公开(公告)号：EP0560587A2

公开(公告)日：1993-09-15

申请号：EP93301810.3

申请日：1993-03-10

申请人： HITACHI, LTD.

发明人： Abe, Masahiro Hitachi Europe Ltd.Research and Deve ,  Sakou, Hiroshi ,  Sagawa, Hirohiko ,  Nitin, Indurkhya, c/o Telecom Research Lab.

IPC分类号： G09B21/00

CPC分类号： G06F17/2872 ,  G06F17/2881 ,  G06K9/00335 ,  G06K9/6293 ,  G06K9/685 ,  G09B19/06 ,  G09B21/00 ,  G09B21/009

摘要： A sign language translation system and method not only recognizes words of a sign language but also supplements omitted words between the words of the sign language, to thereby generate a spoken language. The sign language translation system has an input unit (105) for inputting at least the motion of hands, a language generating unit (102) responsive to the inputted motion of hands for recognizing the words corresponding to the motion of hands and generating a spoken language using the relationship between the recognized words, and an output unit (103) for outputting the generated spoken language. The sign language translation system and method can translate a sign language into an easy-to-understand spoken language.

摘要翻译： 手语翻译系统和方法不仅能识别手语的单词，而且可以补充手语之间的省略词，从而产生口语。 手语翻译系统具有用于至少输入手的运动的输入单元（105），响应于所输入的手的运动来识别与手的运动相对应的单词并产生口语的语言生成单元（102） 使用识别词之间的关系，以及用于输出生成的口语的输出单元（103）。 手语翻译系统和方法可以将手语翻译成易于理解的口语。

公开(公告)号：EP0419259A3

公开(公告)日：1992-12-02

申请号：EP90310300.0

申请日：1990-09-20

申请人： NEC CORPORATION

发明人： Murai, Yasumasa, NEC Corporation ,  Yamauchi, Toshifumi, NEC Corporation

IPC分类号： G06K9/68

CPC分类号： G06K9/685

摘要： On recognizing first through P-th input patterns as first through P-th recognized patterns (P being greater than one) by using a plurality of reference patterns and a plurality of memorized identifiers identifying the reference patterns, a processing circuit (35) processes the first through the P-th input patterns into first through Q-th provisional patterns different from one another, where Q is not greater than the above-mentioned plurality. A calculating circuit (50) calculates dissimilarity degrees had by the reference patterns relative to the first through the Q-th provisional patterns to select, with repetition allowed as selected patterns, the reference patterns which have the dissimilarity degrees less than a predetermined degree. The calculating circuit thereby selects R particular identifiers and S different identifiers from the memorized identifiers, where R plus S is not greater than P and is not less than the plurality. Each particular identifier and each different identifier identify only one pattern and one of at least two patterns of the selected patterns. The other at least one of the above-mentioned at least two patterns is identified by at least one of the R particular identifiers. A determining circuit (60) determines first through P-th output identifiers identifying the first through the P-th recognized patterns by using the R particular idenifiers as ones, R in number, and others, S in number, of the first through the P-th output identifiers.

公开(公告)号：EP0480806A1

公开(公告)日：1992-04-15

申请号：EP91402654.7

申请日：1991-10-04

申请人： THOMSON-CSF

发明人： Ferre, Alain ,  Villalon, Yannick

IPC分类号： G06K9/52 ,  G06K9/68

CPC分类号： G06K9/52 ,  G06K9/685 ,  G06K2009/3291

摘要： Le procédé de reconnaissance d'objets dans des images selon l'invention comporte une phase d'apprentissage au cours de laquelle la zone objet à reconnaître sélectionnée dans une image d'apprentissage est modélisée, et une phase de reconnaissance au cours de laquelle la zone modélisée est comparée à des zones image de même taille prises dans les images à traiter. Le procédé utilise, pour la caractérisation des zones d''images modèle d'objet à reconnaître ou zones d'images à traiter, une décomposition des matrices des valeurs de luminance en valeurs singulières selon un produit matriciel de matrices de projection et d'une matrice diagonale ordonnée, cette dernière constituant une signature caractéristique de la zone considérée ; pour la phase de reconnaissance, des écarts entre la signature de la zone modèle et les signatures de la zones de l'image considérée sont calculés et des zones d'image sont sélectionnées à partir des moindres écarts selon un procédé itératif qui consiste à effectuer un tri grossier sur un grand nombre de zones dans un premier temps, puis un tri de plus en plus fin sur un nombre de plus en plus restreint de zones. Les calculs effectués à chaque étape sont réutilisés à l'étape suivante.

摘要翻译： 用于识别在图像中的对象雅丁到本发明的方法包括其中的过程中学习阶段被在学习图像中识别选择的被建模对象区域，并且其中所述模型化区域的过程中的识别相相比图像 在图像中拍摄的相同大小的区域进行处理。 的方法utilisés，对于对象的模型的图像区域的特征被识别或要被处理的图像区域中，亮度值转换为奇异值的矩阵雅丁到投影矩阵的，纵坐标对角矩阵的矩阵积的分解， 构成有问题的区域的特性的签名的后期; 用于识别阶段，模型区域的签名和所讨论的图像的区域的签名之间的偏差被计算并图像区域在执行粗略哪besteht迭代方法的至少偏差的基础上选择的雅定到 在排序在首位的大量面积，比越来越细的那种在地区一个小数目。 在接下来的步骤在每个步骤中进行的计算被再次使用。 ... ...

公开(公告)号：EP0457534A2

公开(公告)日：1991-11-21

申请号：EP91304283.4

申请日：1991-05-13

申请人： CANON KABUSHIKI KAISHA

发明人： Ikeda, Hiroaki, Canon Kabushiki Kaisha

IPC分类号： G06K9/68

CPC分类号： G06K9/685 ,  G06K9/6807

摘要： An image processing apparatus comprises: a character area extractor to extract a character area from image information which is derived by reading an original document by an image scanner ; a position/size information memory to derive position information and size information of the image information in the character area extracted by the extractor; a comparator to compare the position and size information which are stored in the memory with predetermined standard information ; and a recognizer to decide a recognition character on the basis of a comparison result obtained by the comparator. The predetermined standard information indicates a ratio in the extracted character area. In the case of correcting the recognition character, the predetermined standard information is updated to a value which is compared by the comparator.

摘要翻译： 一种图像处理设备，包括：字符区域提取器，用于从通过图像扫描器读取原始文档得到的图像信息中提取字符区域; 位置/大小信息存储器，用于导出由提取器提取的字符区域中的图像信息的位置信息和大小信息; 比较器，用于将存储在存储器中的位置和大小信息与预定的标准信息进行比较; 以及识别器，用于基于由比较器获得的比较结果来确定识别字符。 预定的标准信息表示提取的字符区域中的比率。 在校正识别字符的情况下，将预定标准信息更新为由比较器比较的值。

公开(公告)号：EP0322920A3

公开(公告)日：1991-11-06

申请号：EP88121865.5

申请日：1988-12-29

申请人： SUMITOMO ELECTRIC INDUSTRIES LIMITED

发明人： Yamaguchi, Mikio c/o Osaka Works ,  Maeda, Naoki c/o Osaka Works

IPC分类号： G06K9/03

CPC分类号： G06K9/03 ,  G06K9/685

摘要： An apparatus and method for determining a character read by an optical character reader during a plurality of scans, whereby the character is determined from the plurality of scans by taking into consideration the tendency that the character will be misread as a second character. If there is no possibility that the second character could be interpreted as a misreading of the character which is read by the optical character reader, a "REJECT" or error signal may be generated, thus improving the reliability of determined characters.

公开(公告)号：EP0204516A2

公开(公告)日：1986-12-10

申请号：EP86304088.7

申请日：1986-05-29

申请人： ADEPT TECHNOLOGY, INC.

发明人： Roth, Scot D

IPC分类号： G06K9/48

CPC分类号： B25J9/1697 ,  G05B19/4182 ,  G06K9/48 ,  G06K9/685 ,  Y02P90/083

摘要： A practical vision system for controlling the positioning of a robot arm recognises and locates objects. The vision system (20, 30) processes binary images, but recognises objects based on boundary features such as lines, arcs, corners and holes instead of "blob features" such as area and best-fit ellipse. Consequently, the vision system can process two common situations not handled by blob analysis: merged blobs due to touching or overlapping parts and incomplete blobs due to low image contrast. The microprocessor-based system is interfaced to the robot system and can recognize up to five parts per second.

摘要翻译： 用于控制机器人手臂的定位的实用视觉系统识别和定位物体。 视觉系统（20,30）处理二进制图像，但是基于诸如线，弧，角和孔的边界特征来识别对象，而不是诸如面积和最佳拟合椭圆的“blob特征”。 因此，视觉系统可以处理不能通过斑点分析处理的两种常见情况：由于低图像对比度，由于触摸或重叠部分以及不完整的斑点而导致的合并斑点。 基于微处理器的系统与机器人系统接口，每秒可识别5个部分。

公开(公告)号：EP3349145A1

公开(公告)日：2018-07-18

申请号：EP18150346.7

申请日：2018-01-04

申请人： Samsung Electronics Co., Ltd.

发明人： YOO, Juwoan ,  LEE, Kwanghyun ,  LEE, Heejun ,  FARTUKOV, Alexei ,  ODINOKIKH, Gleb ,  GNATIUK, Vitalii ,  EREMEEV, Vladimir ,  SHIN, Dae-Kyu ,  PARK, Jeong-Min ,  PARK, Ji-Yoon

IPC分类号： G06K9/00

CPC分类号： G06K9/00604 ,  G06K9/00597 ,  G06K9/0061 ,  G06K9/00617 ,  G06K9/00926 ,  G06K9/685

摘要： An electronic device for iris recognition and an operating method thereof are provided. The electronic device includes a housing including a first surface, a display exposed through a first region of the first surface, a light emitting unit comprising light emitting circuitry disposed in a second region of the first surface, an image device comprising image acquiring circuitry disposed in a third region of the first surface, at least one processor disposed within the housing and electrically connected with the display, the light emitting unit and the image device, and a memory disposed within the housing and electrically connected with the at least one processor. The memory stores instructions that, when executed by the processor, cause the electronic device to store a reference template based on a first iris image which has been recognized using the light emitting unit and the image device, to authenticate a second iris image which has been recognized using the light emitting unit and the image device, using the stored reference template, and to store a template of the second iris image succeeded in authentication, as an additional template.

公开(公告)号：EP3029600A1

公开(公告)日：2016-06-08

申请号：EP14195670.6

申请日：2014-12-01

申请人： Axis AB

发明人： Bengtsson, Jesper ,  Johansson, Emanuel ,  Bylander, Mats ,  Winzell, Thomas

IPC分类号： G06K9/00 ,  G06K9/20 ,  G06K9/66 ,  G06K9/68

CPC分类号： G06K9/00771 ,  G06K9/2018 ,  G06K9/66 ,  G06K9/685

摘要： Shape detection is used in thermal images for detecting objects of one or more object classes. The temperature of a detected object is used to verify that the detection is correct, by determining if a temperature characteristic of a detected object is within a predetermined temperature interval of the relevant object class. Environmental conditions, such as rain, snow or fog, impacting the amount of thermal energy reaching a thermal camera from an object, are taken into account by collecting temperature statistics for the detected objects, and using these statistics to update the temperature interval.

摘要翻译： 形状检测用于热图像中，用于检测一个或多个对象类别的对象。 检测对象的温度通过确定检测对象的温度特性是否处于相关对象类的预定温度间隔内，用于验证检测是否正确。 通过收集检测对象的温度统计数据，并使用这些统计信息来更新温度间隔，考虑了影响从物体到达热敏相机的热能的数量，如雨，雪或雾等环境条件。

公开(公告)号：EP2764468A4

公开(公告)日：2015-11-18

申请号：EP12839143

申请日：2012-10-05

申请人： CIRECA THERANOSTICS LLC

发明人： REMISZEWSKI STANLEY H ,  THOMPSON CLAY M

IPC分类号： G06K9/00

CPC分类号： G06K9/00127 ,  G06K9/00147 ,  G06K9/6219 ,  G06K9/6262 ,  G06K9/685 ,  G06T7/0012 ,  G06T2207/30024

公开(公告)号：EP2674896B1

公开(公告)日：2015-11-18

申请号：EP12786171.4

申请日：2012-05-15

申请人： Panasonic Intellectual Property Management Co., Ltd.

发明人： SETOYAMA, Masahiro ,  HIROTOMI, Haruo ,  OBATA, Kazuhiko ,  FUJIWARA, Masanori ,  MORI, Akimichi ,  HATADA, Naoki

IPC分类号： G06K9/68 ,  B09B3/00 ,  B09B5/00 ,  G06K9/72

CPC分类号： B23P15/26 ,  B09B3/0058 ,  G06K9/685 ,  G06K9/723 ,  G06K2209/01 ,  Y02W30/82 ,  Y02W30/825 ,  Y10T29/49359 ,  Y10T29/53113