CN113065025A - Video duplicate checking method, device, equipment and storage medium - Google Patents
Video duplicate checking method, device, equipment and storage medium Download PDFInfo
- Publication number
- CN113065025A CN113065025A CN202110352954.9A CN202110352954A CN113065025A CN 113065025 A CN113065025 A CN 113065025A CN 202110352954 A CN202110352954 A CN 202110352954A CN 113065025 A CN113065025 A CN 113065025A
- Authority
- CN
- China
- Prior art keywords
- video
- features
- preset
- searched
- feature
- 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.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 66
- 238000012545 processing Methods 0.000 claims abstract description 19
- 238000000605 extraction Methods 0.000 claims description 17
- 238000012549 training Methods 0.000 claims description 11
- 230000006870 function Effects 0.000 claims description 9
- 238000004590 computer program Methods 0.000 claims description 8
- 230000008569 process Effects 0.000 description 12
- 238000010586 diagram Methods 0.000 description 4
- 230000000694 effects Effects 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 4
- 238000004422 calculation algorithm Methods 0.000 description 3
- 230000008878 coupling Effects 0.000 description 3
- 238000010168 coupling process Methods 0.000 description 3
- 238000005859 coupling reaction Methods 0.000 description 3
- HDAJUGGARUFROU-JSUDGWJLSA-L MoO2-molybdopterin cofactor Chemical compound O([C@H]1NC=2N=C(NC(=O)C=2N[C@H]11)N)[C@H](COP(O)(O)=O)C2=C1S[Mo](=O)(=O)S2 HDAJUGGARUFROU-JSUDGWJLSA-L 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- 238000004891 communication Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012552 review Methods 0.000 description 2
- 238000012935 Averaging Methods 0.000 description 1
- 238000003491 array Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 235000019800 disodium phosphate Nutrition 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000011176 pooling Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/70—Information retrieval; Database structures therefor; File system structures therefor of video data
- G06F16/78—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
- G06F16/783—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using metadata automatically derived from the content
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/70—Information retrieval; Database structures therefor; File system structures therefor of video data
- G06F16/71—Indexing; Data structures therefor; Storage structures
Abstract
The invention provides a video duplicate checking method, a video duplicate checking device, video duplicate checking equipment and a storage medium, and relates to the technical field of data processing. The method comprises the following steps: acquiring the important frequency feature to be checked and the important frame feature to be checked of the important video to be checked; searching target video features matched with the important frequency features to be searched in preset video features of an index library according to the important frequency features to be searched, wherein the index library comprises the preset video features corresponding to a plurality of preset videos; and determining whether the video which is repeated frequently with the important image to be checked exists in the index database according to the target frame characteristics corresponding to the target video characteristics and the important image characteristics to be checked. And searching the to-be-searched weighted video features of the to-be-searched video to obtain matched target video features, and then determining whether videos which are repeated with the to-be-searched weighted video exist in the index database according to the target frame features corresponding to the target video features and the to-be-searched weighted frame features. The video characteristics and the frame characteristics are combined to search for the video which is repeated frequently with the important value to be searched, so that the searching accuracy and the user experience are improved.
Description
Technical Field
The invention relates to the technical field of data processing, in particular to a video duplicate checking method, device, equipment and storage medium.
Background
The video fingerprint technology is similar to the finger fingerprint technology and is a unique attribute of videos and people, different videos can be rapidly distinguished and related videos can be searched through the video fingerprint technology, and the video duplicate checking is more and more important to avoid the repetition of the videos.
In the related art, a database is searched according to video features of a video to be retrieved, and a plurality of preset video features matched with the video features in the database are determined, so that a preset video which is repeated with the video to be retrieved in the database is determined.
However, in the related art, the problem of low searching accuracy is easily caused by searching for the repeated video by using the video characteristics, and the user experience is reduced.
Disclosure of Invention
The present invention aims to provide a method, an apparatus, a device and a storage medium for video duplicate checking, so as to solve the problem that in the related art, when a video feature is used to search for a duplicate video, the searching accuracy is low, and the user experience is reduced.
In order to achieve the above purpose, the embodiment of the present invention adopts the following technical solutions:
in a first aspect, an embodiment of the present invention provides a video duplicate checking method, including:
acquiring the important frequency feature to be checked and the important frame feature to be checked of the important video to be checked;
searching target video features matched with the important frequency features to be searched in preset video features of an index library according to the important frequency features to be searched, wherein the index library comprises the preset video features corresponding to a plurality of preset videos;
and determining whether the video which is frequently repeated with the important image to be checked exists in the index database according to the target frame characteristics corresponding to the target video characteristics and the characteristics of the important image to be checked.
Optionally, the retrieving, in an index library, a target video feature matched with the important frequency feature to be searched according to the important frequency feature to be searched includes:
searching repeated video features matched with the important frequency features to be searched in the index database according to the important frequency features to be searched;
determining a repeating group in which the repeated video features are located, wherein the repeating group comprises a plurality of repeated preset video features;
and taking the earliest stored preset video feature in the repeated group as the target video feature.
Optionally, the determining, according to the target frame feature corresponding to the target video feature and the to-be-searched repeated frame feature, whether a video that is frequently repeated with the to-be-searched repeated frame exists in the index library includes:
determining the number of matched frames according to the target frame characteristics and the characteristics of the to-be-checked repeated frames;
judging whether the number of the matched frames is greater than or equal to a preset frame threshold value;
if the number of the matched frames is larger than or equal to the preset frame threshold, determining that videos which are repeated with the important frequency to be searched exist in the index database;
and if the number of the matched frames is smaller than the preset frame threshold, determining that no video which is repeated with the important frequency to be searched exists in the index database, and allocating a new repeat group identifier for the important frequency to be searched.
Optionally, the searching, in the index library, for the repeated video features matched with the important frequency features to be searched according to the important frequency features to be searched includes:
recalling the original video features with the highest matching rate in the index library according to the important video features to be searched;
judging whether the Euclidean distance between the original video features and the important video features to be searched is smaller than or equal to a preset distance threshold value or not;
and if the Euclidean distance is smaller than or equal to the preset distance threshold, determining the original video feature as the repeated video feature.
Optionally, the method further includes:
and if the Euclidean distance is greater than the preset distance threshold, determining that no video which is repeated with the to-be-searched frequency is present in the index database, and allocating a new repeat group identifier for the to-be-searched frequency.
Optionally, before the target video feature matched with the important frequency feature to be searched is retrieved from the preset video features in the index library according to the important frequency feature to be searched, the method further includes:
extracting the convolution characteristics of each frame in the plurality of preset videos by adopting a characteristic extraction network, wherein the characteristic extraction network is obtained by training a training video after data enhancement processing;
determining the preset video characteristics and preset frame characteristics corresponding to the plurality of preset videos according to the convolution characteristics;
and establishing the index library according to the preset frame characteristics and the preset video characteristics.
Optionally, the determining, according to the convolution feature, the preset video feature and the preset frame feature corresponding to the plurality of preset videos includes:
weighting the intermediate features of the convolution features by adopting a preset kernel function to obtain weighted convolution features;
performing feature coding on the weighted convolution features to obtain the preset frame features;
and determining the preset video characteristics of each preset video according to the preset frame characteristics corresponding to each preset video.
In a second aspect, an embodiment of the present invention further provides a video duplicate checking apparatus, including:
the acquisition module is used for acquiring the important frequency feature to be checked and the important frame feature to be checked of the important video to be checked;
the retrieval module is used for retrieving target video features matched with the important frequency features to be searched in preset video features of an index library according to the important frequency features to be searched, wherein the index library comprises the preset video features corresponding to a plurality of preset videos;
and the determining module is used for determining whether the video which is frequently repeated with the to-be-searched importance exists in the index database according to the target frame characteristics corresponding to the target video characteristics and the to-be-searched repeated frame characteristics.
Optionally, the retrieval module is further configured to search, in the index library, a repeated video feature matched with the important frequency feature to be searched according to the important frequency feature to be searched; determining a repeating group in which the repeated video features are located, wherein the repeating group comprises a plurality of repeated preset video features; and taking the earliest stored preset video feature in the repeated group as the target video feature.
Optionally, the determining module is further configured to determine the number of matched frames according to the target frame feature and the to-be-checked repeat frame feature; judging whether the number of the matched frames is greater than or equal to a preset frame threshold value; if the number of the matched frames is larger than or equal to the preset frame threshold, determining that videos which are repeated with the important frequency to be searched exist in the index database; and if the number of the matched frames is smaller than the preset frame threshold, determining that no video which is repeated with the important frequency to be searched exists in the index database, and allocating a new repeat group identifier for the important frequency to be searched.
Optionally, the retrieval module is further configured to recall, in the index library, the original video feature with the highest matching rate according to the important video feature to be searched; judging whether the Euclidean distance between the original video features and the important video features to be searched is smaller than or equal to a preset distance threshold value or not; and if the Euclidean distance is smaller than or equal to the preset distance threshold, determining the original video feature as the repeated video feature.
Optionally, the apparatus further comprises:
and the distribution module is used for determining that no video which is repeated with the to-be-searched valued frequency exists in the index database and distributing a new repeat group identifier for the to-be-searched valued frequency if the Euclidean distance is greater than the preset distance threshold.
Optionally, the apparatus further comprises:
the extraction module is used for extracting the convolution characteristics of each frame in the plurality of preset videos by adopting a characteristic extraction network, and the characteristic extraction network is obtained by training a training video after data enhancement processing;
a first determining module, configured to determine the preset video features and the preset frame features corresponding to the multiple preset videos according to the convolution features;
and the establishing module is used for establishing the index library according to the preset frame characteristics and the preset video characteristics.
Optionally, the first determining module is configured to perform weighting processing on the intermediate feature of the convolution feature by using a preset kernel function, so as to obtain a weighted convolution feature; performing feature coding on the weighted convolution features to obtain preset frame features; and determining the preset video characteristics of each preset video according to the preset frame characteristics corresponding to each preset video.
In a third aspect, an embodiment of the present invention further provides a video duplicate checking device, including: a memory and a processor, wherein the memory stores a computer program executable by the processor, and the processor implements the video duplicate checking method according to any one of the first aspect when executing the computer program.
In a fourth aspect, an embodiment of the present invention further provides a storage medium, where a computer program is stored on the storage medium, and when the computer program is read and executed, the video duplicate checking method according to any one of the above first aspects is implemented.
The invention has the beneficial effects that: the embodiment of the invention provides a video duplicate checking method, which comprises the following steps: acquiring the important frequency feature to be checked and the important frame feature to be checked of the important video to be checked; searching target video features matched with the important frequency features to be searched in preset video features of an index library according to the important frequency features to be searched, wherein the index library comprises the preset video features corresponding to a plurality of preset videos; and determining whether the video which is repeated frequently with the important image to be checked exists in the index database according to the target frame characteristics corresponding to the target video characteristics and the important image characteristics to be checked. And searching the to-be-searched weighted video features of the to-be-searched video to obtain matched target video features, and then determining whether videos which are repeated with the to-be-searched weighted video exist in the index database according to the target frame features corresponding to the target video features and the to-be-searched weighted frame features. The video characteristics and the frame characteristics are combined to search for the video which is repeated frequently with the importance to be searched, so that the searching accuracy is improved, and further the user experience is improved.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings needed to be used in the embodiments will be briefly described below, it should be understood that the following drawings only illustrate some embodiments of the present invention and therefore should not be considered as limiting the scope, and for those skilled in the art, other related drawings can be obtained according to the drawings without inventive efforts.
Fig. 1 is a schematic flowchart of a video duplicate checking method according to an embodiment of the present invention;
fig. 2 is a schematic flowchart of a video duplicate checking method according to an embodiment of the present invention;
fig. 3 is a schematic flowchart of a video duplicate checking method according to an embodiment of the present invention;
fig. 4 is a schematic flowchart of a video duplicate checking method according to an embodiment of the present invention;
fig. 5 is a schematic flowchart of a video duplicate checking method according to an embodiment of the present invention;
fig. 6 is a schematic flowchart of a video duplicate checking method according to an embodiment of the present invention;
fig. 7 is a schematic structural diagram of a video duplicate checking apparatus according to an embodiment of the present invention;
fig. 8 is a schematic structural diagram of a video duplicate checking device according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, but not all, embodiments of the present invention.
Thus, the following detailed description of the embodiments of the present application, presented in the accompanying drawings, is not intended to limit the scope of the claimed application, but is merely representative of selected embodiments of the application. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.
In the description of the present application, it should be noted that if the terms "upper", "lower", etc. are used for indicating the orientation or positional relationship based on the orientation or positional relationship shown in the drawings or the orientation or positional relationship which is usually arranged when the product of the application is used, the description is only for convenience of describing the application and simplifying the description, but the indication or suggestion that the referred device or element must have a specific orientation, be constructed in a specific orientation and operation, and thus, cannot be understood as the limitation of the application.
Furthermore, the terms "first," "second," and the like in the description and in the claims, as well as in the drawings, are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used is interchangeable under appropriate circumstances such that the embodiments of the invention described herein are capable of operation in sequences other than those illustrated or described herein. Furthermore, the terms "comprises," "comprising," and "having," and any variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed, but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.
It should be noted that the features of the embodiments of the present application may be combined with each other without conflict.
The method and the device aim at the problem that in the related technology, repeated videos are searched by adopting video characteristics, the searching accuracy rate is low easily, and the user experience is reduced. The embodiment of the application provides a video duplicate checking method, which is characterized in that matched target video features are obtained based on the retrieval of the to-be-checked frequency-weighted features of the to-be-checked video, and then whether videos which are repeated with the to-be-checked frequency are in an index library or not is determined according to the target frame features corresponding to the target video features and the to-be-checked frequency-weighted features. The video characteristics and the frame characteristics are combined to search for the video which is repeated frequently with the importance to be searched, so that the searching accuracy is improved, and further the user experience is improved.
The embodiment of the application provides a video duplicate checking method, wherein an execution main body of the video duplicate checking method can be video duplicate checking equipment, and the video duplicate checking equipment can be a terminal or a server, and can also be other types of equipment with processing functions. For example, when the video duplication checking device is a terminal, the video duplication checking device may be a desktop computer, a notebook computer, or a tablet computer, and the video duplication checking method provided in the embodiment of the present application is explained below with the terminal as an execution subject.
Fig. 1 is a schematic flow chart of a video duplicate checking method according to an embodiment of the present invention, as shown in fig. 1, the method may include:
s101, acquiring the important frequency feature to be checked and the important frame feature to be checked of the important video to be checked.
The important frequency to be searched can also be referred to as a video to be retrieved, the important frequency to be searched can be a long video or a short video, and the duration or the size of the important frequency to be searched is not particularly limited in the embodiment of the present application.
In some embodiments, the terminal may extract a convolution feature of the duplicate video to be checked by using the feature extraction network, determine a duplicate frame feature to be checked according to the convolution feature of the duplicate video to be checked, and determine the duplicate video feature to be checked based on the duplicate frame feature to be checked.
It should be noted that a to-be-checked significant frequency may include at least one to-be-checked repeated frame, each to-be-checked repeated frame has a corresponding to-be-checked repeated frame feature, that is, there is at least one to-be-checked repeated frame feature, and averaging the at least one to-be-checked repeated frame feature may obtain the to-be-checked significant frequency feature.
And S102, searching target video characteristics matched with the important frequency characteristics to be searched in preset video characteristics of an index library according to the important frequency characteristics to be searched.
The index library comprises preset video features corresponding to a plurality of preset videos.
In a possible implementation manner, the terminal may calculate similarity between each preset video feature and the important frequency feature to be searched, obtain similarity ranking, and determine a target video feature matched with the important frequency feature to be searched based on the similarity ranking.
Optionally, the target video feature may be a preset video feature with the highest similarity to the important video feature to be searched, or the target video feature may be a video feature determined according to the preset video feature with the highest similarity.
S103, determining whether the video which is frequently repeated with the to-be-searched important image exists in the index database according to the target frame characteristics corresponding to the target video characteristics and the to-be-searched repeated frame characteristics.
After the target video characteristics are determined, the terminal can perform real-time calculation according to the target video characteristics to determine the target frame characteristics corresponding to the target video characteristics. Each target frame in the target video corresponds to one target frame characteristic, and at least one target frame characteristic exists; and each frame to be checked in the video to be checked corresponds to one frame to be checked characteristic, and at least one frame to be checked characteristic exists.
In some embodiments, the terminal may match each target frame feature with the to-be-checked repeat frame feature to obtain at least one matching result, and determine whether the at least one matching result meets a preset condition; if so, determining that the video which is repeated frequently with the important value to be searched exists in the index database; if not, determining that the video which is repeated with the important frequency to be searched does not exist in the index database. And each matching result is used for representing that the corresponding target frame characteristic is similar to or dissimilar to the characteristic of the frame to be checked.
In addition, if the video overlapping with the important frequency to be checked exists in the index database, the target video corresponding to the target video feature may be the video overlapping with the important frequency to be checked.
In summary, an embodiment of the present invention provides a video duplicate checking method, including: acquiring the important frequency feature to be checked and the important frame feature to be checked of the important video to be checked; searching target video features matched with the important frequency features to be searched in preset video features of an index library according to the important frequency features to be searched, wherein the index library comprises preset video features and preset frame features corresponding to a plurality of preset videos; and determining whether the video which is repeated frequently with the important image to be checked exists in the index database according to the target frame characteristics corresponding to the target video characteristics and the important image characteristics to be checked. And searching the to-be-searched weighted video features of the to-be-searched video to obtain matched target video features, and then determining whether videos which are repeated with the to-be-searched weighted video exist in the index database according to the target frame features corresponding to the target video features and the to-be-searched weighted frame features. The video characteristics and the frame characteristics are combined to search for the video which is repeated frequently with the importance to be searched, so that the searching accuracy is improved, and further the user experience is improved.
Optionally, in the above description about S101, the process of determining the characteristics of the review frame according to the convolution characteristics of the review video may include: and performing weighting processing on the intermediate features of the convolution features of the video to be repeated by adopting a preset kernel function to obtain weighted convolution features of the weighted important frequency to be repeated, and performing feature coding on the weighted convolution features of the video to be repeated to obtain features of the frame to be repeated.
It should be noted that, the intermediate features of the convolution features of the video to be searched are weighted, so that the obtained frequency-to-be-searched features and the obtained frame-to-be-searched features are more accurate, the frequency-to-be-searched can be more accurately represented, and the accuracy of searching can be further improved.
Optionally, fig. 2 is a flowchart of a video duplicate checking method according to an embodiment of the present invention, as shown in fig. 2, a process of retrieving, in the index library, a target video feature matched with a to-be-checked important frequency feature in S102 according to the to-be-checked important frequency feature may include:
s201, searching repeated video features matched with the important frequency features to be searched in an index library according to the important frequency features to be searched.
Wherein, the repeated video feature may be one of the preset video features.
In some embodiments, the terminal may obtain a similarity ranking for the similarity between the importance frequency feature to be searched and each preset video feature, and directly use the preset video feature corresponding to the highest similarity in the similarity ranking as the repeat video feature.
Optionally, the terminal may further determine the repeated video feature according to the preset video feature corresponding to the highest similarity, which is not specifically limited in this embodiment of the present application.
S202, determining a repeating group where the repeated video features are located, wherein the repeating group comprises a plurality of repeated preset video features.
Wherein, the index library can include: a plurality of preset repeating groups, each preset repeating group may include: there is at least one preset video feature that repeats, each preset repeat group having a corresponding preset repeat group identification.
In this embodiment of the application, the terminal may determine a repeat group identifier where the repeated video feature is located, and determine a repeat group where the repeated video feature is located, where the repeat group may be one of a plurality of preset repeat groups.
And S203, taking the earliest stored preset video feature in the repeated group as a target video feature.
It should be noted that, each of the preset video features in the repeating group has a corresponding storage time, and the preset video feature with the earliest storage time is taken as the target video feature.
In summary, the preset video features stored at the earliest time are used as the target video features, so that the determined target video features and the important video features to be searched can be matched more, and the determined target video features can be more accurate.
Optionally, fig. 3 is a schematic flowchart of a video duplicate checking method according to an embodiment of the present invention, as shown in fig. 3, a process of determining whether a video that is repeated with the duplicate to be checked exists in the index database according to a target frame feature and a duplicate to be checked frame feature corresponding to a target video feature in the above S103 may include:
s301, determining the number of matched frames according to the target frame characteristics and the to-be-checked repeated frame characteristics.
The number of the target frame features may be at least one, and the number of the frame features to be checked may also be at least one.
In a possible implementation manner, a target frame feature and a corresponding to-be-checked repeat frame feature are respectively matched to obtain at least one matching result, each matching result represents that one target frame feature is matched with or not matched with the corresponding to-be-checked repeat frame feature, and the matching results representing matching in the at least one matching result are counted to obtain the number of matched frames.
S302, judging whether the number of the matched frames is larger than or equal to a preset frame threshold value.
It should be noted that the preset frame threshold may be set according to actual requirements, may be set according to empirical values, and may also be set in other manners, which is not specifically limited in this embodiment of the present application.
And S303, if the number of the matched frames is greater than or equal to a preset frame threshold value, determining that the video which is repeated with the important frequency to be searched exists in the index database.
In the embodiment of the application, if the number of the matched frames is greater than or equal to the preset frame threshold, it is indicated that the matching degree of the characteristics of the to-be-searched repeated frame and the characteristics of the target frame is high, it is determined that a video which is frequently repeated with the to-be-searched important exists in the index database, and the repeat group identifier where the characteristics of the target frame are located is used as the repeat group identifier of the to-be-searched repeated video.
S304, if the number of the matched frames is smaller than the preset frame threshold value, determining that the video which is repeated with the important frequency to be searched does not exist in the index database, and allocating a new repeat group identifier for the important frequency to be searched.
Optionally, the terminal may perform the processes of S301 to S304 by using a preset algorithm, where the preset algorithm may be a smith waterman (smith-waterman) algorithm, and may be used for frame matching, and even if the video has a misalignment, a good matching result may still be obtained, so that the accuracy of frame matching may be improved, and the accuracy of video duplicate checking may be improved.
Optionally, fig. 4 is a flowchart of a video duplicate checking method according to an embodiment of the present invention, as shown in fig. 4, a process of searching for a duplicate video feature matching a to-be-checked important frequency feature in an index library according to the to-be-checked important frequency feature in S201 may include:
s401, recalling the original video features with the highest matching rate in an index library according to the important video features to be searched.
S402, judging whether the Euclidean distance between the original video feature and the important video feature to be checked is smaller than or equal to a preset distance threshold value.
And S403, if the Euclidean distance is smaller than or equal to a preset distance threshold, determining that the original video feature is a repeated video feature.
Wherein, after determining that the original video feature is the repeated video feature, the terminal may continue to perform the processes of S202 to S203.
It should be noted that the preset distance threshold may be set according to actual requirements, may be set according to empirical values, and may also be set in other manners, which is not specifically limited in the embodiment of the present application.
Optionally, the method further includes:
and if the Euclidean distance is greater than a preset distance threshold value, determining that no video which is repeated with the to-be-searched frequency is existed in the index database, and allocating a new repeat group identifier for the to-be-searched frequency.
Optionally, fig. 5 is a flowchart of a video duplicate checking method according to an embodiment of the present invention, as shown in fig. 5, before the process of retrieving, in the preset video features in the index library, the target video feature matched with the important video feature to be checked according to the important video feature to be checked in S102, the method may further include:
s501, extracting convolution characteristics of each frame in a plurality of preset videos by adopting a characteristic extraction network, wherein the characteristic extraction network is obtained by training a training video after data enhancement processing.
The feature extraction network trained by the training video after data enhancement processing can better meet the actual service requirements. The data enhancement processing may include a combination of at least one of: adding black edges, adding fuzzy edges, cutting, adding logo (logo), and the like.
In addition, the terminal may perform feature extraction in an unsupervised learning-based manner, for example, the unsupervised learning manner may be the MOCO v2 manner.
In some embodiments, the terminal may adopt resnet34 as a MOCO v2 backbone network, i.e., a feature extraction network, and extract a convolution feature of each frame in a plurality of preset videos through the feature extraction network, where resnet34 may well balance speed and accuracy.
It should be noted that the full-connection feature focuses more on the global feature, and compared with the use of the full-connection feature, the selection of the convolution feature can focus more on the spatial information expression, and meanwhile, the expression of the local feature is more important for video deduplication.
S502, according to the convolution characteristics, the preset video characteristics and the preset frame characteristics corresponding to the preset videos are determined.
The terminal can process the convolution characteristics to obtain preset frame characteristics, and can acquire the preset video characteristics based on the preset frame characteristics, so that the preset video characteristics and the preset frame characteristics can be determined.
S503, establishing an index library according to the preset frame characteristics and the preset video characteristics.
In the embodiment of the application, the terminal can store the preset frame characteristics and the preset video characteristics, and index by using the preset frame characteristics and the preset video characteristics to obtain the established index library. The terminal may establish the index by using a preset frame, and the preset frame may be a faiss (a frame).
Optionally, fig. 6 is a schematic flow chart of a video duplicate checking method according to an embodiment of the present invention, as shown in fig. 6, a process of determining preset video features and preset frame features corresponding to a plurality of preset videos according to a convolution feature in the above S502 may include:
s601, weighting the intermediate features of the convolution features by adopting a preset kernel function to obtain weighted convolution features.
It should be noted that the convolution feature is insensitive to special effects such as clipping, black edges, fuzzy edges, logo, and watermark, which easily results in a large number of missed calls, and the terminal may use a preset kernel function to perform weighting processing on the intermediate feature of the convolution feature, so that the obtained weighted convolution feature may highlight the central area of the video feature.
And S602, performing feature coding on the weighted convolution features to obtain preset frame features.
The weighted convolution characteristics bring a large burden to subsequent storage and calculation, so characteristic coding is required. For example, the weighted convolution characteristic is [14, 14, 512], and the default frame characteristic may be [1, 512 ].
In some embodiments, the terminal may perform feature coding on the weighted convolution features in a sliding window manner, so as to preserve the local features and the spatial expression, and meanwhile, the feature expression is better. Alternatively, the terminal may be characterized by Rmac Pooling.
S603, determining the preset video characteristics of each preset video according to the preset frame characteristics corresponding to each preset video.
Each preset video may correspond to at least one preset frame feature.
In this embodiment of the application, the terminal may average at least one preset frame feature corresponding to each preset video to obtain a preset video feature of each preset video.
It should be noted that, by using a preset kernel function to perform weighting processing on the intermediate features of the convolution features, the central region of the video features can be highlighted, so that the determination of the preset video features and the preset frame features is more accurate, and the video duplicate checking is more accurate.
In summary, an embodiment of the present invention provides a video duplicate checking method, including: acquiring the important frequency feature to be checked and the important frame feature to be checked of the important video to be checked; searching target video features matched with the important frequency features to be searched in preset video features of an index library according to the important frequency features to be searched, wherein the index library comprises preset video features and preset frame features corresponding to a plurality of preset videos; and determining whether the video which is repeated frequently with the important image to be checked exists in the index database according to the target frame characteristics corresponding to the target video characteristics and the important image characteristics to be checked. And searching the to-be-searched weighted video features of the to-be-searched video to obtain matched target video features, and then determining whether videos which are repeated with the to-be-searched weighted video exist in the index database according to the target frame features corresponding to the target video features and the to-be-searched weighted frame features. The video characteristics and the frame characteristics are combined to search for the video which is repeated frequently with the importance to be searched, so that the searching accuracy is improved, and further the user experience is improved.
Moreover, the central area of the video features can be highlighted, so that the preset video features and the preset frame features are determined more accurately, the video duplicate checking is more accurate, and the accuracy and the recall rate of the video duplicate checking are greatly improved.
For specific implementation processes and technical effects of the video duplicate checking device, the apparatus, the storage medium, and the like for executing the video duplicate checking method provided by the present application, reference is made to the relevant contents of the video duplicate checking method, and details are not described below.
Fig. 7 is a schematic structural diagram of a video duplicate checking apparatus according to an embodiment of the present invention, as shown in fig. 7, the apparatus may include:
an obtaining module 701, configured to obtain a to-be-checked importance frequency feature and a to-be-checked importance frame feature of a to-be-checked importance video; a retrieval module 702, configured to retrieve, from preset video features of an index library, a target video feature matched with a to-be-searched important frequency feature according to the to-be-searched important frequency feature, where the index library includes preset video features corresponding to multiple preset videos;
the determining module 703 is configured to determine whether a video that is frequently repeated with the to-be-searched importance exists in the index database according to the target frame feature and the to-be-searched repeated frame feature corresponding to the target video feature.
Optionally, the retrieving module 702 is further configured to search, in the index database, a repeated video feature matched with the important frequency feature to be searched according to the important frequency feature to be searched; determining a repeating group in which the repeated video features are located, wherein the repeating group comprises a plurality of repeated preset video features; and taking the earliest stored preset video feature in the repeated group as a target video feature.
Optionally, the determining module 703 is further configured to determine the number of the matched frames according to the target frame feature and the to-be-checked repeat frame feature; judging whether the number of the matched frames is greater than or equal to a preset frame threshold value or not; if the number of the matched frames is larger than or equal to a preset frame threshold value, determining that videos which are repeated frequently with the important image to be searched exist in the index database; and if the number of the matched frames is less than the preset frame threshold, determining that no video which is repeated with the to-be-checked frequency is existed in the index database, and allocating a new repeat group identifier for the to-be-checked frequency.
Optionally, the retrieval module 702 is further configured to recall, in the index library, the original video feature with the highest matching rate according to the important video feature to be searched; judging whether the Euclidean distance between the original video features and the important frequency features to be checked is smaller than or equal to a preset distance threshold value or not; and if the Euclidean distance is smaller than or equal to the preset distance threshold, determining the original video features as the repeated video features.
Optionally, the apparatus further comprises:
and the distribution module is used for determining that the video which is repeated with the important frequency to be searched does not exist in the index database if the Euclidean distance is greater than a preset distance threshold value, and distributing a new repeat group identifier for the important frequency to be searched.
Optionally, the apparatus further comprises:
the extraction module is used for extracting the convolution characteristics of each frame in a plurality of preset videos by adopting a characteristic extraction network, and the characteristic extraction network is obtained by training a training video after data enhancement processing;
the first determining module is used for determining preset video characteristics and preset frame characteristics corresponding to a plurality of preset videos according to the convolution characteristics;
and the establishing module is used for establishing an index library according to the preset frame characteristics and the preset video characteristics.
Optionally, the first determining module is configured to perform weighting processing on the intermediate feature of the convolution feature by using a preset kernel function, so as to obtain a weighted convolution feature; performing feature coding on the weighted convolution features to obtain preset frame features; and determining the preset video characteristics of each preset video according to the preset frame characteristics corresponding to each preset video.
The above-mentioned apparatus is used for executing the method provided by the foregoing embodiment, and the implementation principle and technical effect are similar, which are not described herein again.
These above modules may be one or more integrated circuits configured to implement the above methods, such as: one or more Application Specific Integrated Circuits (ASICs), or one or more microprocessors (DSPs), or one or more Field Programmable Gate Arrays (FPGAs), among others. For another example, when one of the above modules is implemented in the form of a Processing element scheduler code, the Processing element may be a general-purpose processor, such as a Central Processing Unit (CPU) or other processor capable of calling program code. For another example, these modules may be integrated together and implemented in the form of a system-on-a-chip (SOC).
Fig. 8 is a schematic structural diagram of a video duplicate checking apparatus according to an embodiment of the present invention, and as shown in fig. 8, the video duplicate checking apparatus may include: a processor 801 and a memory 802.
The memory 802 is used for storing programs, and the processor 801 calls the programs stored in the memory 802 to execute the above-mentioned method embodiments. The specific implementation and technical effects are similar, and are not described herein again.
Optionally, the invention also provides a program product, for example a computer-readable storage medium, comprising a program which, when being executed by a processor, is adapted to carry out the above-mentioned method embodiments.
In the embodiments provided in the present invention, it should be understood that the disclosed apparatus and method may be implemented in other ways. For example, the above-described apparatus embodiments are merely illustrative, and for example, the division of the units is only one logical division, and other divisions may be realized in practice, for example, a plurality of units or components may be combined or integrated into another system, or some features may be omitted, or not executed. In addition, the shown or discussed mutual coupling or direct coupling or communication connection may be an indirect coupling or communication connection through some interfaces, devices or units, and may be in an electrical, mechanical or other form.
The units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of the embodiment.
In addition, functional units in the embodiments of the present invention may be integrated into one processing unit, or each unit may exist alone physically, or two or more units are integrated into one unit. The integrated unit can be realized in a form of hardware, or in a form of hardware plus a software functional unit.
The integrated unit implemented in the form of a software functional unit may be stored in a computer readable storage medium. The software functional unit is stored in a storage medium and includes several instructions to enable a computer device (which may be a personal computer, a server, or a network device) or a processor (processor) to execute some steps of the methods according to the embodiments of the present invention. And the aforementioned storage medium includes: a U disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and other various media capable of storing program codes.
The above is only a preferred embodiment of the present invention, and is not intended to limit the present invention, and various modifications and changes will occur to those skilled in the art. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims (10)
1. A video duplicate checking method is characterized by comprising the following steps:
acquiring the important frequency feature to be checked and the important frame feature to be checked of the important video to be checked;
searching target video features matched with the important frequency features to be searched in preset video features of an index library according to the important frequency features to be searched, wherein the index library comprises the preset video features corresponding to a plurality of preset videos;
and determining whether the video which is frequently repeated with the important image to be checked exists in the index database according to the target frame characteristics corresponding to the target video characteristics and the characteristics of the important image to be checked.
2. The method according to claim 1, wherein the retrieving, in an index database, the target video feature matching the important frequency feature to be searched comprises:
searching repeated video features matched with the important frequency features to be searched in the index database according to the important frequency features to be searched;
determining a repeating group in which the repeated video features are located, wherein the repeating group comprises a plurality of repeated preset video features;
and taking the earliest stored preset video feature in the repeated group as the target video feature.
3. The method according to claim 1, wherein the determining whether the video that is repeated frequently with the value to be checked exists in the index database according to the target frame feature corresponding to the target video feature and the frame feature to be checked comprises:
determining the number of matched frames according to the target frame characteristics and the characteristics of the to-be-checked repeated frames;
judging whether the number of the matched frames is greater than or equal to a preset frame threshold value;
if the number of the matched frames is larger than or equal to the preset frame threshold, determining that videos which are repeated with the important frequency to be searched exist in the index database;
and if the number of the matched frames is smaller than the preset frame threshold, determining that no video which is repeated with the important frequency to be searched exists in the index database, and allocating a new repeat group identifier for the important frequency to be searched.
4. The method according to claim 2, wherein said searching for the repeated video features matching the important frequency features to be searched in the index database according to the important frequency features to be searched comprises:
recalling the original video features with the highest matching rate in the index library according to the important video features to be searched;
judging whether the Euclidean distance between the original video features and the important video features to be searched is smaller than or equal to a preset distance threshold value or not;
and if the Euclidean distance is smaller than or equal to the preset distance threshold, determining the original video feature as the repeated video feature.
5. The method of claim 4, further comprising:
and if the Euclidean distance is greater than the preset distance threshold, determining that no video which is repeated with the to-be-searched frequency is present in the index database, and allocating a new repeat group identifier for the to-be-searched frequency.
6. The method according to claim 1, wherein before the retrieving, from the preset video features in the index library, the target video feature matching the important frequency feature to be searched, the method further comprises:
extracting the convolution characteristics of each frame in the plurality of preset videos by adopting a characteristic extraction network, wherein the characteristic extraction network is obtained by training a training video after data enhancement processing;
determining the preset video characteristics and preset frame characteristics corresponding to the plurality of preset videos according to the convolution characteristics;
and establishing the index library according to the preset frame characteristics and the preset video characteristics.
7. The method according to claim 6, wherein the determining the preset video features and the preset frame features corresponding to the plurality of preset videos according to the convolution features comprises:
weighting the intermediate features of the convolution features by adopting a preset kernel function to obtain weighted convolution features;
performing feature coding on the weighted convolution features to obtain the preset frame features;
and determining the preset video characteristics of each preset video according to the preset frame characteristics corresponding to each preset video.
8. A video duplicate checking apparatus, comprising:
the acquisition module is used for acquiring the important frequency feature to be checked and the important frame feature to be checked of the important video to be checked;
the retrieval module is used for retrieving target video features matched with the important frequency features to be searched in preset video features of an index library according to the important frequency features to be searched, wherein the index library comprises the preset video features corresponding to a plurality of preset videos;
and the determining module is used for determining whether the video which is frequently repeated with the to-be-searched importance exists in the index database according to the target frame characteristics corresponding to the target video characteristics and the to-be-searched repeated frame characteristics.
9. A video duplication checking apparatus, comprising: a memory storing a computer program executable by the processor, and a processor implementing the video duplication checking method of any one of claims 1 to 7 when the computer program is executed by the processor.
10. A storage medium having a computer program stored thereon, wherein the computer program is read and executed to implement the video duplication checking method according to any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110352954.9A CN113065025A (en) | 2021-03-31 | 2021-03-31 | Video duplicate checking method, device, equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110352954.9A CN113065025A (en) | 2021-03-31 | 2021-03-31 | Video duplicate checking method, device, equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113065025A true CN113065025A (en) | 2021-07-02 |
Family
ID=76565029
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110352954.9A Pending CN113065025A (en) | 2021-03-31 | 2021-03-31 | Video duplicate checking method, device, equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113065025A (en) |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103617233A (en) * | 2013-11-26 | 2014-03-05 | 烟台中科网络技术研究所 | Method and device for detecting repeated video based on semantic content multilayer expression |
| CN107665261A (en) * | 2017-10-25 | 2018-02-06 | 北京奇虎科技有限公司 | Video duplicate checking method and device |
| CN107705805A (en) * | 2017-10-25 | 2018-02-16 | 北京奇虎科技有限公司 | Audio duplicate checking method and device |
| CN107943849A (en) * | 2017-11-03 | 2018-04-20 | 小草数语（北京）科技有限公司 | The search method and device of video file |
| CN109189991A (en) * | 2018-08-17 | 2019-01-11 | 百度在线网络技术（北京）有限公司 | Repeat video frequency identifying method, device, terminal and computer readable storage medium |
| CN109275008A (en) * | 2018-09-17 | 2019-01-25 | 青岛海信电器股份有限公司 | A kind of method and apparatus of audio-visual synchronization |
| CN109816011A (en) * | 2019-01-21 | 2019-05-28 | 厦门美图之家科技有限公司 | Generate the method and video key frame extracting method of portrait parted pattern |
| CN110490119A (en) * | 2019-08-14 | 2019-11-22 | 腾讯科技（深圳）有限公司 | Repeat video marker method, apparatus and computer readable storage medium |
| CN110502664A (en) * | 2019-08-27 | 2019-11-26 | 腾讯科技（深圳）有限公司 | Video tab indexes base establishing method, video tab generation method and device |
| CN110688524A (en) * | 2019-09-24 | 2020-01-14 | 深圳市网心科技有限公司 | Video retrieval method and device, electronic equipment and storage medium |
-
2021
- 2021-03-31 CN CN202110352954.9A patent/CN113065025A/en active Pending
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103617233A (en) * | 2013-11-26 | 2014-03-05 | 烟台中科网络技术研究所 | Method and device for detecting repeated video based on semantic content multilayer expression |
| CN107665261A (en) * | 2017-10-25 | 2018-02-06 | 北京奇虎科技有限公司 | Video duplicate checking method and device |
| CN107705805A (en) * | 2017-10-25 | 2018-02-16 | 北京奇虎科技有限公司 | Audio duplicate checking method and device |
| CN107943849A (en) * | 2017-11-03 | 2018-04-20 | 小草数语（北京）科技有限公司 | The search method and device of video file |
| CN109189991A (en) * | 2018-08-17 | 2019-01-11 | 百度在线网络技术（北京）有限公司 | Repeat video frequency identifying method, device, terminal and computer readable storage medium |
| CN109275008A (en) * | 2018-09-17 | 2019-01-25 | 青岛海信电器股份有限公司 | A kind of method and apparatus of audio-visual synchronization |
| CN109816011A (en) * | 2019-01-21 | 2019-05-28 | 厦门美图之家科技有限公司 | Generate the method and video key frame extracting method of portrait parted pattern |
| CN110490119A (en) * | 2019-08-14 | 2019-11-22 | 腾讯科技（深圳）有限公司 | Repeat video marker method, apparatus and computer readable storage medium |
| CN110502664A (en) * | 2019-08-27 | 2019-11-26 | 腾讯科技（深圳）有限公司 | Video tab indexes base establishing method, video tab generation method and device |
| CN110688524A (en) * | 2019-09-24 | 2020-01-14 | 深圳市网心科技有限公司 | Video retrieval method and device, electronic equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108491433B (en) | Chat response method, electronic device and storage medium | |
| CN108334533B (en) | Keyword extraction method and device, storage medium and electronic device | |
| WO2020006835A1 (en) | Customer service method, apparatus, and device for engaging in multiple rounds of question and answer, and storage medium | |
| CN109086394B (en) | Search ranking method and device, computer equipment and storage medium | |
| US9348898B2 (en) | Recommendation system with dual collaborative filter usage matrix | |
| CN109634698B (en) | Menu display method and device, computer equipment and storage medium | |
| CN110377558B (en) | Document query method, device, computer equipment and storage medium | |
| CN112883258A (en) | Information recommendation method and device, electronic equipment and storage medium | |
| CN106033455B (en) | Method and equipment for processing user operation information | |
| CN115809371A (en) | Learning demand determination method and system based on data analysis | |
| CN109299238B (en) | Data query method and device | |
| CN110209780B (en) | Question template generation method and device, server and storage medium | |
| CN111708942B (en) | Multimedia resource pushing method, device, server and storage medium | |
| CN110674388A (en) | Mapping method and device for push item, storage medium and terminal equipment | |
| CN113032524A (en) | Trademark infringement identification method, terminal device and storage medium | |
| CN111737443A (en) | Answer text processing method and device and key text determining method | |
| CN108170664B (en) | Key word expansion method and device based on key words | |
| CN113065025A (en) | Video duplicate checking method, device, equipment and storage medium | |
| CN110895538A (en) | Data retrieval method, device, storage medium and processor | |
| CN115130455A (en) | Article processing method and device, electronic equipment and storage medium | |
| CN114741489A (en) | Document retrieval method, document retrieval device, storage medium and electronic equipment | |
| CN111428120B (en) | Information determination method and device, electronic equipment and storage medium | |
| US20180276294A1 (en) | Information processing apparatus, information processing system, and information processing method | |
| CN111475409B (en) | System test method, device, electronic equipment and storage medium | |
| CN114626340B (en) | Behavior feature extraction method based on mobile phone signaling and related device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |