CN109743580A - A kind of method for processing video frequency and device, storage medium and processor - Google Patents
A kind of method for processing video frequency and device, storage medium and processor Download PDFInfo
- Publication number
- CN109743580A CN109743580A CN201811585170.5A CN201811585170A CN109743580A CN 109743580 A CN109743580 A CN 109743580A CN 201811585170 A CN201811585170 A CN 201811585170A CN 109743580 A CN109743580 A CN 109743580A
- Authority
- CN
- China
- Prior art keywords
- privacy
- target video
- video
- content
- target
- 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
Landscapes
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
This application discloses a kind of method for processing video frequency and devices, storage medium and processor.This method comprises: obtaining target video to be processed;The privacy content in target video is positioned using privacy location model, and exports the privacy location information of target video;According to privacy location information, to the privacy content in target video carry out half it is destructive remove, the target video that obtains that treated.By the application, when solving the privacy content removed in video in the related technology, the low problem for the treatment of effeciency.
Description
Technical field
This application involves technical field of video processing, are situated between in particular to a kind of method for processing video frequency and device, storage
Matter and processor.
Background technique
Each large enterprises have accumulated video data of magnanimity, such as monitoring, advertisement, films and television programs etc., need to entrust third party
Company such as is identified to video, is classified at the structured works.In the process, video data can be supplied to third party by enterprise,
Assessment, training, test etc. before for identification.Due to there may be sensitive privacy information for enterprise in video,
There can be serious consequence except this kind of leakage of information to enterprise.
In the prior art, the leakage of private information in order to prevent in video causes damages, the technology that on the one hand can be used
Means are that video is stamped to relevant watermark, once leakage of private information can trace back to the source of leakage, and carry out phase
That answers calls to account.On the other hand the technological means that can be used is manually to edit video, and the related content of privacy information is smeared
It removes.
All there are certain drawbacks in the above-mentioned means for preventing leakage of private information, can only using the first technological means
It calls to account afterwards, consequence caused by leakage not can avoid.Using second of technological means, video privacy information is manually edited,
And it is likely to occur privacy content omission, situations such as mistake is erased, low efficiency, effect is poor.
For the above problem present in the relevant technologies, currently no effective solution has been proposed.
Summary of the invention
The main purpose of the application is to provide a kind of method for processing video frequency and device, storage medium and processor, with solution
When certainly removing the privacy content in video in the related technology, the low problem for the treatment of effeciency.
To achieve the goals above, according to the one aspect of the application, a kind of method for processing video frequency is provided.This method packet
It includes: obtaining target video to be processed;The privacy content in target video is positioned using privacy location model, and is exported
The privacy location information of target video;According to privacy location information, half destructiveness is carried out to the privacy content in target video and is gone
It removes, the target video that obtains that treated.
Further, the privacy content in target video is positioned using privacy location model, and exports target view
Before the privacy location information of frequency, this method further include: obtain the privacy locating rule of user's input;Based on privacy locating rule
Create privacy location model.
Further, the privacy content in target video is positioned using privacy location model, and exports target view
Before the privacy location information of frequency, this method further include: obtain multiple groups video scene, wherein include: hidden in every group of video scene
Private type and specified privacy information;It is trained to obtain initial model based on multiple groups video scene;Initial model is moved
Study is moved, privacy location model is obtained.
Further, the privacy content in target video is positioned using privacy location model, and exports target view
The privacy location information of frequency includes: the timing node for exporting privacy content in target video, wherein timing node includes starting
Time point and end time point;Export the coordinate information of privacy content video frame in target video, wherein the coordinate of video frame
Information includes the frame number and coordinate position in the video frame of video frame.
Further, according to privacy location information, half destructive removal is carried out to the privacy content in target video, is obtained
Treated, and target video includes;It is determined based on the coordinate information of video frame in the timing node and target video in target video
Privacy content in target video;Data Dimensionality Reduction processing is carried out to the privacy content in target video, the target that obtains that treated
Video;Or, carry out feature extraction to the privacy content in target video, the target video that obtains that treated.
Further, after obtaining target video to be processed, this method further include: target video is located in advance
Reason, obtains pretreated target video, wherein pretreatment includes at least following one: at resolution decreasing processing, color characteristic
Reason, noise reduction process.
To achieve the goals above, according to the another aspect of the application, a kind of video process apparatus is provided.The device packet
It includes: first acquisition unit, for obtaining target video to be processed;Positioning unit, for utilizing privacy location model to target
Privacy content in video is positioned, and exports the privacy location information of target video;Removal unit, for fixed according to privacy
Position information, to the privacy content in target video carry out half it is destructive remove, the target video that obtains that treated.
Further, the device further include: second acquisition unit, for utilizing privacy location model in target video
Privacy content is positioned, and before exporting the privacy location information of target video, obtains the privacy locating rule of user's input;
Creating unit, for creating privacy location model based on privacy locating rule.
To achieve the goals above, according to the another aspect of the application, a kind of storage medium is provided, storage medium includes
The program of storage, wherein the program executes a kind of method for processing video frequency of above-mentioned any one.
To achieve the goals above, according to the another aspect of the application, a kind of processor is provided, processor is for running
Program, wherein the program executes a kind of method for processing video frequency of above-mentioned any one when running.
By the application, using following steps: obtaining target video to be processed;Target is regarded using privacy location model
Privacy content in frequency is positioned, and exports the privacy location information of target video;According to privacy location information, target is regarded
Privacy content in frequency carries out half destructive removal, the target video that obtains that treated, solves and removes video in the related technology
In privacy content when, the low problem for the treatment of effeciency, by using the privacy location model being pre-created in target video
Privacy content is positioned, and is avoided and is positioned using manual type to the privacy content in video, to improve processing
Efficiency, while by being removed to the privacy content in video, the leakage to privacy content in the video is avoided, and then reach
Protect the effect of privacy content in video.
Detailed description of the invention
The attached drawing constituted part of this application is used to provide further understanding of the present application, the schematic reality of the application
Example and its explanation are applied for explaining the application, is not constituted an undue limitation on the present application.In the accompanying drawings:
Fig. 1 is according to a kind of flow chart of method for processing video frequency provided by the embodiments of the present application;
Fig. 2 is according to a kind of rule-based privacy location model schematic diagram provided by the embodiments of the present application;
Fig. 3 is according to a kind of privacy location model schematic diagram based on transfer learning provided by the embodiments of the present application;
Fig. 4 is the schematic diagram of the method for processing video frequency provided according to the application another kind embodiment;And
Fig. 5 is according to a kind of schematic diagram of video process apparatus provided by the embodiments of the present application.
Specific embodiment
It should be noted that in the absence of conflict, the features in the embodiments and the embodiments of the present application can phase
Mutually combination.The application is described in detail below with reference to the accompanying drawings and in conjunction with the embodiments.
In order to make those skilled in the art more fully understand application scheme, below in conjunction in the embodiment of the present application
Attached drawing, the technical scheme in the embodiment of the application is clearly and completely described, it is clear that described embodiment is only
The embodiment of the application a part, instead of all the embodiments.Based on the embodiment in the application, ordinary skill people
Member's every other embodiment obtained without making creative work, all should belong to the model of the application protection
It encloses.
It should be noted that the description and claims of this application and term " first " in above-mentioned attached drawing, "
Two " etc. be to be used to distinguish similar objects, without being used to describe a particular order or precedence order.It should be understood that using in this way
Data be interchangeable under appropriate circumstances, so as to embodiments herein described herein.In addition, term " includes " and " tool
Have " and their any deformation, it is intended that cover it is non-exclusive include, for example, containing a series of steps or units
Process, method, system, product or equipment those of are not necessarily limited to be clearly listed step or unit, but may include without clear
Other step or units listing to Chu or intrinsic for these process, methods, product or equipment.
For ease of description, below to the invention relates to part noun or term be illustrated:
Half destructive removal: carrying the partial video of privacy information, only destroy privacy information, be allowed to not restore, according to
So remain the information for machine recognition.
Privacy information identification, location technology: the part where privacy information in automatic identification video, for subsequent burst or
Half destroys.Rule-based identification (needs training data, which forms recurrence phenomenons, need using migration based on content recognition
Study).
Privacy information removal technology: partly devastatingly handling privacy information, makes privacy information that can not be resumed, interpret.
According to an embodiment of the present application, a kind of method for processing video frequency is provided.
Fig. 1 is the flow chart according to a kind of method for processing video frequency of the embodiment of the present application.As shown in Figure 1, this method includes
Following steps:
Step S101 obtains target video to be processed.
Specifically, to video carry out privacy content processing before, first have to obtain original video, using the original video as
Video to be processed, and save video to be processed.
Step S102 positions the privacy content in target video using privacy location model, and exports target view
The privacy location information of frequency.
Above-mentionedly, privacy location model is the model being pre-created, in the embodiment of the present application, fixed by using the privacy
Bit model positions the privacy content in target video, the efficiency of privacy content in positioning target video is promoted, by hidden
The positioning of private location model obtains the location information of privacy content.
Optionally, the privacy content in target video is positioned using privacy location model, and exports target video
Privacy location information before, this method further include: obtain user input privacy locating rule;It is created based on privacy locating rule
Build privacy location model.
Specifically, privacy location model is built upon the Information locating model in privacy location technology.It is adopted in the present embodiment
There are two types of privacy location models, and one of which is rule-based privacy location model, as shown in Fig. 2, fixed based on privacy
Position rule creation privacy location model, needs the information artificially by privacy information locating rule and privacy content in target video
Privacy location model is inputted, privacy location model finds privacy content according to the rule and information that artificially input in target video
The part at place.For example, location information of the privacy content in target video are as follows: time: 00:00:00 --- 15:00:00, depending on
Frequency frame coordinate: 0,0,1920,30, therefore, the content existing for above-mentioned period and video frame coordinate position is privacy content, people
For by time 00:00:00 --- 15:00:00 and coordinate: 0,0,1920,30 input privacy location model, privacy location model is just
Privacy content can be found in target video according to the privacy location information of input.
It should be noted that above-mentioned rule-based privacy location model is compared with suitable for relatively short video or mesh
Mark the less situation of privacy content present in video.
Optionally, the privacy content in target video is positioned using privacy location model, and exports target video
Privacy location information before, this method further include: obtain multiple groups video scene, wherein include: privacy in every group of video scene
Type and specified privacy information;It is trained to obtain initial model based on multiple groups video scene;Initial model is migrated
Study, obtains privacy location model.
Specifically, above-mentioned privacy location model is referred to as the privacy location model based on transfer learning, is suitable for target
The case where including more and more complex privacy content in video.As shown in figure 3, its creation process is as follows: artificially specifying privacy
The data (corresponding to above-mentioned multiple groups video scene) of the video scene of the similar privacy information of the type of information -> prepare in advance ->
Training obtains initial model, and initial model can identify the position of privacy information -> desensitized with a small amount of from video automatically
True privacy information video scene, to initial model transfer learning -> obtain final mask -> with the true privacy to have desensitized
Informational video scrnario testing, iteration.Such as: due to having the presence such as capsule information, article and desktop, artificially specify all desktops
It is all privacy information, firstly, preparing " desktop+data+article " video of non-genuine video scene in advance;Secondly, training obtains
Initial identification model;Again, with the video of the real video scene of " desktop+data+article " that has desensitized on a small quantity, know to initial
Other model carries out transfer learning;Finally, obtaining final " desktop+data+article " identification model, wherein final identification model
It also to be tested, iteration, can be only used for carrying out high accuracy positioning to the desktop position in target video.
Optionally, the privacy content in target video is positioned using privacy location model, and exports target video
Privacy location information include: to export timing node of the privacy content in target video, wherein when timing node includes starting
Between point and end time point;Export the coordinate information of privacy content video frame in target video, wherein the coordinate of video frame is believed
Frame number of the breath including video frame and coordinate position in the video frame.
Specifically, using above-mentioned rule-based privacy location model and the privacy location model based on transfer learning to mesh
Mark video is positioned, and the purpose of positioning is all the more specific location information for exporting privacy content in target video.Therefore, it positions
Afterwards, privacy location model can export specific location of the privacy content in target video, wherein the specific location packet of privacy content
Including the coordinate position in the timing node and video frame in target video, timing node includes initial time and end time,
Or the coordinate in video frame number and video frame.For example, position of the privacy content of output in target video is: the time: 00:
00:00 --- 00:00:20 and coordinate: 0,0,1920,30, i.e., in the beginning of target video to there are privacy contents in 20 seconds, and
And the video frame coordinate of specific privacy content is that 0,0,1920,30 or privacy content are present in 110- in target video
In the video frame of 550 frames, and privacy content is present in the coordinate in video frame are as follows: 0,0,1920,30.
Step S103 carries out half destructive removal to the privacy content in target video, obtains according to privacy location information
Treated target video.
Above-mentionedly, by being removed using half destructive removal means to privacy content, removal privacy content is obtained
Target video, and then while reached privacy content removal, do not influence the effect of subsequent privacy content identification also.
Optionally, according to privacy location information, half destructive removal is carried out to the privacy content in target video, is obtained everywhere
Target video after reason includes;Mesh is determined based on the coordinate information of video frame in the timing node and target video in target video
Mark the privacy content in video;Data Dimensionality Reduction processing is carried out to the privacy content in target video, the target view that obtains that treated
Frequently;Or, carry out feature extraction to the privacy content in target video, the target video that obtains that treated.
Specifically, according to privacy location model export privacy location information, obtain include privacy content piece of video
Section carries out the half destructive target video for removing, obtaining after half destructive removal, privacy content human eye therein to privacy content
Non-readable, the mankind can not interpret, but still remain the characteristic information that machine can be read, and then for example in subsequent criminal investigation
In investigation, it can use machine and privacy content identified.Wherein, using in the processing of privacy content Data Dimensionality Reduction and privacy
Two kinds of technological means of feature extraction of appearance, which can reach, not can recognize privacy content processing to human eye, and the mankind can not interpret,
But the effect that machine can identify, certainly, other reach the means of above-mentioned treatment effect also in the restriction of the application, different
One enumerates and repeats.
It should be noted that the privacy content after destroying removal cannot be recognized by the human eye, the mankind can not be interpreted,
It can not be reconditioned as former visual information.
For example, the clothes of Cleaner are privacy contents in monitor video, and after being destroyed by half destructive removal, monitor video
The clothes visual image of middle Cleaner is destroyed, and still, is present in the video letter of other such as desks, chair in video scene
Breath all also saves, and only the clothes of Cleaner are not regarded by human eye, and the mankind can not interpret, but machine can also be according to base
This characteristic information can be read out the clothes of Cleaner.
Optionally, after obtaining target video to be processed, this method further include: target video is pre-processed,
Obtain pretreated target video, wherein pretreatment includes at least following one: resolution decreasing is handled, color characteristic is handled,
Noise reduction process.
Specifically, target video can also be located in advance before obtaining target video and carrying out privacy information removal processing
Reason, wherein preprocessing means include resolution decreasing processing, color characteristic processing, noise reduction process etc., and its object is to reduce target
Video removes calculation rate when operating in privacy information.
Above-mentionedly, resolution decreasing processing is carried out to the target video of acquisition, since the purpose of the present embodiment is protection target
Privacy content in video, in the case where reducing certain video resolution does not influence the positioning to privacy content, target view
The clarity of frequency is not just the factor mainly considered, therefore, target video is carried out resolution decreasing processing, it is possible to reduce subsequent right
The operating quantity of target video removal privacy content.
Above-mentionedly, if the color in target video is not required to the privacy information of protection, target video is being obtained
After can carry out color characteristic processing to target video, such as color be removed, since reduce privacy in subsequent processing video
The calculation amount of information.
Above-mentionedly, if certain a segment of audio in video is to need privacy content to be protected, it is also necessary to in target video
Audio carry out noise reduction, to make to need audio-frequency information to be protected clearer, Yi Dingwei.For example, if being deposited in target video
Conversation content be to need privacy content to be protected, carrying out noise reduction to the audio in target video can make in target video
Conversation content is relatively sharp, when carrying out positioning searching to privacy audio content, is easier to navigate in the audio in target video
Hold.
A kind of method for processing video frequency provided by the embodiments of the present application, by obtaining target video to be processed;Utilize privacy
Location model positions the privacy content in target video, and exports the privacy location information of target video;According to privacy
Location information, to the privacy content in target video carry out half it is destructive remove, the target video that obtains that treated solves phase
When removing the privacy content in video in the technology of pass, the low problem for the treatment of effeciency, by utilizing half destructive removal means to hidden
Private content is removed, and then while reached privacy content removal, does not influence the effect of subsequent privacy content identification also, together
When, the privacy content in target video is positioned by using the privacy location model being pre-created, is avoided using people
Work mode positions the privacy content in video, so that treatment effeciency is improved, by going to the privacy content in video
It removes, avoids the leakage to privacy content in the video, and then achieved the effect that protect privacy content in video.
The method for processing video frequency that the application another kind embodiment provides, as shown in figure 4, being handled using privacy location model former
Beginning video obtains privacy location information;According to privacy location information, half destructive removal is carried out to privacy content, is removed
The video of privacy information.
Above-mentionedly, it by being positioned using privacy location model to the privacy content in video, avoids using artificial
Mode positions the privacy content in video, so that treatment effeciency is improved, by being removed to the privacy content in video,
The leakage to privacy content in the video is avoided, when solving the privacy content removed in video in the related technology, processing effect
The low problem of rate.Meanwhile half destructive removal is carried out further through the privacy to video, while reaching privacy content removal,
The effect of subsequent privacy content identification is not influenced also.
It should be noted that step shown in the flowchart of the accompanying drawings can be in such as a group of computer-executable instructions
It is executed in computer system, although also, logical order is shown in flow charts, and it in some cases, can be with not
The sequence being same as herein executes shown or described step.
The embodiment of the present application also provides a kind of video process apparatus, it should be noted that one kind of the embodiment of the present application
Video process apparatus can be used for executing provided by the embodiment of the present application for a kind of method for processing video frequency.Below to the application
A kind of video process apparatus that embodiment provides is introduced.
Fig. 5 is the schematic diagram according to a kind of video process apparatus of the embodiment of the present application.As shown in figure 5, the device includes:
First acquisition unit 501, for obtaining target video to be processed;Positioning unit 502, for utilizing privacy location model to mesh
Privacy content in mark video is positioned, and exports the privacy location information of target video;Removal unit 503 is used for basis
Privacy location information, to the privacy content in target video carry out half it is destructive remove, the target video that obtains that treated.
A kind of video process apparatus provided by the embodiments of the present application, mesh to be processed is obtained by first acquisition unit 501
Mark video;Positioning unit 502 positions the privacy content in target video using privacy location model, and exports target view
The privacy location information of frequency;Removal unit 503 carries out half to the privacy content in target video and destroys according to privacy location information
Property removal, the target video that obtains that treated, when solving the privacy content removed in video in the related technology, treatment effeciency is low
The problem of, the privacy content in target video is positioned by using the privacy location model being pre-created, avoids and adopts
Manually mode positions the privacy content in video, so that treatment effeciency is improved, by the privacy in video
Hold removal, avoid the leakage to privacy content in the video, and then achieved the effect that protect privacy content in video, together
When, while removal by being removed using half destructive removal means to privacy content, and then having reached privacy content, also
The effect of subsequent privacy content identification is not influenced.
Optionally, the device further include: second acquisition unit, for utilizing privacy location model to hidden in target video
Private content is positioned, and before exporting the privacy location information of target video, obtains the privacy locating rule of user's input;Wound
Unit is built, for creating privacy location model based on privacy locating rule.
Optionally, the device further include: third acquiring unit, for utilizing privacy location model to hidden in target video
Private content is positioned, and before exporting the privacy location information of target video, obtains multiple groups corpus video scene, wherein every
It include: privacy type and specified privacy information in group corpus video scene;Training unit, for based on multiple groups video scene into
Row training obtains initial model;Unit obtains privacy location model for carrying out transfer learning to initial model.
Optionally, positioning unit 502 includes: the first output subelement, for exporting privacy content in target video
Timing node, wherein timing node includes start time point and end time point;Second output subelement, for exporting privacy
The coordinate information of content video frame in target video, wherein the coordinate information of video frame includes the frame number of video frame and regarding
Coordinate position in frequency frame.
Optionally, removal unit 503 includes;Subelement is determined, for based on the timing node and target in target video
The coordinate information of video frame determines the privacy content in target video in video;Subelement is handled, for in target video
Privacy content carries out Data Dimensionality Reduction processing, the target video that obtains that treated;Or, subelement is extracted, for in target video
Privacy content carry out feature extraction, the target video that obtains that treated.
Optionally, which further includes processing unit, for being regarded after obtaining target video to be processed to target
Frequency is pre-processed, and pretreated target video is obtained, wherein pretreatment is including at least following one: resolution decreasing processing,
Color characteristic processing, noise reduction process.
A kind of video process apparatus includes processor and memory, above-mentioned first acquisition unit 501,502 and of positioning unit
It is stored in memory as program unit except unit 503 is equal, above procedure stored in memory is executed by processor
Unit realizes corresponding function.
Include kernel in processor, is gone in memory to transfer corresponding program unit by kernel.Kernel can be set one
Or more, it is asked by adjusting kernel parameter come when solving to remove the privacy content in video in the related technology, treatment effeciency is low
Topic.
Memory may include the non-volatile memory in computer-readable medium, random access memory (RAM) and/
Or the forms such as Nonvolatile memory, if read-only memory (ROM) or flash memory (flash RAM), memory include that at least one is deposited
Store up chip.
The embodiment of the invention provides a kind of storage mediums, are stored thereon with program, real when which is executed by processor
A kind of existing method for processing video frequency.
The embodiment of the invention provides a kind of processor, processor is for running program, wherein program executes one when running
Kind method for processing video frequency.
The embodiment of the invention provides a kind of equipment, equipment include processor, memory and storage on a memory and can
The program run on a processor, processor performs the steps of when executing program obtains target video to be processed;Using hidden
Private location model positions the privacy content in target video, and exports the privacy location information of target video;According to hidden
Private location information, to the privacy content in target video carry out half it is destructive remove, the target video that obtains that treated.
Optionally, the privacy content in target video is positioned using privacy location model, and exports target video
Privacy location information before, this method further include: obtain user input privacy locating rule;It is created based on privacy locating rule
Build privacy location model.
Optionally, the privacy content in target video is positioned using privacy location model, and exports target video
Privacy location information before, this method further include: obtain multiple groups video scene, wherein include: privacy in every group of video scene
Type and specified privacy information;It carries out obtaining initial model based on multiple groups video scene;Transfer learning is carried out to initial model,
Obtain privacy location model.
Optionally, the privacy content in target video is positioned using privacy location model, and exports target video
Privacy location information include: to export timing node of the privacy content in target video, wherein when timing node includes starting
Between point and end time point;Export the coordinate information of privacy content video frame in target video, wherein the coordinate of video frame is believed
Frame number of the breath including video frame and coordinate position in the video frame.
Optionally, according to privacy location information, half destructive removal is carried out to the privacy content in target video, is obtained everywhere
Target video after reason includes;Mesh is determined based on the coordinate information of video frame in the timing node and target video in target video
Mark the privacy content in video;Data Dimensionality Reduction processing is carried out to the privacy content in target video, the target view that obtains that treated
Frequently;Or, carry out feature extraction to the privacy content in target video, the target video that obtains that treated.
Optionally, after obtaining target video to be processed, this method further include: target video is pre-processed,
Obtain pretreated target video, wherein pretreatment includes at least following one: resolution decreasing is handled, color characteristic is handled,
Noise reduction process.Equipment herein can be server, PC, PAD, mobile phone etc..
Present invention also provides a kind of computer program products, when executing on data processing equipment, are adapted for carrying out just
The program of beginningization there are as below methods step: target video to be processed is obtained;Using privacy location model in target video
Privacy content is positioned, and exports the privacy location information of target video;According to privacy location information, in target video
Privacy content carries out half destructive removal, the target video that obtains that treated.
Optionally, the privacy content in target video is positioned using privacy location model, and exports target video
Privacy location information before, this method further include: obtain user input privacy locating rule;It is created based on privacy locating rule
Build privacy location model.
Optionally, the privacy content in target video is positioned using privacy location model, and exports target video
Privacy location information before, this method further include: obtain multiple groups video scene, wherein include: privacy in every group of video scene
Type and specified privacy information;It carries out obtaining initial model based on multiple groups video scene;Transfer learning is carried out to initial model,
Obtain privacy location model.
Optionally, the privacy content in target video is positioned using privacy location model, and exports target video
Privacy location information include: to export timing node of the privacy content in target video, wherein when timing node includes starting
Between point and end time point;Export the coordinate information of privacy content video frame in target video, wherein the coordinate of video frame is believed
Frame number of the breath including video frame and coordinate position in the video frame.
Optionally, according to privacy location information, half destructive removal is carried out to the privacy content in target video, is obtained everywhere
Target video after reason includes;Mesh is determined based on the coordinate information of video frame in the timing node and target video in target video
Mark the privacy content in video;Data Dimensionality Reduction processing is carried out to the privacy content in target video, the target view that obtains that treated
Frequently;Or, carry out feature extraction to the privacy content in target video, the target video that obtains that treated.
Optionally, after obtaining target video to be processed, this method further include: target video is pre-processed,
Obtain pretreated target video, wherein pretreatment includes at least following one: resolution decreasing is handled, color characteristic is handled,
Noise reduction process.
It should be understood by those skilled in the art that, embodiments herein can provide as method, system or computer program
Product.Therefore, complete hardware embodiment, complete software embodiment or reality combining software and hardware aspects can be used in the application
Apply the form of example.Moreover, it wherein includes the computer of computer usable program code that the application, which can be used in one or more,
The computer program implemented in usable storage medium (including but not limited to magnetic disk storage, CD-ROM, optical memory etc.) produces
The form of product.
The application is referring to method, the process of equipment (system) and computer program product according to the embodiment of the present application
Figure and/or block diagram describe.It should be understood that every one stream in flowchart and/or the block diagram can be realized by computer program instructions
The combination of process and/or box in journey and/or box and flowchart and/or the block diagram.It can provide these computer programs
Instruct the processor of general purpose computer, special purpose computer, Embedded Processor or other programmable data processing devices to produce
A raw machine, so that being generated by the instruction that computer or the processor of other programmable data processing devices execute for real
The device for the function of being specified in present one or more flows of the flowchart and/or one or more blocks of the block diagram.
These computer program instructions, which may also be stored in, is able to guide computer or other programmable data processing devices with spy
Determine in the computer-readable memory that mode works, so that it includes referring to that instruction stored in the computer readable memory, which generates,
Enable the manufacture of device, the command device realize in one box of one or more flows of the flowchart and/or block diagram or
The function of being specified in multiple boxes.
These computer program instructions also can be loaded onto a computer or other programmable data processing device, so that counting
Series of operation steps are executed on calculation machine or other programmable devices to generate computer implemented processing, thus in computer or
The instruction executed on other programmable devices is provided for realizing in one or more flows of the flowchart and/or block diagram one
The step of function of being specified in a box or multiple boxes.
In a typical configuration, calculating equipment includes one or more processors (CPU), input/output interface, net
Network interface and memory.
Memory may include the non-volatile memory in computer-readable medium, random access memory (RAM) and/
Or the forms such as Nonvolatile memory, such as read-only memory (ROM) or flash memory (flash RAM).Memory is computer-readable Jie
The example of matter.
Computer-readable medium includes permanent and non-permanent, removable and non-removable media can be by any method
Or technology come realize information store.Information can be computer readable instructions, data structure, the module of program or other data.
The example of the storage medium of computer includes, but are not limited to phase change memory (PRAM), static random access memory (SRAM), moves
State random access memory (DRAM), other kinds of random access memory (RAM), read-only memory (ROM), electric erasable
Programmable read only memory (EEPROM), flash memory or other memory techniques, read-only disc read only memory (CD-ROM) (CD-ROM),
Digital versatile disc (DVD) or other optical storage, magnetic cassettes, tape magnetic disk storage or other magnetic storage devices
Or any other non-transmission medium, can be used for storage can be accessed by a computing device information.As defined in this article, it calculates
Machine readable medium does not include temporary computer readable media (transitory media), such as the data-signal and carrier wave of modulation.
It should also be noted that, the terms "include", "comprise" or its any other variant are intended to nonexcludability
It include so that the process, method, commodity or the equipment that include a series of elements not only include those elements, but also to wrap
Include other elements that are not explicitly listed, or further include for this process, method, commodity or equipment intrinsic want
Element.In the absence of more restrictions, the element limited by sentence "including a ...", it is not excluded that including element
There is also other identical elements in process, method, commodity or equipment.
It will be understood by those skilled in the art that embodiments herein can provide as method, system or computer program product.
Therefore, complete hardware embodiment, complete software embodiment or embodiment combining software and hardware aspects can be used in the application
Form.It is deposited moreover, the application can be used to can be used in the computer that one or more wherein includes computer usable program code
The shape for the computer program product implemented on storage media (including but not limited to magnetic disk storage, CD-ROM, optical memory etc.)
Formula.
The above is only embodiments herein, are not intended to limit this application.To those skilled in the art,
Various changes and changes are possible in this application.It is all within the spirit and principles of the present application made by any modification, equivalent replacement,
Improve etc., it should be included within the scope of the claims of this application.
Claims (10)
1. a kind of method for processing video frequency characterized by comprising
Obtain target video to be processed;
The privacy content in the target video is positioned using privacy location model, and exports the hidden of the target video
Private location information;
According to the privacy location information, half destructive removal is carried out to the privacy content in the target video, is handled
Target video afterwards.
2. the method according to claim 1, wherein using privacy location model to hidden in the target video
Private content is positioned, and before exporting the privacy location information of the target video, the method also includes:
Obtain the privacy locating rule of user's input;
The privacy location model is created based on the privacy locating rule.
3. the method according to claim 1, wherein using privacy location model to hidden in the target video
Private content is positioned, and before exporting the privacy location information of the target video, the method also includes:
Obtain multiple groups video scene, wherein include: privacy type and specified privacy information in every group of video scene;
It is trained to obtain initial model based on the multiple groups video scene;
Transfer learning is carried out to the initial model, obtains the privacy location model.
4. the method according to claim 1, wherein using privacy location model to hidden in the target video
Private content is positioned, and the privacy location information for exporting the target video includes:
Export timing node of the privacy content in the target video, wherein the timing node includes initial time
Point and end time point;
Export the coordinate information of privacy content video frame in the target video, wherein the coordinate of the video frame is believed
Frame number of the breath including video frame and coordinate position in the video frame.
5. according to the method described in claim 4, it is characterized in that, according to the privacy location information, to the target video
In privacy content carry out half destructive removal, obtaining that treated, target video includes;
The target is determined based on the coordinate information of video frame in the timing node and the target video in the target video
Privacy content in video;
Data Dimensionality Reduction processing is carried out to the privacy content in the target video, the target video that obtains that treated;
Or,
Feature extraction is carried out to the privacy content in the target video, the target video that obtains that treated.
6. the method according to claim 1, wherein after obtaining target video to be processed, the method
Further include: the target video is pre-processed, pretreated target video is obtained, wherein the pretreatment is at least wrapped
Include following one: resolution decreasing processing, color characteristic processing, noise reduction process.
7. a kind of video process apparatus characterized by comprising
First acquisition unit, for obtaining target video to be processed;
Positioning unit for positioning using privacy location model to the privacy content in the target video, and exports institute
State the privacy location information of target video;
Removal unit, for it is destructive to carry out half to the privacy content in the target video according to the privacy location information
Removal, the target video that obtains that treated.
8. device according to claim 7, which is characterized in that described device further include:
Second acquisition unit, for being positioned using privacy location model to the privacy content in the target video, and it is defeated
Out before the privacy location information of the target video, the privacy locating rule of user's input is obtained;
Creating unit, for creating the privacy location model based on the privacy locating rule.
9. a kind of storage medium, which is characterized in that the storage medium includes the program of storage, wherein described program right of execution
Benefit require any one of 1 to 6 described in a kind of method for processing video frequency.
10. a kind of processor, which is characterized in that the processor is for running program, wherein right of execution when described program is run
Benefit require any one of 1 to 6 described in a kind of method for processing video frequency.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811585170.5A CN109743580A (en) | 2018-12-24 | 2018-12-24 | A kind of method for processing video frequency and device, storage medium and processor |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811585170.5A CN109743580A (en) | 2018-12-24 | 2018-12-24 | A kind of method for processing video frequency and device, storage medium and processor |
Publications (1)
Publication Number | Publication Date |
---|---|
CN109743580A true CN109743580A (en) | 2019-05-10 |
Family
ID=66359677
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201811585170.5A Pending CN109743580A (en) | 2018-12-24 | 2018-12-24 | A kind of method for processing video frequency and device, storage medium and processor |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109743580A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111428273A (en) * | 2020-04-23 | 2020-07-17 | 北京中安星云软件技术有限公司 | Dynamic desensitization method and device based on machine learning |
Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2009061420A1 (en) * | 2007-11-07 | 2009-05-14 | Viewdle, Inc. | Object recognition and database population |
CN104616026A (en) * | 2015-01-20 | 2015-05-13 | 衢州学院 | Monitor scene type identification method for intelligent video monitor |
CN105007395A (en) * | 2015-07-22 | 2015-10-28 | 深圳市万姓宗祠网络科技股份有限公司 | Privacy processing method for continuously recording video |
CN106295584A (en) * | 2016-08-16 | 2017-01-04 | 深圳云天励飞技术有限公司 | Depth migration study is in the recognition methods of crowd's attribute |
CN107704877A (en) * | 2017-10-09 | 2018-02-16 | 哈尔滨工业大学深圳研究生院 | A kind of image privacy cognitive method based on deep learning |
CN108040230A (en) * | 2017-12-19 | 2018-05-15 | 司马大大(北京)智能***有限公司 | A kind of monitoring method and device for protecting privacy |
CN108520184A (en) * | 2018-04-16 | 2018-09-11 | 成都博锐智晟科技有限公司 | A kind of method and system of secret protection |
CN108900895A (en) * | 2018-08-23 | 2018-11-27 | 深圳码隆科技有限公司 | The screen method and its device of the target area of a kind of pair of video flowing |
-
2018
- 2018-12-24 CN CN201811585170.5A patent/CN109743580A/en active Pending
Patent Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2009061420A1 (en) * | 2007-11-07 | 2009-05-14 | Viewdle, Inc. | Object recognition and database population |
CN104616026A (en) * | 2015-01-20 | 2015-05-13 | 衢州学院 | Monitor scene type identification method for intelligent video monitor |
CN105007395A (en) * | 2015-07-22 | 2015-10-28 | 深圳市万姓宗祠网络科技股份有限公司 | Privacy processing method for continuously recording video |
CN106295584A (en) * | 2016-08-16 | 2017-01-04 | 深圳云天励飞技术有限公司 | Depth migration study is in the recognition methods of crowd's attribute |
CN107704877A (en) * | 2017-10-09 | 2018-02-16 | 哈尔滨工业大学深圳研究生院 | A kind of image privacy cognitive method based on deep learning |
CN108040230A (en) * | 2017-12-19 | 2018-05-15 | 司马大大(北京)智能***有限公司 | A kind of monitoring method and device for protecting privacy |
CN108520184A (en) * | 2018-04-16 | 2018-09-11 | 成都博锐智晟科技有限公司 | A kind of method and system of secret protection |
CN108900895A (en) * | 2018-08-23 | 2018-11-27 | 深圳码隆科技有限公司 | The screen method and its device of the target area of a kind of pair of video flowing |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111428273A (en) * | 2020-04-23 | 2020-07-17 | 北京中安星云软件技术有限公司 | Dynamic desensitization method and device based on machine learning |
CN111428273B (en) * | 2020-04-23 | 2023-08-25 | 北京中安星云软件技术有限公司 | Dynamic desensitization method and device based on machine learning |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109743579A (en) | A kind of method for processing video frequency and device, storage medium and processor | |
CN107610695A (en) | Driver's voice wakes up the dynamic adjusting method of instruction word weight | |
CN109542789B (en) | Code coverage rate statistical method and device | |
CN103347009B (en) | A kind of information filtering method and device | |
CN104572056B (en) | A kind of method and device of page comparison | |
CN104572431B (en) | A kind of method of testing and device | |
CN110069781B (en) | Entity label identification method and related equipment | |
CN106933887A (en) | A kind of data visualization method and device | |
CN109766683B (en) | Protection method for sensor fingerprint of mobile intelligent device | |
CN111932438B (en) | Image style migration method, device and storage device | |
CN110532983A (en) | Method for processing video frequency, device, medium and equipment | |
CN105657446B (en) | The detection method and device of bumper advertisements in a kind of video | |
CN106909567B (en) | Data processing method and device | |
CN106843820A (en) | Code process method and apparatus | |
CN109683773A (en) | Corpus labeling method and device | |
CN109743580A (en) | A kind of method for processing video frequency and device, storage medium and processor | |
CN108733694A (en) | Method and apparatus are recommended in retrieval | |
CN109766123A (en) | Application program packaging method and device | |
CN108818531A (en) | Robot control method and device | |
CN109561339A (en) | The treating method and apparatus of video file | |
CN107562703A (en) | Dictionary tree reconstructing method and system | |
CN114897723B (en) | Image generation and noise adding method based on generation type countermeasure network | |
CN110019508B (en) | Data synchronization method, device and system | |
CN108806701A (en) | Voice data method for scrambling and device, voice data de-scrambling method and device | |
CN113660658A (en) | Audio data protection method and device and voice equipment |
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 | ||
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190510 |