会员体验
专利管家(专利管理)
工作空间(专利管理)
风险监控(情报监控)
数据分析(专利分析)
侵权分析(诉讼无效)
联系我们
交流群
官方交流:
QQ群: 891211   
微信请扫码    >>>
现在联系顾问~
热词
    • 1. 发明专利
    • Input prediction device
    • 输入预测装置
    • JP2009301198A
    • 2009-12-24
    • JP2008153078
    • 2008-06-11
    • Mitsubishi Electric Corp三菱電機株式会社
    • OKADA YASUHIROYAMAZAKI KO
    • G06F17/22
    • PROBLEM TO BE SOLVED: To provide an input prediction device facilitating predictive input of a desired word by reducing the number of character input of reading and the number of times of operation of the selective input of candidates. SOLUTION: On the basis of word notations matched with a portion of the reading of an input word and the reading and a matched place, the notations of partial words with which a portion of the reading of the word is matched are extracted from a dictionary for input predictive conversion, and the selective input of the candidates of the partial words is accepted by displaying the extracted partial words as candidates, and when it is not possible to uniquely determine a desired word from the candidates of the selectively input partial words, the partial words in the word including the selectively input partial words are successively extracted, and the selectively input of the candidates is successively performed so that the desired word can be determined. COPYRIGHT: (C)2010,JPO&INPIT
    • 要解决的问题:通过减少读取的字符输入的数量和候选的选择性输入的操作次数来提供便于对所需字的预测输入的输入预测装置。 解决方案:根据与输入字的读取部分以及阅读和匹配位置匹配的字符号,从单词的一部分读取匹配的部分词的符号从 用于输入预测转换的字典,并且通过将所提取的部分字显示为候选来接受部分字的候选的选择性输入,并且当不可能从选择性输入的部分字的候选中唯一地确定期望的字时 ,连续地提取包括有选择地输入的部分字的单词中的部分词,并且连续执行候选的选择性输入,从而可以确定期望的单词。 版权所有(C)2010,JPO&INPIT
    • 2. 发明专利
    • Image processor
    • 图像处理器
    • JP2009122990A
    • 2009-06-04
    • JP2007296675
    • 2007-11-15
    • Mitsubishi Electric Corp三菱電機株式会社
    • YONEYAMA SHOGOMIYAHARA HIROYASUOKADA YASUHIRO
    • G06T1/00G06T7/20H04N7/18
    • PROBLEM TO BE SOLVED: To solve the problem that when a moving object is detected by comparing a photographic image with a prepared photographic image in an area other than a mask area where the moving object is not detected, the moving object cannot be detected, which is located within a predetermined mask area for a moving range of a blocking object.
      SOLUTION: A moving area is detected by moving area detecting means from an image obtained by image recording means which records both the moving object and the blocking object. The blocking object is detected by blocking object detecting means in accordance with operation storing means for storing operation information in advance such as the moving range, the moving direction or the speed of the blocking object. The area of the detected blocking object is decided as the mask area and an area excluding the mask area is set as a moving object detecting area by moving object detecting area setting means. The moving object is detected by moving object detecting means from the moving object detecting area and the moving area detected by the moving area detecting means.
      COPYRIGHT: (C)2009,JPO&INPIT
    • 要解决的问题:为了解决通过将拍摄图像与未检测到移动物体的掩模区域以外的区域中的准备摄影图像进行比较来检测移动物体的问题,移动物体不能 被检测到,其位于用于阻挡对象的移动范围的预定掩模区域内。 解决方案:通过移动区域检测装置从记录移动物体和阻挡物体的图像记录装置获得的图像检测移动区域。 根据用于存储诸如移动范围,移动方向或阻挡物体的速度的预先存储操作信息的操作存储装置,通过阻挡物体检测装置来检测阻挡物体。 决定检测到的阻挡对象的区域为掩模区域,通过移动物体检测区域设定装置将除掩模区域之外的区域设定为移动物体检测区域。 通过移动物体检测装置从移动物体检测区域和由移动区域检测装置检测到的移动区域来检测移动物体。 版权所有(C)2009,JPO&INPIT
    • 3. 发明专利
    • Vehicle type determination device
    • 车辆类型确定装置
    • JP2008146154A
    • 2008-06-26
    • JP2006329543
    • 2006-12-06
    • Mitsubishi Electric Corp三菱電機株式会社
    • MIYAHARA HIROYASUOKADA YASUHIRO
    • G08G1/015G06T1/00G08G1/04
    • PROBLEM TO BE SOLVED: To solve a problem in a vehicle type determination device in which its determination accuracy is worsened if using such determination information as it is since the location of a number plate is sometimes shifted from the center of its body when using the number plate and windshield easy to acquire their accurate locations for determining the type of a vehicle.
      SOLUTION: A number plate detecting means detects the number plate from an input image of an image input means, and outputs its coordinates information. A windshield detecting means detects the windshield, and outputs its coordinates information. A number plate position determination means estimates a shifted amount by determining whether the position of the number plate detected by the number plate position determination means is shifted from the center of its body. A shape comparison means compensates the coordinates information of the number plate detected by the shape comparison means with an estimated shifted amount. The compensated coordinates and a shape dictionary in which the coordinates information indicating positional relationship between the number plate and the windshield by using the coordinates of the detected windshield, are compared with each other, and a vehicle type determination means determines the type of a vehicle, based on the compared results.
      COPYRIGHT: (C)2008,JPO&INPIT
    • 要解决的问题:为了解决如果使用这样的确定信息,其确定精度恶化的车辆类型确定装置中的问题,因为牌子板的位置有时从其身体的中心偏移, 使用号牌和挡风玻璃容易获得用于确定车辆类型的准确位置。 数字板检测装置根据图像输入装置的输入图像检测印版,并输出其坐标信息。 挡风玻璃检测装置检测挡风玻璃,并输出其坐标信息。 数字板位置确定装置通过确定由铭牌位置确定装置检测到的号牌的位置是否从其主体的中心偏移来估计移位量。 形状比较装置以估计的偏移量补偿由形状比较装置检测到的印版的坐标信息。 补偿坐标和形状字典,其中通过使用检测到的挡风玻璃的坐标来指示号牌和挡风玻璃之间的位置关系的坐标信息彼此进行比较,并且车辆类型确定装置确定车辆的类型, 基于比较结果。 版权所有(C)2008,JPO&INPIT
    • 5. 发明专利
    • Character recognition device
    • 字符识别装置
    • JP2003346081A
    • 2003-12-05
    • JP2002155928
    • 2002-05-29
    • Mitsubishi Electric Corp三菱電機株式会社
    • MIYAHARA HIROYASUOKADA YASUHIRO
    • G06K9/20G06K9/62G06K9/72
    • PROBLEM TO BE SOLVED: To accurately recognize characters partly hidden by a neutral-colored object.
      SOLUTION: This character recognition device includes a hidden area determining means which cuts out a character pattern having a hidden area expressed by multi-gradient pixels, divides the character pattern into areas, extracts the concentration characteristic quantity of each of the areas, and determines whether or not each of the areas is included in the hidden area according to the concentration characteristic quantity; a recognition dictionary which stores a reference pattern for the collation of character recognition; and a character recognition means which collates the reference pattern with the character pattern as to the areas except the area included in the hidden area and then outputs characters according to the result.
      COPYRIGHT: (C)2004,JPO
    • 要解决的问题:准确识别中性色物体部分隐藏的字符。 解决方案:该字符识别装置包括:隐藏区域确定装置,其切出具有由多梯度像素表示的隐藏区域的字符图案,将字符图案划分为区域,提取每个区域的浓度特征量; 并且根据浓度特征量确定每个区域是否包括在隐藏区域中; 识别字典,其存储用于字符识别的对照的参考图案; 以及字符识别装置,其将参考图案与除了隐藏区域中包括的区域之外的区域的字符图案进行整理,然后根据结果输出字符。 版权所有(C)2004,JPO
    • 6. 发明专利
    • Document retrieval device
    • 文件检索装置
    • JP2010146509A
    • 2010-07-01
    • JP2008326286
    • 2008-12-22
    • Mitsubishi Electric Corp三菱電機株式会社
    • HIRANO TAKASHIOKADA YASUHIRO
    • G06F17/30G06K9/00
    • PROBLEM TO BE SOLVED: To solve such the problem that in the character code normalization of various kinds of documents such as an electronic document, and image document, a document of mixture of these documents, a plurality of methods are used, but a retrieval score having hit by a character code after normalization cannot be made to correspond to respective normalization methods.
      SOLUTION: A document retrieval device includes an equivalent character table, a variant-different pronunciation character table, a decomposition/composition character table, and a same form character table. The device normalizes the character code by referring to a table other than the same form character table with respect to a character code extracted from a retrieval document by an electronic text extraction part and referring to all the tables with respect to a character code with a candidate character which is subjected to character recognition by a character recognition part, calculates coordinate position information of a decomposed and integrated character upon normalization from information before decomposition and integration, generates a retrieval index corresponding to the normalized character code by a retrieval index generation part, normalizes a retrieval keyword with all the table by a retrieval processing part and outputs the result of collation together with the retrieval index.
      COPYRIGHT: (C)2010,JPO&INPIT
    • 要解决的问题为了解决这样的问题,即在诸如电子文档和图像文档的各种文档的字符代码归一化中,使用这些文档的混合的文档,使用多种方法,但是 在归一化之后由字符代码命中的检索分数不能对应于各自的标准化方法。 解决方案:文档检索装置包括等效字符表,变体不同的发音字符表,分解/合成字符表和相同的表单字符表。 该装置通过参照除电子文本提取部分从检索文档提取的字符代码的相同格式字符表以外的表格来对字符代码进行标准化,并且参考关于具有候选的字符代码的所有表 通过字符识别部进行字符识别的字符在归一化之前从分解和积分之前的信息计算出分解和整合字符的坐标位置信息,通过检索索引生成部分生成与归一化字符代码对应的检索索引,将其归一化 通过检索处理部分具有所有表的检索关键字,并将检索结果与检索索引一起输出。 版权所有(C)2010,JPO&INPIT
    • 7. 发明专利
    • Motion recognition device
    • 运动识别装置
    • JP2008310588A
    • 2008-12-25
    • JP2007157769
    • 2007-06-14
    • Mitsubishi Electric Corp三菱電機株式会社
    • YONEYAMA SHOGOMIYAHARA HIROYASUKAWAMATA TAKENORIOKADA YASUHIRO
    • G06F3/033G06F3/038
    • PROBLEM TO BE SOLVED: To efficiently recognize motion with high accuracy when there are many kinds of motion of a recognition target or even when setting a plurality of pieces of similar motion as the recognition target. SOLUTION: This motion recognition device is provided with: an acceleration characteristic point detection part 2 analyzing a temporal change of acceleration of each shaft of an identification target device to detect a characteristic point of the acceleration of each the shaft; and an acceleration characteristic point storage part 3 storing the characteristic point of the temporally changing acceleration of each the shaft in each of the plurality of pieces of the recognition target motion. In the motion recognition device, a motion search part 4 collates the characteristic point of the acceleration of each the shaft detected by the acceleration characteristic point detection part 2 and the characteristic point of the acceleration of each the shaft stored in the acceleration characteristic point storage part 3, and searches for the recognition target motion according with operation of the identification target device. COPYRIGHT: (C)2009,JPO&INPIT
    • 要解决的问题:当识别目标有许多种类的运动或者甚至当设置多个类似的运动作为识别目标时,能够高精度地识别运动。 解决方案:该运动识别装置具有:加速度特征点检测部分2,分析识别目标装置的每个轴的加速度的时间变化,以检测每个轴的加速度的特征点; 以及加速度特征点存储部3,其存储在多个识别目标运动中的每一个中存储每个轴的时间上变化的加速度的特征点。 在运动识别装置中,运动检索部4将由加速度特性点检测部2检测出的各轴的加速度的特征点与存储在加速度特征点存储部中的各轴的加速度的特征点进行比较 根据识别目标装置的操作,搜索识别目标运动。 版权所有(C)2009,JPO&INPIT
    • 8. 发明专利
    • Character reader
    • 字符阅读器
    • JP2007172227A
    • 2007-07-05
    • JP2005368004
    • 2005-12-21
    • Mitsubishi Electric Corp三菱電機株式会社
    • MIYAHARA HIROYASUOKADA YASUHIRO
    • G06K9/32G06T7/60
    • PROBLEM TO BE SOLVED: To provide a character reader, for reading a character of an image taken at an extremely close range by correcting a trapezoidally distorted area into a rectangular shape.
      SOLUTION: This character reader comprises an inclination measuring means determining an inclination of a character in an input image; an inclination calculation expression determination means for determining, from the inclination of the character determined by the measuring means and the position coordinate of the character, a relational expression showing the relation between the position coordinate and the inclination; a distortion correction means for determining, based on the relational expression determined by the determination means, a trapezoidal area, and converting the trapezoidal area into a rectangular shape; and a character recognition means for extracting and recognizing a character pattern from an image in the area corrected by the correction means.
      COPYRIGHT: (C)2007,JPO&INPIT
    • 要解决的问题:为了提供字符读取器,用于通过将梯形失真区域校正为矩形来读取在极近距离处拍摄的图像的字符。 解决方案:该字符读取器包括确定输入图像中的字符的倾斜度的倾斜测量装置; 倾斜计算表达式确定装置,用于根据由测量装置确定的角色和角色的位置坐标的倾斜度确定表示位置坐标和倾斜度之间的关系的关系式; 失真校正装置,用于基于由确定装置确定的关系表达式确定梯形区域,并将梯形区域转换为矩形; 以及字符识别装置,用于从由校正装置校正的区域中的图像中提取和识别字符图案。 版权所有(C)2007,JPO&INPIT
    • 9. 发明专利
    • Document classification device
    • 文件分类装置
    • JP2006040058A
    • 2006-02-09
    • JP2004220666
    • 2004-07-28
    • Mitsubishi Electric Corp三菱電機株式会社
    • KAWAMURA HIDEOOKADA YASUHIRO
    • G06F17/30
    • PROBLEM TO BE SOLVED: To solve the following problems of a conventional document classification device by a document vector: classification is erroneously performed by accumulating even a word of remarks or an additional note in the document vector, and operation such as a change of an attribute vector value of a word table is required when performing classification attaching great importance to a different viewpoints.
      SOLUTION: This document classification device has: a document retrieval means retrieving a document from a document database storing documents by a retrieval keyword designated by a user, and outputting the retrieved document; an attribute extraction means extracting a character string from record contents of the document outputted by the document retrieval means, extracting an attribute of an attribute term and an attribute name of the attribute term from the character string on the basis of a prescribed attribute extraction definition, and outputting it; and a document classification means generating the document vector on the basis of the attribute outputted by the attribute extraction means, the record contents of the document, and the retrieval keyword, and classifying the document into a plurality of categories by a distance between the document vectors.
      COPYRIGHT: (C)2006,JPO&NCIPI
    • 要解决的问题:为了通过文档向量来解决常规文档分类装置的以下问题:通过在文档向量中累加甚至一个备注或附加的注释的字来错误地执行分类,并且诸如改变的操作 当对不同的视点进行分类时,需要单词表的属性向量值。 解决方案:本文档分类装置具有:文档检索装置,通过由用户指定的检索关键字从存储文档的文档数据库中检索文档,并输出检索到的文档; 属性提取装置,从文件检索装置输出的文档的记录内容中提取字符串,根据规定的属性提取定义,从字符串中提取属性项的属性和属性项的属性名, 并输出; 文档分类装置根据由属性提取装置输出的属性,文档的记录内容和检索关键字生成文档向量,并将该文档分类成文档向量之间的距离 。 版权所有(C)2006,JPO&NCIPI