CN101394522A - Detection method and system for video copy - Google Patents
Detection method and system for video copy Download PDFInfo
- Publication number
- CN101394522A CN101394522A CNA2007101220189A CN200710122018A CN101394522A CN 101394522 A CN101394522 A CN 101394522A CN A2007101220189 A CNA2007101220189 A CN A2007101220189A CN 200710122018 A CN200710122018 A CN 200710122018A CN 101394522 A CN101394522 A CN 101394522A
- Authority
- CN
- China
- Prior art keywords
- video
- key frame
- detected
- inquiry
- similarity
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Landscapes
- Image Analysis (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
- Television Signal Processing For Recording (AREA)
Abstract
The invention discloses a video copy detection method and a system. The method comprises the following steps: an inquired video and a detected video are converted to digital videos with the unified format; the inquired video and the detected video are carried out the video structure analysis, shots in the videos are segmented and key frames are extracted; an index is established for the key frames of the detected video, the key frames of the inquired video are utilized for carrying out the preliminary search of a video copy in the index, the detected video containing the similar content with the key frames of the inquired video is obtained, and the detected video is taken as the option video for the next matching; the similarity between the key frames of the inquired video and the sequence of the key frames of the option video is calculated, the matching of the sequence of the key frames is carried out, the similarity between the videos are confirmed, and the video with the high similarity is taken as a copy video source of the inquired video. The video copy detection method and the system can reduce the search space for the video copy detection and effectively improve the detection efficiency.
Description
Technical field
The present invention relates to the Video Detection field, particularly relate to a kind of detection method and system of video copy.
Background technology
Digitizing technique be very easy to video information storage, transmit and duplicate, make the quantity of digital video content rapidly increase.And being subjected to the restriction of conditions such as Network Transmission bandwidth, video quality requirement in actual applications, identical video content often is encoded as different forms, to adapt to different demands.
Usually, claim this identically with former video in terms of content, and the video that changes on form is a video copy, perhaps copies video.
Detect the copy video important practical use arranged in Vision information processing: at first video copy detection the video information retrieval and to Search Results filter, significant aspect the ordering; Secondly, aspect the medium tracking, video copy detection can be used for automatically monitoring the broadcast situation of certain content video; In addition, aspect the digital video copyright protection, for traditional digital watermark technology, owing to have need not to add additional information in medium, feature extraction such as can carry out at good characteristic after media releasing, and copy detection also begins to receive publicity.
Detect through the copy video after the format change, detection method must satisfy two requirements, promptly will change and the video data variation robust of generation Yin Geshi on the one hand, changes sensitivity to change the video data that produces because of video content again on the other hand.
For this reason, the researcher has proposed the various video copy detection method, but because how the diversity of video format and content chooses effective feature, detecting video copy exactly still is an an open question.
Summary of the invention
The object of the present invention is to provide a kind of detection method and system of video copy, it reduces the search volume of video copy detection, has effectively improved detection efficiency.
For realizing the detection method of a kind of video copy that the object of the invention provides, comprise the following steps:
Steps A is converted into inquiry video and detected video the digital video of consolidation form; Described inquiry video and detected video are carried out the video structure analysis, and the camera lens in the divided video extracts key frame;
Step B, key frame to detected video is set up index, utilize the key frame of inquiry video in index, to carry out tentatively searching of video copy, obtain having the detected video of similar content to the key frame of inquiring about video, and with the alternative videos of these detected videos as next step coupling;
Step C calculates the similarity between the key frame sequence of the key frame of inquiry video and alternative videos, and the key frame sequence is mated, and confirms the similarity between the video, and the high video of similarity is as the copy video source of inquiry video.
In the described steps A, described extraction key frame comprises the following steps:
Steps A 1, the HSI value of each frame in the calculating video;
Steps A 2 generates and adds up the histogram of each frame HSI value of whole video, and estimation obtains the video information entropy;
Steps A 3 is determined preferable key frame frame number.
After the described steps A 3, also comprise the following steps:
Steps A 4 according to preferable crucial frame number, is used the initial center of the cluster centre of self adaptation unsupervised clustering as the K-mean cluster, and each frame in the video is carried out the K-mean cluster.
Among the described step B, the feature of the key frame of detected video is set up index, comprise the following steps:
Step B1 carries out the principal component analysis dimensionality reduction to the SIFT feature of the key frame of the detected video that extracts;
Step B2 is to the feature employing R of the detected key frame of video behind the dimensionality reduction
*Tree structure is set up index.
Among the described step C, the similarity between the key frame of calculating inquiry video and the key frame sequence of alternative videos is mated the key frame sequence, comprises the following steps:
Step C1, with a key frame as the value on the time point on the time series, then the key frame sequence can be considered time series, and a promptly time dependent class value is considered as two key frame time serieses respectively with the key frame sequence of inquiry video and the key frame sequence of alternative videos;
Step C2 carries out the dynamic time warping similarity by feature to time series and calculates, and the key frame sequence is mated.
For realizing that the object of the invention also provides a kind of detection system of video copy, comprise the analysis extraction module, index module, detection module, wherein:
Described analysis extraction module is used for inquiry video and detected video are converted into the digital video of consolidation form; Described inquiry video and detected video are carried out the video structure analysis, and the camera lens in the divided video extracts key frame;
Described index module, be used for the key frame of detected video is set up index, utilize the key frame of inquiry video in index, to carry out tentatively searching of video copy, obtain having the video number of the key frame of detected video of similar content and this detected video to the key frame of inquiring about video, and with the alternative videos of these detected videos as next step coupling;
Detection module is used to calculate the similarity between the key frame sequence of the key frame of inquiry video and alternative videos, and the key frame sequence is mated, and confirms the similarity between the video, and the high video of similarity is as the copy video of inquiry video.
Described extraction key frame comprises:
Calculate the HSI value of each frame in the video;
Generate and add up the histogram of each frame HSI value of whole video, estimation obtains the video information entropy;
Determine preferable key frame frame number.
According to preferable crucial frame number, use the initial center of the cluster centre of self adaptation unsupervised clustering as the K-mean cluster, each frame in the video is carried out the K-mean cluster.
The feature of described key frame to detected video is set up index, comprising:
SIFT feature to the key frame of the detected video that extracts is carried out the principal component analysis dimensionality reduction;
Feature to the detected key frame of video behind the dimensionality reduction adopts R
*Tree structure is set up index.
Similarity between the key frame of described calculating inquiry video and the key frame sequence of alternative videos is mated the key frame sequence, is meant:
With a key frame as the value on the time point on the time series, then the key frame sequence can be considered time series, be a time dependent class value, the key frame sequence of inquiry video and the key frame sequence of alternative videos are considered as two key frame time serieses respectively;
By feature time series is carried out the dynamic time warping similarity and calculate, the key frame sequence is mated.
Effective effect of the present invention is: the detection method of video copy of the present invention and system, video is carried out key-frame extraction, and as the basis of analyzing, improved efficient and processing speed with this; To key-frame extraction SIFT feature, this feature to the noise ratio in the video copy process vision signal brought than robust, thereby not affected by noise when carrying out the key frame sequences match.Adopt DTW method in the time series coupling carrying out key frame when coupling, can effectively overcome the not exclusively corresponding and sequences match problem brought in key frame position.
Description of drawings
Fig. 1 is the detection method flow chart of video copy of the present invention;
Fig. 2 is the detection system structural representation of video copy of the present invention.
Embodiment
In order to make purpose of the present invention, technical scheme and advantage clearer,, the detection method and the system of a kind of video copy of the present invention is further elaborated below in conjunction with drawings and Examples.Should be appreciated that specific embodiment described herein only in order to explanation the present invention, and be not used in qualification the present invention.
The detection method of video copy of the present invention and system, be whether video segment is detected for another video content copies, it carries out content analysis to the digital video that has obtained, obtain content element, each unit is extracted the key frame of video that can express the unit main contents, determine that by the similarity of analysis of key frame sequence a video segment is all or part of video copy of another video.
In embodiments of the present invention, with video (being called detected video) with copyright, to coming from the inquiry video on the Internet, the content (promptly inquiring about the part or all of content whether video has copied detected video) whether detection inquiry video wherein comprises in the detected video is an example, video copying detection method of the present invention is described, but it is not the qualification to range of application of the present invention, and the present invention also goes for the detection of other video copies.
Describe the detection method of a kind of video copy of the present invention below in detail, as shown in Figure 1, comprise the following steps:
Step S100 is converted into inquiry video and detected video the digital video of consolidation form; Described inquiry video and detected video are carried out the video structure analysis, and the camera lens in the divided video extracts key frame.
With detected video and a plurality of inquiry videos of downloading from the Internet, be converted into the video format of consolidation form, make that size, the coded format of video are unified, as being mpeg format etc. from AVI, DV-AVI, MOV format conversion.This conversion is the state of the art, and those skilled in the art can realize its conversion process according to content disclosed by the invention, therefore, describes in detail no longer one by one in the present invention.
Detected video and inquiry video are all carried out the video structure analysis, and the camera lens in the divided video (shot) extracts key frame.。
The video structure fractional analysis be meant to video flowing carry out that camera lens is cut apart, processing such as key-frame extraction and scene are cut apart.
The key that camera lens is cut apart is to determine the border of camera lens.The many discontinuities with video content of existing camera lens dividing method are the foundation of dividing camera lens, and the discontinuity that certain feature of common selecting video is measured video content is as color characteristic, motion vector feature, edge feature etc.
Cut apart by camera lens, can carry out succinct expression, thereby under the prerequisite that keeps the video substance, video content is carried out expressing based on the key frame of lens unit video content based on lens unit;
After camera lens is cut apart, can extract some key frames, and represent camera lens compactly with key frame to each camera lens.
Owing to have the redundancy of certain degree between each two field picture in the same camera lens, the two field picture that therefore can choose the main information content in the reflection camera lens is as key frame.Redundancy between the key frame has been lowered like this, can improve the efficient of content-based Video processing.
Wherein, in embodiments of the present invention, the concrete grammar of the key-frame extraction of described video is:
Step S110, the HSI value of each frame in the calculating video;
Wherein, and HSI (Hue, Saturation, Intensity, or claim HSV) expression form and aspect, saturation and brightness.
The computational methods of HSI are a kind of prior aries, as a kind of embodiment, can utilize the method as describing among " Technical Basis of Multimedia Computer and its Application " (Zhong Yu chisel etc., Higher Education Publishing House, 1999) to realize.
With this image of equal value representation of each pixel HSV in every two field picture, simultaneously HSI non-uniform quantizing to 0～165 totally 166 grades.Quantization method also is a kind of prior art, as a kind of enforceable mode, can utilize (MichelLantage as " VIP:Vision tool for comparing Images of People ", Marc Parizeau and Robert Bergevin, In:Vision Interface 2003) method of describing in realizes.
Step S120 generates and adds up the histogram of each frame HSI value of whole video, and estimation obtains video information entropy H (X);
Wherein the computational methods of comentropy are:
Wherein totalFrameNum represents the totalframes of this video flowing, and FrameNumi represents to be quantized in the video flowing frame number on i rank, N=1, and 2 ...
Step S130 determines preferable key frame frame number;
The relation of key frame number and comentropy can be expressed as follows:
For key frame, from the needs that key frame is simplified, key frame is to have nothing in common with each other, and for K key frame, can be considered as K the symbol that equiprobability occurs, and comentropy at this moment is
Bit; From the angle of reflecting video content, many more important more hypothesis appear in video based on picture frame, if
The randomness of video flowing information source is greater than the randomness of simplifying the back key frame, and the number of key frame is not enough to the variation of reflecting video, and if
The randomness of video flowing information source is less than the randomness of simplifying the back key frame, and the number of key frame is too much, may choose in video the frame that occurs seldom as key frame.Therefore, preferable situation is that the frame number K of key frame and the comentropy of video are satisfied:
Be K=2
H (X)
Step S140 according to preferable crucial frame number, uses the initial center of the cluster centre of self adaptation unsupervised clustering as the K-mean cluster, and each frame in the video is carried out the K-mean cluster.
Cluster is the tolerance based on the similitude between different objects of numerical expression.The K-average is a simple unsupervised learning method (algorithm), and it is used to solve some known clustering problem.K-mean cluster method is used to produce cluster and classification fixed dimension, straight according to the distance metric of similitude.Its given number (for example, K cluster) by fixing in advance cluster is classified given data set.It is a kind of prior art, and those skilled in the art can realize the K-mean cluster process to each frame in the video according to content disclosed by the invention, therefore, describes in detail no longer one by one in the present invention.
Step S200, key frame to detected video is set up index, utilize the key frame of inquiry video in index, tentatively to search, obtain having the video number of the key frame of detected video of similar content and this detected video to the key frame of inquiring about video, and with the alternative videos of these detected videos as next step coupling.
In the embodiment of the invention, key frame to detected video is set up index, be to key-frame extraction SIFT (Scale Invariant Feature Transform) feature, and set up index, be used for tentatively searching the content that whether exists in the detected video with inquiry video coupling.
It can express robust (robustness) feature of video content to key-frame extraction, by setting up high dimensional indexing with these features, accelerates the speed of searching similar key frame.
When tentatively searching, record is through searching, has the key frame of the detected video of similar content to the key frame of inquiry video (i.e. the video of submitting to), and the video of this detected video number, with the alternative videos of these detected videos as next step coupling, the result who searches according to key frame tentatively selectes the preliminary scope of further carrying out video copy detection.
Wherein, in embodiments of the present invention, the concrete grammar of the key frame of detected video being set up index is:
Step S210 carries out dimensionality reduction to the SIFT feature of the key frame of the detected video that extracts;
Because the SIFT characteristic point that generates is many, data volume is big, and therefore, the matrix that the description of SIFT characteristic point is constituted carries out dimensionality reduction.
Preferably, (Principal Component Analysis, PCA) effective aspect the concentration of energy meaning, the embodiment of the invention adopts PCA that the SIFT eigenmatrix is carried out dimensionality reduction in view of principal component analysis.
In the embodiment of the invention, regard each point SIFT feature the delegation of eigenmatrix as, then the feature of a two field picture becomes an eigenmatrix, and this eigenmatrix is carried out the PCA conversion. get low dimensional feature vector after the conversion as the key frame feature of low dimension.As a kind of embodiment, the PCA conversion can utilize (the David Salomon as " The Complete Reference (2nd Edition) "; Data Compression:The Complete Reference (2nd Edition), Springer, ISBN 0-387-95045-1,2000) the middle method realization of describing.
Step S220 is to the feature employing R of the detected key frame of video behind the dimensionality reduction
*Tree structure is set up index.
Wherein, R
*Tree is the distressed structure of R tree (region tree, zone tree), is the index structure commonly used in the present large scale database, has Stability Analysis of Structures, retrieves advantages such as rapid.The present invention is directed to the characteristics of image feature data, adopted R
*Tree structure is realized the index of image library.
R
*Tree is based on the R tree mutation that different pieces of information distributes and proposes.It is different from the R tree only at little leaf zone volume optimization, R
*Tree has also been optimized: the covering between (1) leaf node zone; (2) surface in leaf node zone; (3) memory by using rate.Different with R tree insertion algorithm, R
*Tree is when being the suitable insertion leaf node of data target selection, and when point was not included in any leaf node zone, the processing method of leaf node and directory node was different.For leaf node, to select to make to cover the zone that produces minimum expansion value, expansion is worth when identical, further relatively the volume expansion value and the own vol size of leaf node.For directory node, judge earlier the minimum volume enlarged area, if can not judge the time, then by self volume decision.R
*The characteristics of tree splitting algorithm have been to propose the notion of a kind of " inserting again by force ", if promptly a node overflows, just delete the target away from the central area of certain percentage, insert these targets again by the insertion method again.Though R
*" inserting again " algorithm of tree can increase computational complexity and visit disk number of times, but this recruitment has been avoided the division of a lot of nodes.
Step S300, similarity between the key frame of calculating inquiry video and the key frame sequence of alternative videos, the key frame sequence is mated, confirm the similarity between the video, the high video of similarity is as the copy video source of inquiry video, promptly thinks video on this coupling in the detected video that come from online inquiry video copy.
Among the present invention, calculation of similarity degree between the video is converted into the key frame sequence similarity to be calculated, by the time series matching process, can effectively calculate the similarity between the key frame sequence, thereby judge effectively according to similarity whether a video is all or part of video copy of another video.
Wherein, calculate the similarity between the key frame sequence of the key frame of inquiry video and alternative videos in embodiments of the present invention, the concrete grammar that the key frame sequence is mated is:
Step S310, as the value on the time point on the time series, then the key frame sequence can be considered time series with a key frame, a promptly time dependent class value.The key frame sequence of inquiry video and the key frame sequence of alternative videos are considered as two key frame time serieses respectively;
Time series is the important data object of a class, extensively exists in many applications such as economy, the hydrology, meteorologies.Characteristics such as time series has the dimension height, data volume is big and noise jamming is serious, the seasonal effect in time series pattern list is shown with the benefit of tripartite face: the one, time series is compressed, bring littler storage and calculation cost; The 2nd, only kept the main form of seasonal effect in time series, removed the details interference, more can reflect the seasonal effect in time series unique characteristics, help improving the efficient and the accuracy of data mining; The 3rd, a lot of applications are concerned about is changing pattern and rule in time series a period of time, rather than the value of single sequence of points in the time series, and modal representation more can meet its characteristics.
Step S320, (Dynamic Time Warping, DTW) similarity is calculated, and the key frame sequence is mated by feature time series to be carried out dynamic time warping.
When carrying out the calculating of DTW method, adopt the fast algorithm of introducing the Keogh lower bound to carry out the DTW distance calculation;
By feature time series being carried out the calculating of DTW similarity is a kind of prior art, as a kind of embodiment, can utilize (Michail Vlachos as " Indexing Multi-Dimensional Time-Series with SupportforMultiple Distance Measures ", Marios Hadjieleftheriou, Dimitrios Gunopulos, Eamonn Keogh, SIGKDD ＇ 03, August 24-27,2003, Washington, DC, USA) the middle method of describing realizes.
At last, can be by preestablishing a threshold values, when similarity is higher than the video that preestablishes threshold values, then as the copy video of inquiring about video.
The detection method of video copy of the present invention by definite alternative videos, has been dwindled the scope of the video that will carry out the video segment coupling; Video in this scope is carried out the key frame sequences match one by one, thereby obtain final result of determination, the video of being chosen is promptly as the copy video of inquiring about video.
Correspondingly, the present invention also provides a kind of detection system 20 of video copy, and as shown in Figure 2, this system comprises analysis extraction module 21, index module 22, detection module 23.Wherein:
Described analysis extraction module 21 is used for inquiry video and detected video are converted into the digital video of consolidation form; Described inquiry video and detected video are carried out the video structure analysis, and the camera lens in the divided video extracts key frame.
Described index module 22, be used for the key frame of detected video is set up index, utilize the key frame of inquiry video in index, tentatively to search, obtain having the video number of the key frame of detected video of similar content and this detected video to the key frame of inquiring about video, and with the alternative videos of these detected videos as next step coupling.
The detection system of video copy of the present invention is carried out work with the process identical with the detection method of video copy of the present invention, therefore, in embodiments of the present invention, repeats no longer one by one in detail.
The detection method of the video copy of the embodiment of the invention and system carry out key-frame extraction to video, as the basis of analyzing, have improved efficient and processing speed with this; To key-frame extraction SIFT feature, this feature to the noise ratio in the video copy process vision signal brought than robust, thereby not affected by noise when carrying out the key frame sequences match.Adopt DTW method in the time series coupling carrying out key frame when coupling, can effectively overcome the not exclusively corresponding and sequences match problem brought in key frame position.
In conjunction with the drawings to the description of the specific embodiment of the invention, others of the present invention and feature are conspicuous to those skilled in the art.
More than specific embodiments of the invention are described and illustrate it is exemplary that these embodiment should be considered to it, and be not used in and limit the invention, the present invention should make an explanation according to appended claim.
Claims (10)
1, a kind of detection method of video copy is characterized in that, comprises the following steps:
Steps A is converted into inquiry video and detected video the digital video of consolidation form; Described inquiry video and detected video are carried out the video structure analysis, and the camera lens in the divided video extracts key frame;
Step B, key frame to detected video is set up index, utilize the key frame of inquiry video in index, tentatively to search, obtain having the detected video of similar content to the key frame of inquiring about video, and with the alternative videos of these detected videos as next step coupling;
Step C calculates the similarity between the key frame sequence of the key frame of inquiry video and alternative videos, and the key frame sequence is mated, and confirms the similarity between the video, and the high video of similarity is as the copy video source of inquiry video.
2, the detection method of video copy according to claim 1 is characterized in that, in the described steps A, described extraction key frame comprises the following steps:
Steps A 1, the HSI value of each frame in the calculating video;
Steps A 2 generates and adds up the histogram of each frame HSI value of whole video, and estimation obtains the video information entropy;
Steps A 3 is determined preferable key frame frame number.
3, the detection method of video copy according to claim 2 is characterized in that, after the described steps A 3, also comprises the following steps:
Steps A 4 according to preferable crucial frame number, is used the initial center of the cluster centre of self adaptation unsupervised clustering as the K-mean cluster, and each frame in the video is carried out the K-mean cluster.
4, the detection method of video copy according to claim 1 is characterized in that, among the described step B, the feature of the key frame of detected video is set up index, comprises the following steps:
Step B1 carries out the principal component analysis dimensionality reduction to the SIFT feature of the key frame of the detected video that extracts;
Step B2 is to the feature employing R of the detected key frame of video behind the dimensionality reduction
*Tree structure is set up index.
5, the detection method of video copy according to claim 1 is characterized in that, among the described step C, the similarity between the key frame of calculating inquiry video and the key frame sequence of alternative videos is mated the key frame sequence, comprises the following steps:
Step C1, with a key frame as the value on the time point on the time series, then the key frame sequence can be considered time series, and a promptly time dependent class value is considered as two key frame time serieses respectively with the key frame sequence of inquiry video and the key frame sequence of alternative videos;
Step C2 carries out the dynamic time warping similarity by feature to time series and calculates, and the key frame sequence is mated.
6, a kind of detection system of video copy is characterized in that, comprises the analysis extraction module, index module, and detection module, wherein:
Described analysis extraction module is used for inquiry video and detected video are converted into the digital video of consolidation form; Described inquiry video and detected video are carried out the video structure analysis, and the camera lens in the divided video extracts key frame;
Described index module, be used for the key frame of detected video is set up index, utilize the key frame of inquiry video in index, tentatively to search, obtain having the detected video of similar content to the key frame of inquiring about video, and with the alternative videos of these detected videos as next step coupling;
Detection module is used to calculate the similarity between the key frame sequence of the key frame of inquiry video and alternative videos, and the key frame sequence is mated, and confirms the similarity between the video, and the high video of similarity is as the copy video source of inquiry video.
7, the detection system of video copy according to claim 6 is characterized in that, described extraction key frame comprises:
Calculate the HSI value of each frame in the video;
Generate and add up the histogram of each frame HSI value of whole video, estimation obtains the video information entropy;
Determine preferable key frame frame number.
8, the detection system of video copy according to claim 7 is characterized in that, described extraction key frame also comprises:
According to preferable crucial frame number, use the initial center of the cluster centre of self adaptation unsupervised clustering as the K-mean cluster, each frame in the video is carried out the K-mean cluster.
9, the detection system of video copy according to claim 6 is characterized in that, the feature of described key frame to detected video is set up index, comprising:
SIFT feature to the key frame of the detected video that extracts is carried out the principal component analysis dimensionality reduction;
Feature to the detected key frame of video behind the dimensionality reduction adopts R
*Tree structure is set up index.
10, the detection system of video copy according to claim 6 is characterized in that, the similarity between the key frame of described calculating inquiry video and the key frame sequence of alternative videos is mated the key frame sequence, is meant:
With a key frame as the value on the time point on the time series, then the key frame sequence can be considered time series, be a time dependent class value, the key frame sequence of inquiry video and the key frame sequence of alternative videos are considered as two key frame time serieses respectively;
By feature time series is carried out the dynamic time warping similarity and calculate, the key frame sequence is mated.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2007101220189A CN101394522B (en) | 2007-09-19 | 2007-09-19 | Detection method and system for video copy |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2007101220189A CN101394522B (en) | 2007-09-19 | 2007-09-19 | Detection method and system for video copy |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN101394522A true CN101394522A (en) | 2009-03-25 |
| CN101394522B CN101394522B (en) | 2010-07-21 |
Family
ID=40494564
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2007101220189A Expired - Fee Related CN101394522B (en) | 2007-09-19 | 2007-09-19 | Detection method and system for video copy |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101394522B (en) |
Cited By (32)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101872415A (en) * | 2010-05-06 | 2010-10-27 | 复旦大学 | Video copying detection method being suitable for IPTV |
| CN102129707A (en) * | 2011-02-28 | 2011-07-20 | 浙江大学 | Heterogeneous feature dimension reduction-based two-dimensional role cartoon generation method |
| CN102693299A (en) * | 2012-05-17 | 2012-09-26 | 西安交通大学 | System and method for parallel video copy detection |
| CN102737135A (en) * | 2012-07-10 | 2012-10-17 | 北京大学 | Video copy detection method and system based on soft cascade model sensitive to deformation |
| CN103065660A (en) * | 2012-12-11 | 2013-04-24 | 天津天地伟业数码科技有限公司 | Video file locating method of embedded video recorder |
| CN103327356A (en) * | 2013-06-28 | 2013-09-25 | Tcl集团股份有限公司 | Video matching method and device |
| CN103440640A (en) * | 2013-07-26 | 2013-12-11 | 北京理工大学 | Method for clustering and browsing video scenes |
| CN103514196A (en) * | 2012-06-25 | 2014-01-15 | 联想(北京)有限公司 | Information processing method and electronic device |
| CN103582882A (en) * | 2011-01-07 | 2014-02-12 | 阿尔卡特朗讯公司 | Method and apparatus for comparing videos |
| CN103780902A (en) * | 2014-02-14 | 2014-05-07 | 北京奇艺世纪科技有限公司 | Approximate video detection method and device |
| WO2014071558A1 (en) * | 2012-11-06 | 2014-05-15 | Nokia Corporation | Method and apparatus for summarization based on facial expressions |
| CN103905694A (en) * | 2014-04-10 | 2014-07-02 | 中央电视台 | Key frame processing method and system |
| CN104239420A (en) * | 2014-10-20 | 2014-12-24 | 北京畅景立达软件技术有限公司 | Video fingerprinting-based video similarity matching method |
| CN104331450A (en) * | 2014-10-29 | 2015-02-04 | 聂秀山 | Video copy detection method based on multimodal features and tensor decomposition |
| CN104504307A (en) * | 2015-01-08 | 2015-04-08 | 北京大学 | Method and device for detecting audio/video copy based on copy cells |
| CN105072455A (en) * | 2015-08-11 | 2015-11-18 | 精硕世纪科技（北京）有限公司 | Video matching method and device |
| CN105139014A (en) * | 2015-07-14 | 2015-12-09 | 北京联合大学 | Method for describing image local characteristic descriptor |
| CN105227907A (en) * | 2015-08-31 | 2016-01-06 | 电子科技大学 | Based on the nothing supervision anomalous event real-time detection method of video |
| CN105608423A (en) * | 2015-12-17 | 2016-05-25 | 天脉聚源（北京）科技有限公司 | Video matching method and device |
| CN106021321A (en) * | 2016-05-06 | 2016-10-12 | 成都索贝数码科技股份有限公司 | Picture based online real-time video search method |
| CN106570165A (en) * | 2016-11-07 | 2017-04-19 | 北京航空航天大学 | Content-based video retrieval method and apparatus |
| CN106649663A (en) * | 2016-12-14 | 2017-05-10 | 大连理工大学 | Video copy detection method based on compact video representation |
| CN107135401A (en) * | 2017-03-31 | 2017-09-05 | 北京奇艺世纪科技有限公司 | Key frame extraction method and system |
| CN108255970A (en) * | 2017-12-26 | 2018-07-06 | 努比亚技术有限公司 | A kind of video retrieval method, terminal and computer readable storage medium |
| CN110147469A (en) * | 2019-05-14 | 2019-08-20 | 腾讯音乐娱乐科技（深圳）有限公司 | A kind of data processing method, equipment and storage medium |
| CN110175267A (en) * | 2019-06-04 | 2019-08-27 | 黑龙江省七星农场 | A kind of agriculture Internet of Things control processing method based on unmanned aerial vehicle remote sensing technology |
| CN110321454A (en) * | 2019-08-06 | 2019-10-11 | 北京字节跳动网络技术有限公司 | Processing method, device, electronic equipment and the computer readable storage medium of video |
| CN111601115A (en) * | 2020-05-12 | 2020-08-28 | 腾讯科技（深圳）有限公司 | Video detection method, related device, equipment and storage medium |
| CN112507875A (en) * | 2020-12-10 | 2021-03-16 | 上海连尚网络科技有限公司 | Method and equipment for detecting video repetition |
| CN112651336A (en) * | 2020-12-25 | 2021-04-13 | 深圳万兴软件有限公司 | Method, device and computer readable storage medium for determining key frame |
| CN113239855A (en) * | 2021-05-27 | 2021-08-10 | 北京字节跳动网络技术有限公司 | Video detection method and device, electronic equipment and storage medium |
| CN113886632A (en) * | 2021-12-03 | 2022-01-04 | 杭州并坚科技有限公司 | Video retrieval matching method based on dynamic programming |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR100371513B1 (en) * | 1999-12-06 | 2003-02-07 | 주식회사 팬택앤큐리텔 | Method and apparatus of summerizing and browsing video sequences using fidelity values of key frame stored in its edge of key frame hierarchy |
| CN1851710A (en) * | 2006-05-25 | 2006-10-25 | 浙江大学 | Embedded multimedia key frame based video search realizing method |
| CN100530239C (en) * | 2007-01-25 | 2009-08-19 | 复旦大学 | Video stabilizing method based on matching and tracking of characteristic |
-
2007
- 2007-09-19 CN CN2007101220189A patent/CN101394522B/en not_active Expired - Fee Related
Cited By (53)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101872415A (en) * | 2010-05-06 | 2010-10-27 | 复旦大学 | Video copying detection method being suitable for IPTV |
| CN103582882A (en) * | 2011-01-07 | 2014-02-12 | 阿尔卡特朗讯公司 | Method and apparatus for comparing videos |
| CN102129707A (en) * | 2011-02-28 | 2011-07-20 | 浙江大学 | Heterogeneous feature dimension reduction-based two-dimensional role cartoon generation method |
| CN102693299A (en) * | 2012-05-17 | 2012-09-26 | 西安交通大学 | System and method for parallel video copy detection |
| CN103514196B (en) * | 2012-06-25 | 2020-07-24 | 联想(北京)有限公司 | Information processing method and electronic equipment |
| CN103514196A (en) * | 2012-06-25 | 2014-01-15 | 联想(北京)有限公司 | Information processing method and electronic device |
| CN102737135A (en) * | 2012-07-10 | 2012-10-17 | 北京大学 | Video copy detection method and system based on soft cascade model sensitive to deformation |
| CN102737135B (en) * | 2012-07-10 | 2015-07-01 | 北京大学 | Video copy detection method and system based on soft cascade model sensitive to deformation |
| CN104769611A (en) * | 2012-11-06 | 2015-07-08 | 诺基亚技术有限公司 | Method and apparatus for summarization based on facial expressions |
| WO2014071558A1 (en) * | 2012-11-06 | 2014-05-15 | Nokia Corporation | Method and apparatus for summarization based on facial expressions |
| US9754157B2 (en) | 2012-11-06 | 2017-09-05 | Nokia Technologies Oy | Method and apparatus for summarization based on facial expressions |
| CN103065660A (en) * | 2012-12-11 | 2013-04-24 | 天津天地伟业数码科技有限公司 | Video file locating method of embedded video recorder |
| CN103327356B (en) * | 2013-06-28 | 2016-02-24 | Tcl集团股份有限公司 | A kind of video matching method, device |
| CN103327356A (en) * | 2013-06-28 | 2013-09-25 | Tcl集团股份有限公司 | Video matching method and device |
| CN103440640A (en) * | 2013-07-26 | 2013-12-11 | 北京理工大学 | Method for clustering and browsing video scenes |
| CN103440640B (en) * | 2013-07-26 | 2016-02-10 | 北京理工大学 | A kind of video scene cluster and browsing method |
| CN103780902B (en) * | 2014-02-14 | 2016-05-11 | 北京奇艺世纪科技有限公司 | A kind of approximate video detecting method and device |
| CN103780902A (en) * | 2014-02-14 | 2014-05-07 | 北京奇艺世纪科技有限公司 | Approximate video detection method and device |
| CN103905694A (en) * | 2014-04-10 | 2014-07-02 | 中央电视台 | Key frame processing method and system |
| CN104239420A (en) * | 2014-10-20 | 2014-12-24 | 北京畅景立达软件技术有限公司 | Video fingerprinting-based video similarity matching method |
| CN104239420B (en) * | 2014-10-20 | 2017-06-06 | 北京畅景立达软件技术有限公司 | A kind of video Similarity Match Method based on video finger print |
| CN104331450A (en) * | 2014-10-29 | 2015-02-04 | 聂秀山 | Video copy detection method based on multimodal features and tensor decomposition |
| CN104504307B (en) * | 2015-01-08 | 2017-09-29 | 北京大学 | Audio frequency and video copy detection method and device based on copy cell |
| CN104504307A (en) * | 2015-01-08 | 2015-04-08 | 北京大学 | Method and device for detecting audio/video copy based on copy cells |
| CN105139014A (en) * | 2015-07-14 | 2015-12-09 | 北京联合大学 | Method for describing image local characteristic descriptor |
| CN105139014B (en) * | 2015-07-14 | 2018-12-04 | 北京联合大学 | A method of calculating image local feature description |
| CN105072455A (en) * | 2015-08-11 | 2015-11-18 | 精硕世纪科技（北京）有限公司 | Video matching method and device |
| CN105227907A (en) * | 2015-08-31 | 2016-01-06 | 电子科技大学 | Based on the nothing supervision anomalous event real-time detection method of video |
| CN105227907B (en) * | 2015-08-31 | 2018-07-27 | 电子科技大学 | Unsupervised anomalous event real-time detection method based on video |
| CN105608423A (en) * | 2015-12-17 | 2016-05-25 | 天脉聚源（北京）科技有限公司 | Video matching method and device |
| CN106021321A (en) * | 2016-05-06 | 2016-10-12 | 成都索贝数码科技股份有限公司 | Picture based online real-time video search method |
| CN106021321B (en) * | 2016-05-06 | 2019-05-21 | 成都索贝数码科技股份有限公司 | The search method of online real-time video based on picture |
| CN106570165A (en) * | 2016-11-07 | 2017-04-19 | 北京航空航天大学 | Content-based video retrieval method and apparatus |
| CN106570165B (en) * | 2016-11-07 | 2019-09-13 | 北京航空航天大学 | A kind of content based video retrieval system method and device |
| CN106649663A (en) * | 2016-12-14 | 2017-05-10 | 大连理工大学 | Video copy detection method based on compact video representation |
| CN106649663B (en) * | 2016-12-14 | 2018-10-16 | 大连理工大学 | A kind of video copying detection method based on compact video characterization |
| CN107135401A (en) * | 2017-03-31 | 2017-09-05 | 北京奇艺世纪科技有限公司 | Key frame extraction method and system |
| CN107135401B (en) * | 2017-03-31 | 2020-03-27 | 北京奇艺世纪科技有限公司 | Key frame selection method and system |
| CN108255970A (en) * | 2017-12-26 | 2018-07-06 | 努比亚技术有限公司 | A kind of video retrieval method, terminal and computer readable storage medium |
| CN110147469A (en) * | 2019-05-14 | 2019-08-20 | 腾讯音乐娱乐科技（深圳）有限公司 | A kind of data processing method, equipment and storage medium |
| CN110147469B (en) * | 2019-05-14 | 2023-08-08 | 腾讯音乐娱乐科技（深圳）有限公司 | Data processing method, device and storage medium |
| CN110175267A (en) * | 2019-06-04 | 2019-08-27 | 黑龙江省七星农场 | A kind of agriculture Internet of Things control processing method based on unmanned aerial vehicle remote sensing technology |
| CN110175267B (en) * | 2019-06-04 | 2020-07-07 | 黑龙江省七星农场 | Agricultural Internet of things control processing method based on unmanned aerial vehicle remote sensing technology |
| CN110321454B (en) * | 2019-08-06 | 2023-03-24 | 北京字节跳动网络技术有限公司 | Video processing method and device, electronic equipment and computer readable storage medium |
| CN110321454A (en) * | 2019-08-06 | 2019-10-11 | 北京字节跳动网络技术有限公司 | Processing method, device, electronic equipment and the computer readable storage medium of video |
| CN111601115A (en) * | 2020-05-12 | 2020-08-28 | 腾讯科技（深圳）有限公司 | Video detection method, related device, equipment and storage medium |
| CN112507875A (en) * | 2020-12-10 | 2021-03-16 | 上海连尚网络科技有限公司 | Method and equipment for detecting video repetition |
| CN112651336A (en) * | 2020-12-25 | 2021-04-13 | 深圳万兴软件有限公司 | Method, device and computer readable storage medium for determining key frame |
| CN112651336B (en) * | 2020-12-25 | 2023-09-29 | 深圳万兴软件有限公司 | Method, apparatus and computer readable storage medium for determining key frame |
| CN113239855A (en) * | 2021-05-27 | 2021-08-10 | 北京字节跳动网络技术有限公司 | Video detection method and device, electronic equipment and storage medium |
| CN113239855B (en) * | 2021-05-27 | 2023-04-18 | 抖音视界有限公司 | Video detection method and device, electronic equipment and storage medium |
| CN113886632A (en) * | 2021-12-03 | 2022-01-04 | 杭州并坚科技有限公司 | Video retrieval matching method based on dynamic programming |
| CN113886632B (en) * | 2021-12-03 | 2022-04-01 | 杭州并坚科技有限公司 | Video retrieval matching method based on dynamic programming |
Also Published As
| Publication number | Publication date |
|---|---|
| CN101394522B (en) | 2010-07-21 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101394522B (en) | Detection method and system for video copy | |
| JP3568117B2 (en) | Method and system for video image segmentation, classification, and summarization | |
| CN112818906A (en) | Intelligent full-media news cataloging method based on multi-mode information fusion understanding | |
| CN109783691B (en) | Video retrieval method for deep learning and Hash coding | |
| CN104504307A (en) | Method and device for detecting audio/video copy based on copy cells | |
| CN112163122A (en) | Method and device for determining label of target video, computing equipment and storage medium | |
| CN101467145A (en) | Method and apparatus for automatically annotating images | |
| Awad et al. | Content-based video copy detection benchmarking at TRECVID | |
| CN113010703A (en) | Information recommendation method and device, electronic equipment and storage medium | |
| CN113591530A (en) | Video detection method and device, electronic equipment and storage medium | |
| CN102236714A (en) | Extensible markup language (XML)-based interactive application multimedia information retrieval method | |
| Haroon et al. | Video scene detection using compact bag of visual word models | |
| Fei et al. | Learning user interest with improved triplet deep ranking and web-image priors for topic-related video summarization | |
| Yang et al. | Keyframe recommendation based on feature intercross and fusion | |
| CN116069973B (en) | Video abstract generation method based on semantic self-mining | |
| Vega et al. | A robust video identification framework using perceptual image hashing | |
| CN115379233B (en) | Big data video information analysis method and system | |
| Li et al. | A confidence based recognition system for TV commercial extraction | |
| CN115379301A (en) | Video processing method and related equipment | |
| Nie et al. | Key-frame based robust video hashing using isometric feature mapping | |
| Liu et al. | Within and between shot information utilisation in video key frame extraction | |
| Gharbi et al. | Key frames extraction based on local features for efficient video summarization | |
| Yang et al. | A repeated video clip identification system | |
| CN117710777B (en) | Model training method, key frame extraction method and device | |
| Tian et al. | VISFF: an approach for video summarization based on feature fusion |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20100721 Termination date: 20210919 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |