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
    • Vehicle detection device
    • 车辆检测装置
    • JP2014074939A
    • 2014-04-24
    • JP2012220429
    • 2012-10-02
    • Toshiba Corp株式会社東芝
    • AOKI YASUHIROSATO TOSHIOTAKAHASHI YUSUKEKARASAWA SHIGERU
    • G08G1/04G06T1/00H04N7/18
    • H04N7/18G06T7/254G06T2207/10021G06T2207/10028G06T2207/30236G08G1/04
    • PROBLEM TO BE SOLVED: To provide a vehicle detection device capable of accurately detecting a vehicle entering and exiting from a pickup image without being influenced by a change in optical conditions in a case where a stereo camera is adopted.SOLUTION: A vehicle detection device performs: enhancing edges in vertical directions of a vehicle to obtain parallax data between edge-enhanced pickup images; calculating variance of pixels within the pickup image to selectively discarding parallax data from the result; obtaining distance data from the parallax data to generate a distance image having a distance for each pixel; dividing the distance image into plural zones in a travel direction of the vehicle to generate plural reference sections; calculating a difference between background distance data and distance data for each of the plural reference sections; and calculating a change amount of the difference for each reference section and comparing the change amount of the difference with a threshold for each reference section, to thereby determine a state corresponding to the presence or absence of an article for each clock-time, retain the determination of the state for each clock-time, determine transition of the vehicle, and detect the vehicle entering and exiting.
    • 要解决的问题:提供一种车辆检测装置,其能够在采用立体照相机的情况下,不会受到光学条件的变化的影响,能够准确地检测进入和离开拾取图像的车辆。解决方案:车辆检测装置执行 :增强车辆的垂直方向的边缘,以获得边缘增强的拾取图像之间的视差信息; 计算拾取图像内的像素的方差以选择性地从结果中丢弃视差数据; 从所述视差数据获取距离数据以产生具有每个像素的距离的距离图像; 将距离图像沿车辆的行进方向分割成多个区域以生成多个基准部分; 计算所述多个参考部分中的每一个的背景距离数据和距离数据之间的差; 并且计算每个参考部分的差异的变化量,并且将每个参考部分的差异的变化量与阈值进行比较,从而确定与每个时钟时间有无物品相对应的状态, 确定每个时钟时间的状态,确定车辆的过渡,并检测车辆进入和离开。
    • 2. 发明专利
    • Image processing device and image processing method
    • 图像处理装置和图像处理方法
    • JP2013074321A
    • 2013-04-22
    • JP2011209658
    • 2011-09-26
    • Toshiba Corp株式会社東芝
    • KIMIYAMA KENJISATO TOSHIOSUZUKI YOSHIHIKO
    • H04N7/01
    • PROBLEM TO BE SOLVED: To improve resolution and fidelity with simple processing while solving a problem that information is lost and noise is mixed in.SOLUTION: An image processing device which performs super-resolution processing to an interlaced input image comprises, as components thereof: field super-resolution processing means which executes super-resolution processing for each field to the input image; frame super-resolution processing/field extract means which performs, for the input image, super-resolution processing for each frame formed by combining Top field and Bottom field of the input image in parallel with the processing by the field super-resolution processing means to output it in a field unit; characteristic analysis means which analyzes a characteristic of the input image and outputs a selecting signal corresponding to the characteristic; and first selection means which outputs in an interlaced form output of the field super-resolution processing means and the frame super-resolution processing/field extract means corresponding to the selecting signal.
    • 要解决的问题:通过简单的处理来提高分辨率和保真度,同时解决信息丢失和噪声混合的问题。解决方案:一种对隔行输入图像执行超分辨率处理的图像处理装置 包括作为其组成的场地超分辨率处理装置,其对于输入图像执行每场的超分辨率处理; 框架超分辨率处理/场提取装置,其对于输入图像执行与通过场超分辨率处理装置的处理并行地组合输入图像的顶场和底场形成的每个帧的超分辨率处理, 以现场单位输出; 特征分析装置,其分析输入图像的特性并输出与该特征对应的选择信号; 以及第一选择装置,其以隔行格式输出场超分辨率处理装置的输出和对应于选择信号的帧超分辨率处理/场提取装置。 版权所有(C)2013,JPO&INPIT
    • 3. 发明专利
    • Moving object tracking system and moving object tracking method
    • 移动对象跟踪系统和移动对象跟踪方法
    • JP2012059224A
    • 2012-03-22
    • JP2010204830
    • 2010-09-13
    • Toshiba Corp株式会社東芝
    • SAITO KODAISUKEGAWA HIROSHIYAMAGUCHI OSAMUSATO TOSHIO
    • G06T7/20H04N5/232H04N7/18
    • PROBLEM TO BE SOLVED: To provide a moving object tracking system capable of omitting a time-consuming operation such as answer teaching by automatically adjusting tracking parameters even for variation derived from a change in a photographing device or even for variation derived from a change in a photographing environment when tracking plural moving objects.SOLUTION: In a moving object tracking system that tracks a moving object by detecting plural moving objects included in plural frames in time series of images and associating the same moving objects among the frames with each other, reliability for tracking processing of the moving object is calculated, and if the calculated reliability is high, tracking parameters are automatically learned and adjusted.
    • 要解决的问题:提供一种运动对象跟踪系统,其能够通过自动调整跟踪参数来省略诸如回答教学的耗时的操作,即使对于从拍摄装置的变化导出的变化,或者甚至对于从 在跟踪多个移动物体时改变摄影环境。 解决方案:在移动物体跟踪系统中,通过检测多个帧中包含的多个移动物体的图像并在帧间相互关联相同的移动物体,跟踪移动物体,跟踪移动物体的移动物体的跟踪处理的可靠性 计算对象,如果计算的可靠性较高,则自动学习和调整跟踪参数。 版权所有(C)2012,JPO&INPIT
    • 4. 发明专利
    • Pattern recognition method, character recognition method, pattern recognition program, character recognition program, pattern recognition apparatus and character recognition apparatus
    • 图案识别方法,字符识别方法,图案识别程序,字符识别程序,图案识别装置和字符识别装置
    • JP2011065370A
    • 2011-03-31
    • JP2009214811
    • 2009-09-16
    • Toshiba Corp株式会社東芝
    • HAMAMURA TOMOYUKISATO TOSHIO
    • G06K9/62
    • G06K9/6215G06K9/6278
    • PROBLEM TO BE SOLVED: To provide a pattern recognition method which determines a recognition result by values obtained by evaluating similarities of an input pattern with respect to categories, on the basis of past recognition results, and efficiently achieves pattern recognition with high accuracy, to provide a character recognition method, a pattern recognition program, a character recognition program, a pattern recognition apparatus and a character recognition apparatus.
      SOLUTION: In pattern recognition, similarities of an input pattern with respect to a plurality of categories registered in a recognition dictionary are calculated, and the calculated similarities are converted into first evaluation values obtained from probabilities of the similarities under the condition that the calculated similarities are correct solutions and those of the similarities under the condition that the calculated similarities are not correct solutions, and second evaluation values are calculated from the first evaluation values and prior probabilities, and a category corresponding to a maximum second evaluation value is selected.
      COPYRIGHT: (C)2011,JPO&INPIT
    • 要解决的问题:提供一种模式识别方法,其通过基于过去的识别结果评估输入图案相似于类别的相似度而获得的值来确定识别结果,并且以高精度有效地实现模式识别 提供字符识别方法,图案识别程序,字符识别程序,图案识别装置和字符识别装置。 解决方案:在模式识别中,计算相对于在识别字典中登记的多个类别的输入模式的相似度,并且将所计算的相似度转换为从相似性的概率获得的第一评估值, 计算出的相似度是在计算出的相似度不是正确的解决方案的条件下的相似性的正确解决方案,并且从第一评估值和先验概率计算第二评估值,并且选择对应于最大第二评估值的类别。 版权所有(C)2011,JPO&INPIT
    • 5. 发明专利
    • Image pickup device
    • 图像拾取器件
    • JP2011002718A
    • 2011-01-06
    • JP2009146797
    • 2009-06-19
    • Toshiba Corp株式会社東芝
    • SATO TOSHIOHASHITANI SEIICHINAKANO NAOHISASAITO TSUTOMU
    • G03B7/18G03B15/00G03B15/02H04N5/225
    • PROBLEM TO BE SOLVED: To provide an image pickup device capable of photographing a person such as a driver in a vehicle by lessening influence by reflection of illuminating light or sunlight on a windshield of the vehicle.SOLUTION: The image pickup device simultaneously picks up an image of a face of the driver in the vehicle running on a toll acceptance lane of a tollgate and an image of a number plate of the vehicle. The image pickup device is provided with a polarizing filter 103 which allows only light in one direction (for example, longitudinal direction) to pass through on a light radiation side of an illuminator 101 radiating the illuminating light to field including the windshield 111 and the number plate 112 of the running vehicle 110, and also a polarizing filter 104 which allows only light in a direction (for example, lateral direction) different in the angle from the polarizing filter 103 on the illumination side to pass through on an imaging surface side of the camera 106 picking up the image in the field.
    • 要解决的问题:提供一种能够通过减轻照明光或阳光反射在车辆挡风玻璃上的影响来拍摄诸如车辆中的驾驶员的人的图像拾取装置。解决方案:图像拾取装置同时拾取 在车辆的收费车道上行驶的车辆中的驾驶员脸部的图像和车辆的牌照的图像。 图像拾取装置设置有偏振滤光器103,该偏振滤光器103仅允许在一个方向(例如,纵向方向)上的光在照射灯101的照射光101的光辐射侧通过,包括挡风玻璃111和数字 运行车辆110的板112以及仅在与照明侧的偏振光滤光器103的角度不同的方向(例如,横向)上的光仅在照射侧的偏振光滤光器103的摄像面侧通过的偏振光滤光器104 相机106拾取现场的图像。
    • 6. 发明专利
    • Person recognition apparatus and method and entry/exit management system
    • 人员识别装置和方法进入/退出管理系统
    • JP2007156541A
    • 2007-06-21
    • JP2005346876
    • 2005-11-30
    • Toshiba Corp株式会社東芝
    • SUKEGAWA HIROSHISATO TOSHIO
    • G06T7/00
    • PROBLEM TO BE SOLVED: To provide a person recognition apparatus and method and an entry/exit management system capable of recognizing subjects without depending on speeds at which the subjects pass and the number of subjects. SOLUTION: The person recognition apparatus, for recognizing previously registered persons by means of the persons' face feature information, controls the accuracy of a recognition process and the subjects depending on the degree of load applied to the recognition process according to information about speeds at which the subjects pass, the number of passengers, distances to the passengers, and the time elapsed since the start of the recognition process. COPYRIGHT: (C)2007,JPO&INPIT
    • 要解决的问题:提供能够识别被摄体而不依赖于被摄体通过的速度和被摄体数量的人物识别装置和方法以及入口/出口管理系统。 解决方案:用于通过人脸特征信息来识别先前注册的人的人识别装置根据关于识别过程的信息的程度根据关于识别过程的信息的程度来控制识别过程的精度和对象 受试者通过的速度,乘客人数,到乘客的距离以及自识别过程开始以来经过的时间。 版权所有(C)2007,JPO&INPIT
    • 7. 发明专利
    • Character recognition device and character recognition method
    • 字符识别装置和字符识别方法
    • JP2007058304A
    • 2007-03-08
    • JP2005239916
    • 2005-08-22
    • Toshiba Corp株式会社東芝
    • AKAGI TAKUMASATO TOSHIO
    • G06K9/34
    • G06K9/342G06K2209/01
    • PROBLEM TO BE SOLVED: To detect characters with high precision for character strings entered in a plurality of character entry frames described on a paper sheet. SOLUTION: A character recognition device conducts character recognition by generating individual character candidates from character strings entered in a plurality of character entry frames described on a paper sheet such as a business form. A line image separation unit 32 acquires images including the plurality of character entry frames and the character strings. A character detection processing unit including a variety of processing units 33-37 detects individual points at which a plurality of lines are contacted or intersected with one another in the character strings included in the images to determine the points at which the character strings should be segmented or re-combined on the basis of an arrangement relation between the corresponding points and character entry frames, and then, generates individual character candidates consisting of the character strings by conducting the segmentation or recombination. A character recognition unit 38 performs character recognition for the individual character candidates generated by the character detection processing unit. COPYRIGHT: (C)2007,JPO&INPIT
    • 要解决的问题:检测在纸张上描述的多个字符输入框中输入的字符串的高精度字符。 解决方案:字符识别装置通过在诸如业务表格的纸张上描述的多个字符输入帧中输入的字符串生成单个字符候选来进行字符识别。 线图像分离单元32获取包括多个字符输入帧和字符串的图像。 包括各种处理单元33-37的字符检测处理单元检测包括在图像中的字符串中的多个线彼此接触或相交的各个点,以确定字符串应被分割的点 或者基于对应点和字符输入帧之间的排列关系重新组合,然后通过进行分割或重组来生成由字符串组成的单个字符候选。 字符识别单元38对由字符检测处理单元产生的单个字符候选进行字符识别。 版权所有(C)2007,JPO&INPIT
    • 8. 发明专利
    • FACE IMAGE RECOGNIZING DEVICE AND METHOD
    • JP2001266152A
    • 2001-09-28
    • JP2000077160
    • 2000-03-17
    • TOSHIBA CORP
    • SATO TOSHIO
    • G06T7/00G06T1/00
    • PROBLEM TO BE SOLVED: To provide a face image recognizing device and a passage controller by which a face image can be recognized precisely by surely fetching even face images of optional persons different in body height. SOLUTION: A wide visual field lens 101 magnifies a visual field in the direction of the body heights of the persons 107 and 108, and an image input part 102 inputs an image in the visual field magnified by lens 101. An image correction processing part 103 corrects the distortion of the image inputted by the part 102, and a face pattern detecting part 104 detects the face patterns of the persons 107 and 108 from the image whose distortion is corrected by the part 103. A face pattern recognizing part 106 recognizes the face patterns of the persons 107 and 108 by collating the face pattern detected by the part 104 with the reference face pattern stored previously in a face pattern storing part 105.
    • 10. 发明专利
    • CONTAMINATION GRADE INSPECTING DEVICE FOR PRINTED MATTER
    • JP2000062299A
    • 2000-02-29
    • JP23685298
    • 1998-08-24
    • TOSHIBA CORP
    • DOBASHI HIROYOSHISATO TOSHIO
    • B41J29/46G01B11/28G07D7/182H04N1/00
    • PROBLEM TO BE SOLVED: To provide a contamination grade inspecting device for a printed matter capable of executing highly accurate inspection of a contamination grade of the printed material likely to a sense of a man. SOLUTION: A positioning section 102 executes positioning of image data of a printed matter P inputted from an image inputting section 101 and an average density calculating section 104 calculates an average density in each visual field region by using the positioned image data and a printing kind of the printed material and an area of the visual field obtained from a region information memory section 103. A density difference measuring section 105 measures a density difference by each visual field region by using the calculated average density of the printed material and a reference density of the visual field region of the printed material obtained from the region information memory section 103. A density difference normalization operating section 106 normalizes the density difference by each visual field region by using the measured density difference of the visual field region, the density average of the visual field region obtained from the region information memory section 103 and a variation level of the density distribution. An integrating section 107 judges a contamination grade of the printed matter P by integrating the normalized density differences of the respective visual field regions.