公开(公告)号：US08781255B2

公开(公告)日：2014-07-15

申请号：US13434028

申请日：2012-03-29

申请人： Zhe Lin ,  Jonathan W. Brandt

发明人： Zhe Lin ,  Jonathan W. Brandt

IPC分类号： G06K9/00

CPC分类号： G06F17/30262 ,  G06F17/3025 ,  G06K9/4676

摘要： Each image of a set of images is characterized with a set of sparse feature descriptors and a set of dense feature descriptors. In some embodiments, both the set of sparse feature descriptors and the set of dense feature descriptors are calculated based on a fixed rotation for computing texture descriptors, while color descriptors are rotation invariant. In some embodiments, the descriptors of both sparse and dense features are then quantized into visual words. Each database image is represented by a feature index including the visual words computed from both sparse and dense features. A query image is characterized with the visual words computed from both sparse and dense features of the query image. A rotated local Bag-of-Features (BoF) operation is performed upon a set of rotated query images against the set of database images. Each of the set of images is ranked based on the rotated local Bag-of-Features operation.

摘要翻译： 一组图像的每个图像用一组稀疏特征描述符和一组密集特征描述符来表征。 在一些实施例中，基于用于计算纹理描述符的固定旋转来计算稀疏特征描述符集合和密集特征描述符集合，而颜色描述符是旋转不变量。 在一些实施例中，稀疏和密集特征的描述符然后被量化为视觉词。 每个数据库图像由特征索引表示，包括从稀疏和密集特征计算的视觉词。 查询图像的特征在于从查询图像的稀疏和密集特征计算的视觉词。 一组旋转的本地特征（BoF）操作是针对一组数据库图像进行旋转的查询图像执行的。 基于旋转的本地Bag-of-Features操作对该组图像中的每一个进行排名。

公开(公告)号：US20140164297A1

公开(公告)日：2014-06-12

申请号：US13709773

申请日：2012-12-10

申请人： HEWLETT-PACKARD DEVELOPMENT COMPAN

发明人： Vinay Deolalikar ,  Hernan Laffitte

IPC分类号： G06N99/00

CPC分类号： G06N99/005 ,  G06K9/00442 ,  G06K9/00463 ,  G06K9/4676

摘要： A method of generating training documents for training a classifying device comprises, with a processor, sampling from a distribution of words in a number of original documents, and creating a number of pseudo-documents from the distribution of words, the pseudo-documents comprising a similar distribution of words as the original documents. A device for classifying textual documents comprises a processor; and a memory communicatively coupled to the processor, the memory comprising a sampling module to, when executed by the processor, determine the distribution of words in a number of original documents, a pseudo-document creation module to, when executed by the processor, create a number of pseudo-documents from the distribution of words, the pseudo-documents comprising a similar distribution of words as the original documents, and a training module to, when executed by the processor, train the device to classify textual documents based on the pseudo-documents.

摘要翻译： 一种生成用于训练分类装置的训练文档的方法包括：利用处理器对来自多个原始文档中的单词分布进行采样，以及从分发单词创建多个伪文档，所述伪文档包括 与原始文件相似的分词。 用于分类文本文档的设备包括处理器; 以及通信地耦合到所述处理器的存储器，所述存储器包括采样模块，以在由所述处理器执行时确定多个原始文档中的字的分布;伪文档创建模块，当被所述处理器执行时，创建 来自分发单词的多个伪文档，包含与原始文档相似的单词分类的伪文档，以及训练模块，当由处理器执行时，训练模块基于该伪文档对文本文档进行分类 文件。

公开(公告)号：US20140149376A1

公开(公告)日：2014-05-29

申请号：US14129005

申请日：2011-06-23

申请人： Ken Kutaragi ,  Takashi Usuki ,  Yasuhiko Yokote

发明人： Ken Kutaragi ,  Takashi Usuki ,  Yasuhiko Yokote

IPC分类号： G06F17/30

CPC分类号： G06F17/30277 ,  G06F17/30259 ,  G06F17/30864 ,  G06K9/4676 ,  G06K9/6224 ,  G06K9/6263 ,  G06T11/206

摘要： A system for comprehensively and effectively acquiring, as an interest graph, targets and regions of interest unique to a user is provided. A system according to the invention is a search system using as input means image information containing various objects and subjects, in which by querying an image recognition engine on the server side via a network about a selected image, the image recognition engine extracts and recognizes in real time various generic objects and the like contained in the specified image, and notifies a relevance search engine on the server side of recognized image components contained in the input image, and the relevance search engine extracts related elements for the individual image components, and visually presents a relevance graph with the extracted related elements as nodes together with the depths of relationships between the nodes.

摘要翻译： 提供了一种用于全面有效地获取用户特有的目标和兴趣区域作为兴趣图的系统。 根据本发明的系统是一种搜索系统，其使用包含各种对象和对象的图像信息作为输入装置，其中通过关于所选图像的网络查询服务器侧的图像识别引擎，图像识别引擎提取并识别 实时地包含在指定图像中的各种通用对象等，并且将包含在输入图像中的识别图像分量的服务器侧的相关性搜索引擎通知相关性搜索引擎，并且可视地提取各个图像分量的相关元素 将提取的相关元素作为节点与节点之间的关系深度一起呈现相关图。

公开(公告)号：US20140105509A1

公开(公告)日：2014-04-17

申请号：US13651827

申请日：2012-10-15

申请人： CANON KABUSHIKI KAISHA

发明人： Dariusz T. Dusberger ,  Bradley Scott Denney

IPC分类号： G06K9/68

CPC分类号： G06K9/4676

摘要： Systems and methods for scoring similarity count a number of matching visual words between a query image and a candidate image, generate an image similarity score for the query image and the candidate image based at least in part on the number of matching visual words, and generate a normalized image similarity score based at least in part on the image similarity score and one or more of a complexity of the query image and a complexity of the candidate image.

摘要翻译： 用于评分相似性的系统和方法计数查询图像和候选图像之间的多个匹配的视觉词，至少部分地基于匹配的视觉词的数量生成用于查询图像和候选图像的图像相似性得分，并且生成 至少部分地基于图像相似度得分和查询图像的复杂度和候选图像的复杂度中的一个或多个的归一化图像相似性得分。

公开(公告)号：US20140105505A1

公开(公告)日：2014-04-17

申请号：US13832122

申请日：2013-03-15

申请人： GOOGLE INC.

发明人： Sergey Ioffe ,  Mohamed Aly ,  Charles J. Rosenberg

IPC分类号： G06F17/30

CPC分类号： G06F17/30247 ,  G06F17/3025 ,  G06K9/4676 ,  G06K9/6202

摘要： Methods, systems, and apparatus, including computer programs encoded on computer storage media, for determining image search results. One of the methods includes generating a plurality of feature vectors for each image in a collection of images, wherein each feature vector is associated with an image tile of an image, wherein each feature vector corresponds to one of a plurality of predetermined visual words. All images in the collection of images that share at least a threshold number of matching visual words associated with matching image tiles are classified as near-duplicate images.

摘要翻译： 方法，系统和装置，包括在计算机存储介质上编码的计算机程序，用于确定图像搜索结果。 一种方法包括为图像集合中的每个图像生成多个特征向量，其中每个特征向量与图像的图像块相关联，其中每个特征向量对应于多个预定视觉词中的一个。 共享与匹配的图像块相关联的至少阈值数量的匹配视觉词的图像集合中的所有图像被分类为近似重复的图像。

公开(公告)号：US20140016863A1

公开(公告)日：2014-01-16

申请号：US13929506

申请日：2013-06-27

申请人： Samsung Electronics Co., LTD

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

IPC分类号： G06F17/30 ,  G06K9/46

CPC分类号： G06F17/3025 ,  G06K9/4652 ,  G06K9/4676

摘要： 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.

摘要翻译： 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：US20140003716A1

公开(公告)日：2014-01-02

申请号：US13537100

申请日：2012-06-29

申请人： Elena A. Fedorovskaya ,  Jeffrey Clarence Snyder ,  Sharon Lynn Chu Yew Yee ,  Andrew F. Kurtz

发明人： Elena A. Fedorovskaya ,  Jeffrey Clarence Snyder ,  Sharon Lynn Chu Yew Yee ,  Andrew F. Kurtz

IPC分类号： G06K9/68

CPC分类号： G06K9/00677 ,  G06K9/00684 ,  G06K9/4676

摘要： A method for presenting digital images having a high interest level to a particular person selected from a set of candidate digital images. The candidate digital image are analyzed to designate one or more image elements, and familiarity levels are determined of the designated image elements to the particular person. For each candidate digital image, an associated interest level to the particular person is determined responsive to the determined familiarity levels. One or more of the candidate digital images are selected based on the determined interest levels and are presented to the particular person.

摘要翻译： 一种用于向从一组候选数字图像中选择的特定人物呈现具有高兴趣度的数字图像的方法。 分析候选数字图像以指定一个或多个图像元素，并且确定对特定人的指定图像元素的熟悉程度。 对于每个候选数字图像，响应于所确定的熟悉程度来确定与特定人的关联兴趣水平。 基于所确定的兴趣级别选择候选数字图像中的一个或多个，并将其呈现给特定的人。

公开(公告)号：US08577153B2

公开(公告)日：2013-11-05

申请号：US13056001

申请日：2010-10-28

申请人： Manni Duan ,  Xinci Wen ,  Xurui Feng

发明人： Manni Duan ,  Xinci Wen ,  Xurui Feng

IPC分类号： G06K9/62 ,  G06K9/54

CPC分类号： G06K9/00664 ,  G06F17/30256 ,  G06K9/4676

摘要： The present disclosure discloses a method and apparatus for creating a sample image index table, filtering image, and searching image, to improve accuracy of monitoring images. A method for image filtering comprises: establishing a sample image index table; extracting regional characteristics from an image to be searched; clustering the regional characteristics of the image to be searched into corresponding nodes; obtaining a corresponding sample image identification by indexing the sample image index table using node identifications of the nodes of the image to be searched; determining a number of duplicate nodes between the image to be searched and the sample image; obtaining a degree of similarity of the image to be searched based on a number of the nodes of the image to be searched and a number of the nodes of the sample image; and filtering out the image to be searched when a degree of similarity between the image to be searched and the sample image exceeds a similarity threshold.

摘要翻译： 本公开公开了一种用于创建样本图像索引表，过滤图像和搜索图像以提高监视图像的精度的方法和装置。 一种用于图像滤波的方法包括：建立样本图像索引表; 从要搜索的图像中提取区域特征; 将要搜索的图像的区域特征聚类到对应的节点中; 通过使用要搜索的图像的节点的节点标识索引样本图像索引表来获得相应的样本图像标识; 确定要搜索的图像与样本图像之间的重复节点的数量; 基于要搜索的图像的节点数和样本图像的节点数，获得要搜索的图像的相似度; 并且当要搜索的图像与样本图像之间的相似度超过相似阈值时，滤出要搜索的图像。

公开(公告)号：US20130254191A1

公开(公告)日：2013-09-26

申请号：US13707251

申请日：2012-12-06

申请人： THE TRUSTEES OF COLUMBIA UNIV IN THE CITY OF NEW

发明人： Junfeng He ,  Shih-Fu Chang ,  Tai-Hsu Lin

IPC分类号： G06F17/30

CPC分类号： G06F17/3053 ,  G06F17/30247 ,  G06K9/4676

摘要： Determining ranked candidate media in response to media query data corresponding to a query media includes receiving the media query data including feature data of the query media, coordinate data, and boundary data, matching the features with corresponding features of an media database using the feature data to identify features in the media database within a predetermined hamming distance in a hash table from the corresponding features of the query media to obtain matched features in the media database, determining candidate media whose number of matched features exceeds a matched feature threshold, generating a geometry similarity score between the query media and each candidate media using the feature data and the coordinate data, generating a boundary similarity score between the query media and each candidate media using the boundary data, ranking the candidate media based on the numbers of matched features, the geometry similarity scores and the boundary similarity scores.

摘要翻译： 响应于与查询媒体相对应的媒体查询数据确定排名的候选媒体包括：使用特征数据接收包括查询媒体的特征数据，坐标数据和边界数据的媒体查询数据，使特征与媒体数据库的相应特征相匹配 以从查询媒体的相应特征在哈希表中的预定汉明距离内识别媒体数据库中的特征，以获得媒体数据库中的匹配特征，确定匹配特征数量超过匹配特征阈值的候选媒体，生成几何 使用特征数据和坐标数据在查询媒体和每个候选媒体之间的相似性得分，使用边界数据在查询媒体和每个候选媒体之间生成边界相似性得分，基于匹配特征的数量对候选媒体进行排名， 几何相似度分数和边界相似性分数。

公开(公告)号：US20130251340A1

公开(公告)日：2013-09-26

申请号：US13425455

申请日：2012-03-21

申请人： Wei Jiang ,  Alexander C. Loui

发明人： Wei Jiang ,  Alexander C. Loui

IPC分类号： H04N5/91

CPC分类号： G06K9/00718 ,  G06K9/00744 ,  G06K9/4676

摘要： A method for determining a semantic concept classification for a digital video clip based on a grouplet dictionary that includes a plurality of temporally-correlated grouplets. The temporally-correlated grouplets include textual codewords and either visual codewords or audio codewords, wherein the codewords in a particular temporally-correlated grouplet were determined to be correlated with each other based on analysis of a set of training videos. Reference video codeword similarity scores are determined for a set of reference video clips, and codeword similarity scores are determined for the digital video clip. A reference video similarity score is determined for each reference video clip representing a similarity between the digital video clip and the reference video clip based on the reference video codeword similarity scores, the codeword similarity scores, and the temporally-correlated grouplets. One or more semantic concept classifications are determined using trained semantic classifiers responsive to the determined reference video similarity scores.

摘要翻译： 一种用于基于包括多个时间相关小区的小组字典来确定数字视频剪辑的语义概念分类的方法。 时间相关小区包括文本码字和视觉码字或音频码字，其中基于一组训练视频的分析，将特定时间相关小区小区中的码字确定为彼此相关。 对于一组参考视频剪辑确定参考视频码字相似性分数，并且为数字视频剪辑确定码字相似性分数。 基于参考视频码字相似度分数，码字相似度分数和时间相关小区，为表示数字视频剪辑和参考视频剪辑之间的相似性的每个参考视频剪辑确定参考视频相似性分数。 响应于所确定的参考视频相似性分数，使用经过训练的语义分类器来确定一个或多个语义概念分类。