry is configured to generate to display the image of the first object and the image of the second object in the second image.

Abstract translation: 提供电子设备。 电子设备包括处理电路。 处理电路被配置为接收包括第一对象的图像的第一图像和包括第二对象的图像的第二图像。 处理电路还被配置为识别第二图像中的第二对象的深度。 处理电路还被配置为基于第二对象的深度将第一对象的图像插入到深度位置处的第二图像中。 处理电路被配置为生成以在第二图像中显示第一对象的图像和第二对象的图像。

公开(公告)号：US09235780B2

公开(公告)日：2016-01-12

申请号：US14101047

申请日：2013-12-09

Applicant: Samsung Electronics Co., LTD

Inventor： Zhu Li ,  Xin Xin ,  Abhishek Nagar ,  Gaurav Srivastava ,  Kong Posh Bhat ,  Felix Carlos Fernandes

IPC: G06K9/68 ,  G06K9/62 ,  G06K9/46

CPC classification number: G06K9/6211 ,  G06K9/4671

Abstract: To improve feature selection accuracy during a visual search, interest points within a query image are two-way matched to features in an affine transformed image or otherwise transformed version of the query image. A user device implements a method for selecting local descriptors in the visual search. The method includes: detecting a first set of interest points for the original image; computing an affine transform matrix; computing a new image as a transformation of the original image using the affine transform matrix; detecting a second set of interest points from the and new image; performing a two-way matching between the first set of interest points and the second set of interest points; sorting matching pairs according to a specified self-matching score (SMS); assigning an infinite value to SMS of unmatched interest points from the original image; selecting the interest points based on SMS. Significant performance gains reduce false positive matches.

Abstract translation: 为了在视觉搜索期间提高特征选择精度，查询图像内的兴趣点与仿射变换图像中的特征或查询图像的其他变换版本进行双向匹配。 用户设备实现在视觉搜索中选择本地描述符的方法。 该方法包括：检测原始图像的第一组兴趣点; 计算仿射变换矩阵; 使用仿射变换矩阵计算新图像作为原始图像的变换; 从新图像检测第二组兴趣点; 执行第一组兴趣点与第二组兴趣点之间的双向匹配; 根据指定的自匹配分数（SMS）对匹配对进行排序; 从原始图像向无缝匹配的兴趣点的SMS分配无限值; 基于短信选择兴趣点。 显着的业绩增益减少假阳性匹配。

公开(公告)号：US09727586B2

公开(公告)日：2017-08-08

申请号：US14048988

申请日：2013-10-08

Applicant: Samsung Electronics Co., LTD

Inventor： Zhu Li ,  Ankur Saxena ,  Abhishek Nagar ,  Gaurav Srivastava ,  Kong Posh Bhat

IPC: G06F17/30

CPC classification number: G06F17/30277 ,  G06F17/30247 ,  G06K9/4671

Abstract: To reduce communication costs and computational complexity, only a subset of ranked SIFT points within a query image for a visual search request is transmitted to the visual search server in each iteration of an incremental search. For each candidate match, a flag identifying the matching points is returned by the server for use in computing holistic (e.g., histogram) information for a bounding box within the query image including the matching points. Distance from that query image holistic information is used to reject images from a short list used for a subsequent iteration, if any. If all images are rejected or a match criteria is met during one iteration, the search may terminate early without consideration of remaining SIFT points.

公开(公告)号：US09256617B2

公开(公告)日：2016-02-09

申请号：US13929506

申请日：2013-06-27

Applicant: Samsung Electronics Co., LTD

Inventor： Ankur Saxena ,  Serhat Bucak ,  Abhishek Nagar ,  Felix Carlos Fernandes ,  Gaurav Srivastava

IPC: G06K9/00 ,  G06F17/30 ,  G06K9/46

CPC classification number: G06F17/3025 ,  G06K9/4652 ,  G06K9/4676

Abstract: A method comprises computing a color factor value indicating an amount of color gradients in at least one color channel from the query image. The method comprises combining the color-keypoints with the gray-keypoints when the color factor value is greater than a threshold. A method for performing a visual search comprises extracting a plurality of local descriptors from a query image and then selecting a subset of them based on various criteria's such as visual meaning score. A method comprises aggregating each mean vector for each visual codeword from distances between each visual codeword and local descriptors. The method comprises aggregating variance vector for each visual codeword from the distance between each visual codeword, and local descriptors. The method comprises transmitting aggregated mean vector information and aggregated variance vector information to a search server for efficient image retrieval.

Abstract translation: 一种方法包括计算指示来自查询图像的至少一个颜色通道中的颜色渐变量的颜色因子值。 该方法包括当颜色因子值大于阈值时将颜色关键点与灰色关键点组合。 用于执行视觉搜索的方法包括从查询图像中提取多个局部描述符，然后基于诸如视觉含义分数的各种标准来选择它们的子集。 一种方法包括从每个视觉码字和本地描述符之间的距离聚合每个可视码字的每个平均向量。 该方法包括从每个视觉码字与本地描述符之间的距离聚集每个可视码字的方差向量。 该方法包括向搜索服务器发送聚合的平均向量信息和聚合的方差向量信息以获得有效的图像检索。

公开(公告)号：US20140016863A1

公开(公告)日：2014-01-16

申请号：US13929506

申请日：2013-06-27

Applicant: Samsung Electronics Co., LTD

Inventor： Ankur Saxena ,  Serhat Bucak ,  Abhishek Nagar ,  Felix Carlos Fernandes ,  Gaurav Srivastava

IPC: G06F17/30 ,  G06K9/46

CPC classification number: G06F17/3025 ,  G06K9/4652 ,  G06K9/4676

Abstract: A method comprises computing a color factor value indicating an amount of color gradients in at least one color channel from the query image. The method comprises combining the color-keypoints with the gray-keypoints when the color factor value is greater than a threshold. A method for performing a visual search comprises extracting a plurality of local descriptors from a query image and then selecting a subset of them based on various criteria's such as visual meaning score. A method comprises aggregating each mean vector for each visual codeword from distances between each visual codeword and local descriptors. The method comprises aggregating variance vector for each visual codeword from the distance between each visual codeword, and local descriptors. The method comprises transmitting aggregated mean vector information and aggregated variance vector information to a search server for efficient image retrieval.

Abstract translation: 一种方法包括计算指示来自查询图像的至少一个颜色通道中的颜色渐变量的颜色因子值。 该方法包括当颜色因子值大于阈值时将颜色关键点与灰色关键点组合。 用于执行视觉搜索的方法包括从查询图像中提取多个局部描述符，然后基于诸如视觉含义分数的各种标准来选择它们的子集。 一种方法包括从每个视觉码字和本地描述符之间的距离聚合每个可视码字的每个平均向量。 该方法包括从每个视觉码字与本地描述符之间的距离聚集每个可视码字的方差向量。 该方法包括向搜索服务器发送聚合的平均向量信息和聚合的方差向量信息以获得有效的图像检索。

公开(公告)号：US10140507B2

公开(公告)日：2018-11-27

申请号：US14982299

申请日：2015-12-29

Applicant: Samsung Electronics Co., Ltd.

Inventor： Gaurav Srivastava

IPC: G06K9/00 ,  G02B27/01 ,  G06F3/00 ,  G06F3/01 ,  G06T19/00 ,  G06F3/03 ,  G09G5/00

Abstract: A virtual reality (VR) headset configured to be worn by a user. The VR headset comprises: i) a forward-looking vision sensor for detecting objects in the forward field of view of the VR headset; ii) a downward-looking vision sensor for detecting objects in the downward field of view of the VR headset; iii) a controller coupled to the forward-looking vision sensor and the downward-looking vision sensor. The controller is configured to: a) detect a hand in a first image captured by the forward-looking vision sensor; b) detect an arm of the user in a second image captured by the downward-looking vision sensor; and c) determine whether the detected hand in the first image is a hand of the user.

公开(公告)号：US20140310314A1

公开(公告)日：2014-10-16

申请号：US14249929

申请日：2014-04-10

Applicant: Samsung Electronics Co., Ltd.

Inventor： Zhu Li ,  Abhishek Nagar ,  Gaurav Srivastava ,  Felix Carlos Fernandes

IPC: G06F17/30

CPC classification number: G06F16/583

Abstract: A method and apparatus include extracting a global descriptor from a query image with a plurality of segments. The method also includes identifying segments with a desirable discriminating potential by analyzing data of the plurality of segments based on an available image database. The method also includes creating a bitmask where the identified segments are active. The method also includes masking any segment of the plurality of segments of the global descriptor that are inactive according to the bitmaskA method includes extracting a global descriptor from a query image and identifying one or more reference global descriptors. The method also includes determining a distance between the global descriptor and each of the one or more reference global descriptors. In addition, the method includes, responsive to the distance satisfying a threshold, adding an image associated with each of the one or more reference global descriptors that satisfy the threshold to a list.

Abstract translation: 一种方法和装置包括从具有多个段的查询图像提取全局描述符。 该方法还包括基于可用图像数据库分析多个片段的数据来识别具有期望的鉴别潜力的片段。 该方法还包括创建一个位掩码，其中标识的段是活动的。 该方法还包括根据位掩码掩蔽全局描述符的多个段的不活动的任何段。方法包括从查询图像提取全局描述符并识别一个或多个参考全局描述符。 该方法还包括确定全局描述符与一个或多个参考全局描述符中的每一个之间的距离。 此外，该方法包括响应于满足阈值的距离，将与满足阈值的一个或多个参考全局描述符中的每一个相关联的图像添加到列表。