-
公开(公告)号：KR100687732B1
公开(公告)日：2007-02-27
申请号：KR1020050112966
申请日：2005-11-24
Applicant: 한국전자통신연구원 , 한국정보통신대학교 산학협력단
IPC: G06T7/00
Abstract: A method for blocking a malice moving picture by using a contents-based multi-modal feature value and an apparatus therefor are provided to gradually block the moving picture including malice information by using a key frame-based judging engine and a frame-based judging engine and block the malice moving picture by using multi-modal feature values such as contents-based visual feature information, audio feature information, and temporal axis feature information in which malice feature information is extracted from the moving picture. An SVM(Support Vector Machine) model unit constructs a discrimination model including a malice moving picture classification model and a non-malice moving picture classification model(S110). A separating unit separates a video stream and an audio stream from moving picture contents(S120). The first extracting unit extracts a certain visual feature value from the separated video stream, detects a shot boundary, extracts at least one or more key frames from each shot, and generates the shot and key frame information(S130). The second extracting unit performs a certain pre-process for the video stream, and extracts a multi-modal feature value of the key frame on the basis of the shot and the key frame information(S140). A discriminating unit inputs the multi-modal feature value to the discrimination mode, and judges the malice of the shot(S150). The discriminating unit synthesizes the judged result of the shot, and judges the malice of the moving picture contents(S160).
Abstract translation: 通过使用基于内容的多模态特征值来阻止恶意运动图像的方法及其装置被提供以通过使用基于关键帧的判断引擎和基于帧的判断引擎来逐渐地阻挡包括恶意信息的运动图像 并且通过使用诸如基于内容的视觉特征信息，音频特征信息和从运动图片中提取恶意特征信息的时间轴特征信息的多模式特征值来阻止恶意活动图片。 SVM（支持向量机）模型单元构建包括恶意运动图片分类模型和非恶意运动图片分类模型的判别模型（S110）。 分离单元从运动图像内容分离视频流和音频流（S120）。 第一提取单元从分离的视频流中提取特定视觉特征值，检测镜头边界，从每个镜头中提取至少一个或多个关键帧，并产生镜头和关键帧信息（S130）。 第二提取单元对视频流执行特定的预处理，并且基于镜头和关键帧信息提取关键帧的多模式特征值（S140）。 识别单元将多模特征值输入到识别模式，并判断拍摄的恶意（S150）。 鉴别单元合成镜头的判定结果，并判定运动画面内容的恶意（S160）。
-
公开(公告)号：KR1020160110625A
公开(公告)日：2016-09-22
申请号：KR1020150032784
申请日：2015-03-09
Applicant: 한국전자통신연구원
IPC: G06T7/20
CPC classification number: G06T7/20
Abstract: 격자형마스킹탐지장치는프로그램에따라영상에적용된격자형마스킹패턴을탐지하는프로세서및 프로그램을저장하는메모리를포함하되, 프로그램은영상에대한십자형특징점을검출하고, 십자형특징점중 미리지정된조건을만족하는주요특징점을선별하고, 주요특징점을포함하는십자형특징점리스트를생성하고, 십자형특징점리스트를참조하여격자형마스킹처리가된 영역으로추정되는영역의위치와크기를산출하고, 십자형특징점리스트에포함된특징점에대해격자형마스킹영역의중심위치와격자의가로및 세로간격을포함하는격자형마스킹배치변수를산출하고, 격자형마스킹배치변수에따라격자형마스킹탐지장치는십자형특징점리스트에포함된각각의객체에대해서격자형마스킹영역이존재를탐지하도록구성된다.
Abstract translation: 本发明涉及一种用于检测网格掩蔽区域的装置和方法。 提供了一种用于检测网格屏蔽区域的装置，该装置包括：处理器，其通过使用程序来检测应用于图像的网格屏蔽图案; 和存储程序的内存; 其中，所述程序从所述图像检测交叉特征点，在所述交叉特征点中选择满足预定条件的主要特征点，生成包括所述主要特征点的交叉特征点列表，计算估计为的区域的位置和大小 作为通过参照交叉特征点列表进行网格掩蔽处理的区域，并且计算包括网格掩蔽区域的中心的位置和网格的宽度和高度间隔相对于所包含的特征点的网格屏蔽处置变量 在交叉特征点列表中，并且该装置基于网格屏蔽处置变量来检测包括在交叉特征点列表中的每个对象是否存在网格掩蔽区域。
-
公开(公告)号：KR1020150051711A
公开(公告)日：2015-05-13
申请号：KR1020130133547
申请日：2013-11-05
Applicant: 한국전자통신연구원
IPC: G06T7/00
CPC classification number: G06T7/408 , G06K9/00362 , G06T7/11 , G06T7/194 , G06T7/90 , G06T2207/10024
Abstract: 영상으로부터피부영역과배경영역에대한확률밀도함수를산출하고, 산출한확률밀도함수를근거로 MLE 방식으로피부영역을추출하여영상으로부터피부영역을추출하는과정에서발생하는오탐지율을최소화하도록한 유해콘텐츠영상차단을위한피부영역추출장치및 방법이제시된다. 제시된유해콘텐츠영상차단을위한피부영역추출장치는영상매체로부터영상을추출하고, 기저장된사전정보를근거로영상의피부샘플영역및 배경샘플영역을추출하고, 피부샘플영역의확률밀도함수및 배경샘플영역의확률밀도함수를근거로영상으로부터피부영역을추출한다.
Abstract translation: 提供了一种用于提取皮肤区域以阻止有害图像内容的装置和方法，其中可以通过导出皮肤区域的概率密度函数来最小化在从图像提取皮肤区域的过程中发生的错误检测的速率， 背景区域，并且基于导出的概率密度函数，通过最大似然估计（MLE）提取皮肤区域。 在本发明中，提取皮肤区域以阻止有害图像内容的方法包括以下步骤：从图像媒体中提取图像; 基于先前存储的先前信息提取皮肤区域样本和背景区域样本; 并且基于皮肤区域样本的概率密度函数和背景区域样本的概率密度函数从图像中提取皮肤区域。
-
公开(公告)号：KR1020130101624A
公开(公告)日：2013-09-16
申请号：KR1020120015697
申请日：2012-02-16
Applicant: 한국전자통신연구원
CPC classification number: G06F17/30286 , G06F21/6218 , G06F2221/2143 , G06F2221/2149 , H04L63/0245 , H04L63/1425
Abstract: PURPOSE: A harmfulness contents remote management method is provided to soundly use a smart terminal by limiting a teenager child to be exposed to harmfulness contents. CONSTITUTION: A management target terminal transmits connection information of the management target terminal and connection information of a management terminal to a management server and makes the management server to be stored in a database unit (S100). The management target terminal analyzes contents which are stored in the management server and produces an analyzing result for the contents (S200). The management target terminal requests the connection information of the management terminal to the management server. The management target terminal transmits the analyzing result for the contents to the management server using the connection information of the management terminal which is transmitted from the management server (S300). The management target terminal processes the contents according to a processing request for the content which is transmitted from the management terminal (S400). [Reference numerals] (AA) Start; (BB) End; (S100) Store the connection information of a managed terminal and a management terminal; (S200) Managed terminal analyzes a content and generates an analyzing result; (S300) Managed terminal transmits the analyzing result to the management terminal by using the connection information of the management terminal; (S400) Control and manage the managed terminal depending on the analyzing result
Abstract translation: 目的：通过限制青少年儿童暴露于有害物质内容，提供有害物质远程管理方法，以健全使用智能终端。 构成：管理目标终端将管理对象终端的连接信息和管理终端的连接信息发送到管理服务器，并使管理服务器存储在数据库单元中（S100）。 管理目标终端分析存储在管理服务器中的内容，并生成内容的分析结果（S200）。 管理对象终端向管理服务器请求管理终端的连接信息。 管理对象终端使用从管理服务器发送的管理终端的连接信息，向管理服务器发送内容的分析结果（S300）。 管理目标终端根据从管理终端发送的内容的处理请求来处理内容（S400）。 （附图标记）（AA）开始; （BB）结束; （S100）存储被管理终端和管理终端的连接信息; （S200）管理终端分析内容并生成分析结果; （S300）管理终端通过管理终端的连接信息将分析结果发送给管理终端; （S400）根据分析结果控制和管理被管终端
-
公开(公告)号：KR1020110073238A
公开(公告)日：2011-06-29
申请号：KR1020100107618
申请日：2010-11-01
Applicant: 한국전자통신연구원
IPC: G06T7/40
CPC classification number: H04N21/454 , G06T7/90
Abstract: PURPOSE: According to the kind and category of a video, a harmful video blocking apparatus and method based on multimodal and multiscale features are provided to perform multi-stage harmfulness filtering suitable for each scale. CONSTITUTION: A multiscale feature analysis unit(110) creates multiscale features. A harmfulness classification model generator(120) creates multilevel harmfulness classification models. A harmfulness decision unit(130) decides the harmfulness of a video with comparison between at least one of the multiscale feature of the input video and at least one of the multilevel harmfulness classification model. In case the video is harmful, a harmful video blocking unit(140) secludes the input video.
Abstract translation: 目的：根据视频的种类和类别，提供了一种基于多模态和多尺度特征的有害视频拦截装置和方法，以执行适合每个尺度的多级有害滤波。 构成：多尺度特征分析单元（110）创建多尺度特征。 有害分类模型生成器（120）创建多级有害分类模型。 危害判定单元（130）通过比较输入视频的多尺度特征中的至少一个和多级有害性分类模型中的至少一个来判断视频的有害性。 如果视频有害，则有害视频阻挡单元（140）隐藏输入视频。
-
Patent Agency Ranking
公开(公告)号：KR1020110073225A
公开(公告)日：2011-06-29
申请号：KR1020100084657
申请日：2010-08-31
Applicant: 한국전자통신연구원
CPC classification number: G06F17/3089 , G06F9/4494 , G06F9/453 , G06F15/00 , G06F17/2229 , G06F17/30867 , G06F17/30902 , G11B2020/1099
Abstract: PURPOSE: A device and method for analyzing obscene content based on audio data are provided to block the section of content in which obscene grade is marked while reproducing the content by marking the obscene grade on the obscene section. CONSTITUTION: A content input unit(110) inputs content. An input data buffering unit(121) stores the content corresponding to the analysis section longer than the preset analysis section. An obscenity analyzer and obscenity decider(133, 145) determine the obscenity of the analysis section for the audio data extracted from the content. If the content is obscene, the obscene grade is marked in the analysis section. A content player(140) blocks the analysis section in which obscene grade is marked and reproduces the content.
Abstract translation: 目的：提供一种用于基于音频数据分析淫秽内容的装置和方法，以通过标记色情部分上的淫秽等级来再现内容来阻止在其中标记色情等级的内容部分。 构成：内容输入单元（110）输入内容。 输入数据缓冲单元（121）存储与分析单元相对应的内容比预设分析单元长。 淫秽分析器和淫秽判定器（133,145）确定从内容中提取的音频数据的分析部分的淫秽。 如果内容是淫秽的，淫秽等级在分析部分被标记。 内容播放器（140）阻止其中标记淫秽等级的分析部分并再现内容。
-
公开(公告)号：KR1020110056999A
公开(公告)日：2011-05-31
申请号：KR1020090113523
申请日：2009-11-23
CPC classification number: G06K9/00718 , G06T7/11 , G06T7/41 , G06T7/90
Abstract: PURPOSE: A harmful image discriminating and blocking apparatus is provided to efficiently model big harmful meaning through a meaning based feature and to efficiently block a harmful image. CONSTITUTION: A pre-processing unit(2000) divides an input image into one or more areas. A visual feature generator(3000) generates visual feature about the divided areas. A meaning feature generator(4000) generates the meaning feature from visual features by area. A harmful image blocking unit(5000) decides the final harmfulness of the image. A meaning feature modeling unit(6000) models meaning feature models which are stored in a meaning feature model storage(4300).
Abstract translation: 目的：提供有害的图像识别和阻挡装置，通过基于含义的特征有效地建模大的有害意义，并有效地阻止有害的图像。 规定：预处理单元（2000）将输入图像分成一个或多个区域。 视觉特征生成器（3000）生成关于分割区域的视觉特征。 意义特征生成器（4000）根据区域的视觉特征生成意义特征。 有害图像阻挡单元（5000）决定图像的最终有害性。 意义特征建模单元（6000）对存储在意义特征模型存储（4300）中的特征模型进行建模。
-
公开(公告)号：KR1020100066169A
公开(公告)日：2010-06-17
申请号：KR1020080124847
申请日：2008-12-09
Applicant: 한국전자통신연구원
IPC: G06F21/30
CPC classification number: H04L63/0407 , H04L63/0823
Abstract: PURPOSE: A system and a method for managing private information using anonymous authentication are provided to reinforce the decision right of a user about the exposure or usage of user information, thereby managing the user information. CONSTITUTION: A PIM(Private Information Management) server certifies a user through an anonymous certificate. The PIM server stores user information in a user terminal(S410). If a user terminal requests a web service from a web service provider, the web service provider requests user information which is necessary to certify a web provider(S460). The web service provider provides the web service to the user terminal(S470).
Abstract translation: 目的：提供一种使用匿名认证管理私人信息的系统和方法，以加强用户关于用户信息的曝光或使用的决定权，从而管理用户信息。 规定：PIM（私人信息管理）服务器通过匿名证书证明用户。 PIM服务器将用户信息存储在用户终端中（S410）。 如果用户终端从Web服务提供商请求Web服务，则Web服务提供商请求认证web提供者所需的用户信息（S460）。 Web服务提供商向用户终端提供Web服务（S470）。
-
公开(公告)号：KR100759798B1
公开(公告)日：2007-09-20
申请号：KR1020050101741
申请日：2005-10-27
Applicant: 한국전자통신연구원
CPC classification number: G06K9/00711
Abstract: 본 발명은 개인용 컴퓨터(이하 "PC"라고 함)에서 지능적 화면 감시를 통하여 유해 멀티미디어를 차단하는 장치 및 그 방법에 관한 것으로서, PC 상태에 기초하여 화면을 캡쳐할 시점을 결정하는 화면 캡쳐 판단부; 상기 결정 시점에 활성화되어 있는 프로그램의 화면을 캡쳐하는 활성 화면 캡쳐부; 상기 캡쳐된 화면의 유해성을 판단하는 이미지 유해성 판단부; 및 유해한 것으로 판단되면 상기 캡쳐된 화면을 사용하는 응용 프로그램을 차단하는 유해 프로그램 차단부;를 포함하는 것을 특징으로 하며, 지능적으로 PC화면을 캡쳐하고, 캡쳐된 화면의 유해성을 검사하여 유해하다고 판정 되는 경우에 캡쳐된 화면을 사용한 해당 응용 프로그램을 차단함으로써 화면 캡쳐 방법을 사용하여 실시간으로 유해 멀티미디어 정보에 접근하는 것을 차단할 수 있다.
지능적 화면 감시, 화면 감시, 멀티미디어, 유해 정보 차단 장치-
公开(公告)号：KR100670815B1
公开(公告)日：2007-01-19
申请号：KR1020050119996
申请日：2005-12-08
Applicant: 한국전자통신연구원
CPC classification number: H04L63/1416 , G06F21/563 , G06F2221/033 , H04L63/0236
Abstract: A device and a method for blocking a malicious multimedia service based on sequential data processing are provided to classify/block the malicious multimedia service by sequentially processing data of diverse types of multimedia services such as real-time multimedia streaming, and multimedia file which is stored and played from a storage space through a sequential data processing technique. A maliciousness classification model learner(110) generates a maliciousness determining module of each level through machine learning after extracting feature from more than one kind of moving images. A malicious data classifier(120) determines the malicious level of the inputted moving image based on probability for a specific malicious level of the inputted moving image at a determining time point and accumulated malicious probability until a current time point by sequentially inputting the inputted moving image to the maliciousness determining module. A malicious information blocker(130) stops the service if the malicious level is included to a reference malicious level.
Abstract translation: 提供了一种基于顺序数据处理来阻止恶意多媒体服务的设备和方法，用于通过顺序处理多种类型的多媒体服务的数据来分类/阻止恶意多媒体服务，诸如实时多媒体流和存储的多媒体文件 并通过顺序数据处理技术从存储空间播放。 恶意分类模型学习者（110）从多种运动图像提取特征后，通过机器学习生成每个级别的恶意确定模块。 恶意数据分类器（120）基于在确定时间点输入的运动图像的特定恶意级别的概率来确定输入的运动图像的恶意级别，并且通过顺序输入输入的运动图像累积恶意概率直到当前时间点 到恶意确定模块。 如果恶意级别被包括在参考恶意级别中，则恶意信息阻止器（130）停止服务。
-
-
-
-
-
-
-
-
-
Search Results
66
records
(took 0.032 seconds)
