专利快速检索-快速检索全球专利，免费商用专利数据库-IPRDB

41. 发明公开

EP3163498A3 ALARMING METHOD AND DEVICE 有权
标题翻译：警报方法和装置
公开(公告)号：EP3163498A3
公开(公告)日：2017-07-26
申请号：EP16162165.1
申请日：2016-03-24
申请人： Xiaomi Inc.
发明人： ZHANG, Tao , CHEN, Zhijun , WANG, Pingze
IPC分类号： G06K9/00 , G08B13/196 , G06K9/46 , G06K9/62
摘要： Described are an alarming method and device, belonging to the technical field of Internet. The method includes: acquiring a monitoring video (201); determining whether a monitoring target exists in a sensitive area of the monitoring video (202); and sending alarming information to a terminal to enable the terminal to perform alarming when the monitoring target exists in the sensitive area (203).
摘要翻译：所描述的是一种报警方法和装置，属于互联网技术领域。该方法包括：获取监控视频（201）; 确定监视目标是否存在于监视视频的敏感区域（202）中; 并在监测目标存在于敏感区域时向终端发送告警信息，使终端能够进行告警。

42. 发明公开

EP3173977A1 METHOD AND DEVICE FOR FEATURE EXTRACTION 审中-公开
标题翻译：用于特征提取的方法和设备
公开(公告)号：EP3173977A1
公开(公告)日：2017-05-31
申请号：EP16199669.9
申请日：2016-11-18
申请人： Xiaomi Inc.
发明人： LONG, Fei , CHEN, Zhijun , ZHANG, Tao
IPC分类号： G06K9/46 , G06K9/62
CPC分类号： G06K9/4642 , G06K9/4647 , G06K9/6249 , G06K2009/4695 , G06K2009/485 , G06T7/0081
摘要： Method and device for feature extraction are provided in the disclosure. The method may include: partitioning (102) an image into a plurality of blocks, wherein each of the blocks includes a plurality of cells; performing (104) a sparse signal decomposition on each of the cells using a predetermined dictionary D to obtain a sparse vector corresponding to each of the cells respectively; and extracting (106) a HOG feature of the image according to the sparse vectors. The following fact may be solved that the process of the extraction of the HOG feature is directly calculated with respect to the space domain of the image, which causes that detection rate and accuracy are relatively low in pattern recognition; the effect may be reached that the detection rate and accuracy are improved in pattern recognition by extracting the HOG feature of the image in the frequency domain.
摘要翻译：本发明提供了用于特征提取的方法和设备。该方法可以包括：将图像分割（102）成多个块，其中每个块包括多个单元; 使用预定字典D对每个单元执行（104）稀疏信号分解以获得分别对应于每个单元的稀疏矢量; 以及根据所述稀疏矢量提取（106）所述图像的HOG特征。可以解决以下事实：HOG特征的提取过程直接针对图像的空间域进行计算，这导致模式识别中的检测率和准确度相对较低; 可以达到通过在频域中提取图像的HOG特征来提高模式识别中的检测率和准确度的效果。

43. 发明公开

EP3173976A1 METHOD AND DEVICE FOR EXTRACTING CHARACTERISTIC 审中-公开
标题翻译：用于提取特性的方法和装置
公开(公告)号：EP3173976A1
公开(公告)日：2017-05-31
申请号：EP16199236.7
申请日：2016-11-17
申请人： Xiaomi Inc.
发明人： LONG, Fei , CHEN, Zhijun , ZHANG, Tao
IPC分类号： G06K9/46 , G06K9/52
CPC分类号： G06T5/40 , G06K9/4642 , G06K9/522 , G06K9/6298 , G06T5/001 , G06T5/10 , G06T7/0081 , G06T2207/20021 , G06T2207/20052 , G06T2207/20056
摘要： Method and device for extracting characteristics are provided in the disclosure, which belongs to the technical field of image processing. The method for extracting characteristics includes: segmenting (102) an image into a plurality of blocks, wherein each block comprises a plurality of cells; converting (104) each cell from a spatial domain to a frequency domain; and extracting (106) a Histogram of Oriented Gradient (HOG) characteristic of the image in the frequency domain..
摘要翻译：本发明提供了一种提取特征的方法和装置，属于图像处理技术领域。用于提取特征的方法包括：将图像分割（102）成多个块，其中每个块包括多个单元; 将每个小区从空间域转换（104）到频域; 以及在频域中提取（106）图像的定向梯度直方图（HOG）特征。

44. 发明公开

EP3171308A1 INFORMATION PROCESSING METHOD, INFORMATION PROCESSING APPARATUS, TERMINAL, AND SERVER 审中-公开
标题翻译：信息管理系统，信息管理系统
公开(公告)号：EP3171308A1
公开(公告)日：2017-05-24
申请号：EP16174298.6
申请日：2016-08-10
申请人： Xiaomi Inc.
发明人： WANG, Pingze , ZHANG, Tao , CHEN, Zhijun
IPC分类号： G06Q10/10 , G06Q10/02
CPC分类号： G06F17/30864 , G06F17/30312 , G06Q10/02 , G06Q10/1093
摘要： The present invention is related to an information processing method, an information processing apparatus, a terminal, and a server. The method comprises: when information is received, judging whether the information is a predetermined type of information; when the information is the predetermined type of information, matching the predetermined type of information with a key word template; extracting key information in the information based on the successfully matched key word template. By employing the embodiments of the present invention, it is able to accurately identify and extract key information such as train number, seat, departure/arrival station, departure/arrival time, order number, passenger from ticket information in which the key information is distributed very intensively, so as to remind the user based on the key information to thereby improve the user experience.
摘要翻译：本发明涉及信息处理方法，信息处理装置，终端和服务器。该方法包括：当接收到信息时，判断该信息是否是预定类型的信息; 当信息是预定类型的信息时，将预定类型的信息与关键字模板相匹配; 根据成功匹配的关键字模板提取信息中的关键信息。通过采用本发明的实施例，能够准确地识别和提取关键信息，例如列车号，座位，出发/到达站，出发/到达时间，订单号，乘客从密钥信息分发的票信息非常集中，以便基于关键信息提醒用户，从而改善用户体验。

45. 发明公开

EP3163504A1 METHOD, DEVICE AND COMPUTER-READABLE MEDIUM FOR REGION EXTRACTION 有权
标题翻译：用于区域提取的方法，装置和计算机可读介质
公开(公告)号：EP3163504A1
公开(公告)日：2017-05-03
申请号：EP16195231.2
申请日：2016-10-24
申请人： Xiaomi Inc.
发明人： LONG, Fei , ZHANG, Tao , CHEN, Zhijun
IPC分类号： G06K9/00 , G06K9/20 , G06K9/46 , G06K9/34
CPC分类号： G06K9/4604 , G06K9/00442 , G06K9/00449 , G06K9/2054 , G06K9/34 , G06K9/4647 , G06T7/11
摘要： Method, device and computer-readable medium for region extraction are provided in the disclosure. The method may comprise: obtaining (101) a region position of a first information region in a credential image; determining (102) a second information region according to the region position of the first information region; and segmenting (103) the second information region to obtain at least one character zone. The problems that in the related art the recognition of some information regions in a credential image photographed directly is difficult to implement and the positioning of some information regions is inaccurate, may be solved. The following effects may be reached that determining the second information region based on the region position of the first information region in a credential image, and segmenting the second information region, thus precisely positioning the second information region and recognizing the character zone in the second information region.
摘要翻译：本发明提供了用于区域提取的方法，设备和计算机可读介质。该方法可以包括：获得（101）凭证图像中的第一信息区域的区域位置; 根据所述第一信息区域的区域位置确定（102）第二信息区域; 以及分割（103）第二信息区域以获得至少一个字符区域。现有技术中直接拍摄的凭证图像中某些信息区域的识别难以实现，某些信息区域的定位不准确的问题可以得到解决。可以达到以下效果：基于凭证图像中的第一信息区域的区域位置来确定第二信息区域，并且分割第二信息区域，从而精确地定位第二信息区域并且识别第二信息中的字符区域地区。

46. 发明公开

EP3160105A1 METHOD AND DEVICE FOR PUSHING INFORMATION 审中-公开
标题翻译：用于推送信息的方法和设备
公开(公告)号：EP3160105A1
公开(公告)日：2017-04-26
申请号：EP16159949.3
申请日：2016-03-11
申请人： Xiaomi Inc.
发明人： JI, Xinhua , LI, Haifeng , CHEN, Zhijun
IPC分类号： H04L29/06 , H04L29/08
CPC分类号： H04L67/26 , H04L65/1059 , H04L67/22
摘要： Method, device, apparatus and system for pushing information are provided in the disclosure. The method may include: acquiring (101) target information stored locally in a terminal; determining (102) an interest parameter based on the target information; transmitting (103) to the terminal push information corresponding to the interest parameter. Information push may be practiced based on the interest parameter determined by the present disclosure.
摘要翻译：本发明提供了推送信息的方法，设备，装置和系统。该方法可以包括：获取（101）本地存储在终端中的目标信息; 基于目标信息确定（102）兴趣参数; 发送（103）到与兴趣参数对应的终端推送信息。信息推送可以基于由本公开确定的兴趣参数来实施。

47. 发明公开

EP3136258A1 TRANSLATION STRING AND CONTEXT SAVING IN TRANSLATION MEMORY DATABASE 审中-公开
标题翻译：欧洲腓特烈华星报
公开(公告)号：EP3136258A1
公开(公告)日：2017-03-01
申请号：EP16162185.9
申请日：2016-03-24
申请人： Xiaomi Inc.
发明人： CHEN, Zhijun , WANG, Pingze , ZHANG, Tao
IPC分类号： G06F17/28 , G06F17/30 , G06F3/048
CPC分类号： G06F17/2205 , G06F3/048 , G06F3/0481 , G06F17/2705 , G06F17/28 , G06F17/289
摘要： The disclosure relates to method and device for string saving, which belongs to field of network technology. The method includes: acquiring (101, 201) a designated string to be saved; acquiring (102, 202) a translation result of the designated string; acquiring (103, 203) context information of the designated string; saving (104, 204) the designated string, the translation result of the designated string and the context information of the designated string in a preset string database in a correspondence manner. The disclosure enables saving of a designated string and context information thereof at the same time such that the designated string and the context information thereof can be displayed at the same time when an instruction of viewing the string saved in the preset string database from a user is detected, and thus prompt and learning effect can be improved and the user is able to quickly remind the string and its use scenes. (FIG.1)
摘要翻译：本公开涉及字符串保存的方法和装置，属于网络技术领域。该方法包括：获取（101,201）要保存的指定字符串; 获取（102,202）指定字符串的翻译结果; 获取（103,203）指定字符串的上下文信息; 以指定的方式保存（104,204）指定字符串，指定字符串的翻译结果和指定字符串的上下文信息。本公开使得能够同时保存指定的字符串及其上下文信息，使得当从用户观看保存在预设字符串数据库中的字符串的指令是同时显示指定字符串及其上下文信息时检测到，从而可以提高和学习效果，并且用户能够快速提醒字符串及其使用场景。（图。1）

48. 发明公开

EP3125162A1 METHOD AND DEVICE FOR MARKING UNKNOWN NUMBER 审中-公开
标题翻译： VERFAHREN UND VORRICHTUNG ZUR MARKIERUNG EINER UNBEKANNTEN ZAHL
公开(公告)号：EP3125162A1
公开(公告)日：2017-02-01
申请号：EP16176963.3
申请日：2016-06-29
申请人： Xiaomi Inc.
发明人： CHEN, Zhijun , LONG, Fei , ZHANG, Tao
IPC分类号： G06Q10/00 , G06Q50/00
CPC分类号： H04M3/42068 , G06Q10/00 , G06Q10/10 , G06Q50/01 , H04L67/306 , H04W4/16
摘要： Described are a method and device for marking a strange number. The method includes: an information acquisition request including a first phone number which is sent from a terminal is received, wherein the information acquisition request is sent from the terminal when it is detected that the first phone number of a correspondent node communicating with the terminal is a strange number (201); according to a preset corresponding relationship between a phone number and a social account, a first social account corresponding to the first phone number is queried (202); it is detected whether the first social account is a friend account of a second social account, wherein the second social account is a social account corresponding to a second phone number of the terminal (203); when the first social account is the friend account, account information of the first social account is obtained and sent to the terminal (204).
摘要翻译：描述了用于标记奇数的方法和装置。该方法包括：接收到包括从终端发送的第一电话号码的信息获取请求，其中当检测到与终端通信的对方节点的第一电话号码是一个奇怪的数字（201）; 根据电话号码和社交账户之间的预设对应关系，查询对应于第一电话号码的第一社交账户（202）; 检测第一社交帐户是否是第二社交账户的朋友帐户，其中第二社交账户是与终端的第二电话号码对应的社交账户（203）; 当第一社交帐户是朋友账户时，获得第一个社交账户的账户信息并将其发送到终端（204）。

49. 发明公开

EP3125158A2 METHOD AND DEVICE FOR DISPLAYING IMAGES 有权
标题翻译： VERFAHREN UND VORRICHTUNG ZUR ANZEIGE VON BILDERN
公开(公告)号：EP3125158A2
公开(公告)日：2017-02-01
申请号：EP16181425.6
申请日：2016-07-27
申请人： Xiaomi Inc.
发明人： ZHANG, Tao , LONG, Fei , CHEN, Zhijun
IPC分类号： G06K9/00 , G06K9/03
CPC分类号： G06K9/00624 , G06K9/00281 , G06K9/00677 , G06K9/036 , G06K9/4661 , H04N1/40012
摘要： The present disclosure relates to a method and a device for displaying images, which pertains to the field of terminal technology. The method includes: when receiving an instruction for opening an album, clustering (101, 201) similar images in the album to obtain a plurality of groups of images; selecting (102) an image from each of the groups of images as a representative image of the group of images; and displaying (103, 206) the representative images of the groups of images. Thereby, it can improve the display effect of the representative images of the groups of images.
摘要翻译：本公开涉及一种用于显示与终端技术领域有关的图像的方法和装置。该方法包括：当接收到打开专辑的指令时，在相册中聚类（101,201）相似图像以获得多组图像; 从所述图像组中的每一组中选择（102）图像作为所述图像组的代表图像; 以及显示（103,206）图像组的代表图像。从而可以提高图像组的代表图像的显示效果。

50. 发明公开

EP3125135A1 PICTURE PROCESSING METHOD AND DEVICE 有权
标题翻译：图像处理方法和装置
公开(公告)号：EP3125135A1
公开(公告)日：2017-02-01
申请号：EP16161513.3
申请日：2016-03-22
申请人： Xiaomi Inc.
发明人： ZHANG, Tao , CHEN, Zhijun , LONG, Fei
IPC分类号： G06F17/30
CPC分类号： G06K9/6215 , G06F17/30247 , G06F17/30265 , G06F17/30274 , G06F17/30551 , G06K9/46
摘要： An embodiment herein relates to a picture processing method and device, belonging to the field of terminal technology. The method includes that: during memory space management, all pictures in a picture directory are scanned (101, 203); at least one group of similar pictures is generated according to information on an attribute of each picture and pre-extracted picture features of the each picture (102, 204), the information on the attribute including at least a photographing time; pictures in each group of similar pictures are displayed in units of groups of similar pictures in an order of photographing time (103, 205); and a picture in each group of similar pictures is processed according to a detected operation (104, 206).
摘要翻译：这里的一个实施例涉及一种属于终端技术领域的图片处理方法和设备。该方法包括：在内存空间管理期间，扫描图片目录中的所有图片（101,203）; 根据关于每个图片的属性和每个图片（102,204）的预先提取的图片特征的信息生成至少一组相似图片，关于该属性的信息至少包括拍摄时间; 按照拍摄时间的顺序以相似图片组为单位显示每组相似图片中的图片（103,205）; 并根据检测到的操作处理每组相似图片中的图片（104,206）。

你已经成功收藏专利！

检索式保存成功!

IPRDB

热门服务

关于我们

友情链接

联系方式