-
公开(公告)号：KR1020140028172A
公开(公告)日：2014-03-10
申请号：KR1020120075371
申请日：2012-07-11
Applicant: 한국전자통신연구원
CPC classification number: G10L19/018 , G06F17/30743 , G10L25/18 , G10L25/69
Abstract: Disclosed are a device and a method for measuring an audio quality using fingerprints and feeling feature information. The device for measuring an audio quality according to the present invention comprises a distortion signal generation unit for generating a plurality of distortion signals according to the predetermined distortion reference for an audio; an extraction unit for extracting fingerprints and feeling feature information corresponding to the audio and the distortion signals; a distortion degree measurement unit for measuring a fingerprint distance difference between the fingerprints corresponding to the audio and the fingerprints corresponding to the distortion signals and an AV distance difference between the feeling feature information corresponding to the audio and the feeling feature information corresponding to the distortion signals; a distortion function generation unit for generating a fingerprint distortion function and an AV distortion function based on the respective distributions corresponding to the fingerprint distance difference and the AV distance difference; and a search and measurement unit for extracting the fingerprints and the feeling feature information from an optional inquiry audio and comparing the extracted results with the fingerprint distortion function or the AV distortion function to measure an audio quality. [Reference numerals] (100) Distortion signal generation unit; (200) Extraction unit; (300) Distortion degree measurement unit; (400) Distortion function generation unit; (500) Search and measurement unit; (AA) Audio
Abstract translation: 公开了使用指纹和感觉特征信息来测量音频质量的装置和方法。 根据本发明的用于测量音频质量的装置包括：失真信号产生单元，用于根据音频的预定失真参考产生多个失真信号; 提取单元，用于提取与所述音频和所述失真信号相对应的指纹和感觉特征信息; 扭曲度测量单元，用于测量与音频对应的指纹与对应于失真信号的指纹之间的指纹距离差以及与音频对应的感觉特征信息与对应于失真信号的感觉特征信息之间的AV距离差 ; 失真函数生成单元，用于基于与指纹距离差和AV距离差对应的各个分布来生成指纹失真函数和AV失真函数; 以及搜索和测量单元，用于从可选择的查询音频中提取指纹和感觉特征信息，并将提取的结果与指纹失真功能或AV失真功能进行比较，以测量音频质量。 （附图标记）（100）失真信号生成单元; （200）提取单位; （300）失真度测量单位; （400）失真函数生成单元; （500）检索单位; （AA）音频
-
公开(公告)号：KR1020130119781A
公开(公告)日：2013-11-01
申请号：KR1020120042845
申请日：2012-04-24
Applicant: 한국전자통신연구원
CPC classification number: G06F17/30799 , H04N21/4586
Abstract: The present invention is a video duplication detecting apparatus comprising: a frame structuring part for structuring temporally continuous frames by decoding inputted videos; a feature extracting part for extracting the features of a corresponding video from the frames structured by the frame structuring part and coding the extracted features; a database for storing the coded features of one or more videos; a feature matching part for determining duplication of videos according as a video having the same coded features as the videos coded by the feature extracting part is detected from the database; and a duplicated video removing part for removing the duplicated videos except one in case that the inputted videos are determined to be duplicated by the feature matching part. [Reference numerals] (110) Frame structuring part;(120) Feature extracting part;(130) Feature matching part;(150) Duplicated video removing part
Abstract translation: 本发明是一种视频复制检测装置，包括：帧结构部分，用于通过解码输入的视频来构造时间上连续的帧; 特征提取部分，用于从由所述帧结构化部分构成的帧中提取对应视频的特征，并对所提取的特征进行编码; 用于存储一个或多个视频的编码特征的数据库; 从数据库检测特征匹配部分，用于根据具有与由特征提取部分编码的视频相同的编码特征的视频来确定视频的重复; 以及复制的视频消除部分，用于在确定输入的视频被特征匹配部分复制的情况下，去除除了一个的重复视频。 （110）帧结构部分;（120）特征提取部分;（130）特征匹配部分;（150）重复的视频去除部分
-
公开(公告)号：KR101287628B1
公开(公告)日：2013-08-23
申请号：KR1020090113718
申请日：2009-11-24
Applicant: 한국전자통신연구원
IPC: G06F21/10 , H04N21/4623 , H04N21/262 , H04N21/4405
Abstract: 도메인 라이센스 발급 장치, 방법 및 이에 의한 디지털 콘텐츠 재생 장치 및 방법을 제시한다. 제시된 도메인 라이센스 발급 장치는 도메인 내에서 디지털 콘텐츠의 공유를 위한 도메인 키를 생성하는 도메인 키 관리부, 도메인 키를 포함하는 도메인 라이센스를 생성하는 도메인 라이센스 생성부, 및 도메인에 속하는 기기로부터의 도메인 라이센스 발급 요청에 따라 도메인 라이센스를 전송하는 도메인 라이센스 발급부를 포함한다. 제시된 디지털 콘텐츠 재생 장치는 도메인 라이센스 발급 장치로부터의 도메인 라이센스에서 도메인 키를 추출하고 디지털 콘텐츠에 대응되는 도메인 공유 정보를 생성하여 디지털 콘텐츠 및 도메인 공유 정보를 하나의 세트로 하는 도메인 공유된 디지털 콘텐츠를 생성하는 도메인 클라이언트, 도메인 키를 이용하여 도메인 공유된 디지털 콘텐츠의 디지털 콘텐츠를 복호화하는 DRM 클라이언트, 및 복호화된 디지털 콘텐츠를 재생하는 재생기를 포함한다. 기기나 콘텐츠에 적용된 DRM 기술에 관계없이 정해진 도메인 내에서 콘텐츠의 공유가 가능하기 때문에 정당한 사용자에게 편의성을 제공함으로써 사용자들에게 DRM에 대한 거부감을 줄여줌으로써 콘텐츠 이용을 활성화시킬 수 있는 효과가 있다.
-
公开(公告)号：KR1020130058939A
公开(公告)日：2013-06-05
申请号：KR1020110124950
申请日：2011-11-28
Applicant: 한국전자통신연구원
CPC classification number: G06F17/3074 , G06F17/30775 , G10H1/0008 , G10H2210/061 , G10H2240/081 , G10L25/21 , G10L25/54 , G10L25/87 , G11B27/28 , H04N21/4384 , H04N21/4394
Abstract: PURPOSE: A device and a method for extracting a musical highlight section are provided to offer a technique for extracting the musical highlight section for a listener so that the listener confirms desired music rapidly or the existence of preference with respect to the music easily. CONSTITUTION: A device(100) for extracting a musical highlight section includes: a frame dividing unit which divides an audio file into a plurality of frames having constant length; an average signal calculating unit(140) which calculates signals per each frame indicating the intensity of average audio energy with respect to samples included in each frame; and a highlight section selecting unit(160) which extracts low-frequency signals from the signals indicating the intensity of average audio energy per each frame, and determines a frame section including a maximum point of the low-frequency signals as the highlight section. [Reference numerals] (120) Frame dividing unit; (140) Average energy signal calculating unit; (160) Highlight section selecting unit; (180) Start point search unit; (AA) Audio file; (BB) Reproduce a highlight section
Abstract translation: 目的：提供用于提取音乐高亮部分的装置和方法，以提供用于提取收听者的音乐高亮部分的技术，使得听众快速确认所需音乐或者容易地相对于音乐存在偏好。 构成：用于提取音乐高光部分的设备（100）包括：帧分割单元，其将音频文件分成具有恒定长度的多个帧; 平均信号计算单元（140），其针对每个帧中包括的采样来计算每帧指示平均音频能量的强度的信号; 以及高亮部分选择单元，其从表示每帧的平均音频能量的强度的信号中提取低频信号，并且确定包括低频信号的最大点的帧部分作为高亮部分。 （附图标记）（120）框架分割单元; （140）平均能量信号计算单元; （160）突出显示选区; （180）起点搜索单位; （AA）音频文件; （BB）重现高光部分
-
公开(公告)号：KR1020110117881A
公开(公告)日：2011-10-28
申请号：KR1020100037365
申请日：2010-04-22
Applicant: 한국전자통신연구원
CPC classification number: H04L9/0866 , G06F21/16 , H04L2209/606 , H04L2209/608 , H04N21/8358 , H04N21/8456
Abstract: A forensic mark insertion apparatus includes: an original content segment storage unit for storing multiple original contents; and a service control unit for, in response to a request for a content from a user, retrieving an original content corresponding to the request for the content from the original content segment storage unit, separating the retrieved original content into plurality of content segments, selectively inserting a forensic mark into some of the content segments, and assembling the content segments to produce an forensic-mark inserted content to be provided to the user.
-
Patent Agency Ranking
公开(公告)号：KR1020110071749A
公开(公告)日：2011-06-29
申请号：KR1020090128395
申请日：2009-12-21
Applicant: 한국전자통신연구원
IPC: G06F17/30 , H04N21/4728 , H04N21/8549 , G06Q50/10
CPC classification number: G06F17/30029 , G06F17/3005 , G06Q50/10 , H04N21/4728 , H04N21/8549
Abstract: PURPOSE: A contents information management device and method is provided to generate/manage the highlight section of the contents by using the contents-related information. CONSTITUTION: A section searcher(101) searches the similarity section between the section contents inputted from the user terminal and whole contents corresponding to the section contents. A highlight setting unit(102) calculates/updates the similarity section, the highlight section, and the feature information of whole contents. A search information storage unit(103) manages the highlight section which is updated in real time and the feature information according to the whole contents.
Abstract translation: 目的：提供内容信息管理装置和方法，通过使用内容相关信息来生成/管理内容的高亮部分。 构成：区段搜索器（101）搜索从用户终端输入的区段内容与对应于区段内容的全部内容之间的相似部分。 高亮度设定单元（102）计算/更新全部内容的相似度部分，高亮部分和特征信息。 搜索信息存储单元（103）根据整个内容管理实时更新的高光部分和特征信息。
-
公开(公告)号：KR1020110001697A
公开(公告)日：2011-01-06
申请号：KR1020090059357
申请日：2009-06-30
Applicant: 한국전자통신연구원
CPC classification number: H04N5/931 , H04N2005/91385
Abstract: PURPOSE: A copyright protection apparatus is provided to decide whether a moving picture is illegally distributed one or not by applying copyright information to the moving picture file. CONSTITUTION: A processing unit(100) decomposes a moving picture to a plurality of unit images. A processing unit is created a main image, a sub image, and an intensity image based on the unit images. The sub-image has a copyright information of a moving picture. An intensity image makes pair with the sub image. According to a merge intensity of the intensity image, a merging unit(200) merges the sub image with the main image. According to the time sequence, an output unit(300) consecutively outputs the merged image.
Abstract translation: 目的：提供一种版权保护装置，通过向运动图像文件应用版权信息来决定运动图像是否被非法分配。 构成：处理单元（100）将运动图像分解成多个单位图像。 基于单位图像创建处理单元的主图像，子图像和强度图像。 子图像具有运动图像的版权信息。 强度图像使得与子图像成对。 根据强度图像的合并强度，合并单元（200）将副图像与主图像合并。 根据时间顺序，输出单元（300）连续地输出合并的图像。
-
公开(公告)号：KR1020090065197A
公开(公告)日：2009-06-22
申请号：KR1020070132670
申请日：2007-12-17
Applicant: 한국전자통신연구원
IPC: G06F17/30
CPC classification number: G06F17/30247
Abstract: An apparatus of generating an index database, a generating method thereof, a search apparatus thereof, and a search method thereof are provided to reduce the amount of indexes to be stored in a database considering distribution features of feature indexes extracted from contents, thereby reducing a storage space. A feature extractor(110) extracts a feature from contents to be recorded in a database. An index generator(120) generates indexes from the extracted features. An index selector(130) selects an index to be stored based on frequency features of the generated indexes. A feature index database(140) stores the selected index and a location of the index.
Abstract translation: 提供一种生成索引数据库的装置，其生成方法，其搜索装置及其搜索方法，以考虑从内容中提取的特征索引的分布特征来减少存储在数据库中的索引的数量，从而减少 储存空间。 特征提取器（110）从要记录在数据库中的内容中提取特征。 索引生成器（120）从所提取的特征生成索引。 索引选择器（130）根据所生成的索引的频率特征来选择要存储的索引。 特征索引数据库（140）存储所选择的索引和索引的位置。
-
公开(公告)号：KR1020080111376A
公开(公告)日：2008-12-23
申请号：KR1020070129389
申请日：2007-12-12
Applicant: 한국전자통신연구원
IPC: H04N5/92
CPC classification number: H04N5/91
Abstract: A method for comparing digital video feature points and a system for managing digital video by using video features are provided to minimize the waste of resources such as time and storage space consumed for managing high volume digital video. A video frame of input digital video is normalized to extract a feature point(300). A digital video candidate group having similarity higher than a predetermined level is selected by searching a feature point of various digital video stored in a database by using the extracted feature point(400). Digital video information of a candidate group in which a difference between the feature point of the input digital video and the feature point of the candidate digital video is within a predetermined level is outputted as a result value(500).
Abstract translation: 提供了一种用于比较数字视频特征点的方法和用于通过使用视频特征来管理数字视频的系统，以最小化诸如用于管理大容量数字视频所消耗的时间和存储空间等资源的浪费。 将输入数字视频的视频帧归一化以提取特征点（300）。 通过使用所提取的特征点（400）搜索存储在数据库中的各种数字视频的特征点，选择具有高于预定级别的相似度的数字视频候选组。 作为结果值（500）输出其中输入数字视频的特征点与候选数字视频的特征点之间的差异在预定电平内的候选组的数字视频信息。
-
公开(公告)号：KR100797024B1
公开(公告)日：2008-01-22
申请号：KR1020060039649
申请日：2006-05-02
Applicant: 한국전자통신연구원
CPC classification number: G06T1/0071 , G06T2201/0052
Abstract: 본 발명은 하이브리드 이미지 워터마크 방법 및 장치에 관한 것으로, 이미지의 에지맵을 이용하여 에지맵의 특정 값을 기준으로 저주파 영역과 고주파 영역으로 구분한 후에 저주파 영역은 HVS를 이용한 워터마킹 방법을 적용하고, 고주파 영역은 비교적 큰 양자화 계수를 사용하는 QIM 방법을 적용하며, 추출 시에는 워터마크가 삽입된 이미지에서 에지 성분을 구한 후 삽입 시 사용했던 에지의 특정값을 대략적으로 추정한 다음 그 값을 기준으로 일정 범위 안에서 워터마크 이미지의 자기상관계수의 최대값이 다시 최대가 되는 특정값을 찾아내어 적용된 삽입방법을 예측하여 워터마크를 추출함으로써, 워터마크의 비가시적인 성질이 배가되며, 워터마크의 강인성이 증대되고, 추출 시에 두 가지의 워터마킹 방법을 쉽게 분별할 수 있어서 워터마크 추출의 용이성이 향상되는 이점이 있다.
디지털 워터마킹, 하이브리드, QIM, 양자화, 저작권보호
-
-
-
-
-
-
-
-
-
Search Results
91
records
(took 0.026 seconds)
