CN108182444A - The method and device of video quality diagnosis based on scene classification - Google Patents
The method and device of video quality diagnosis based on scene classification Download PDFInfo
- Publication number
- CN108182444A CN108182444A CN201711298596.8A CN201711298596A CN108182444A CN 108182444 A CN108182444 A CN 108182444A CN 201711298596 A CN201711298596 A CN 201711298596A CN 108182444 A CN108182444 A CN 108182444A
- Authority
- CN
- China
- Prior art keywords
- scene
- image data
- checkup item
- video quality
- confidence level
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N17/00—Diagnosis, testing or measuring for television systems or their details
Abstract
The present invention is suitable for video quality diagnostic techniques field, provides a kind of method and device of the video quality diagnosis based on scene classification, the method includes:Obtain realtime image data;Classified using trained scene classifier to the scene of realtime image data, obtain scene type corresponding with realtime image data and classification confidence level;If classification confidence level is greater than or equal to predetermined threshold value, according to the correspondence of scene type, checkup item and checkup item sensitivity three, the checkup item corresponding with scene type and checkup item sensitivity are obtained;The checkup item of adjustment video quality diagnosis system reaches the checkup item sensitivity, so as to carry out the diagnosis of the video quality of the realtime image data;The checkup item sensitivity the invention enables the video quality diagnosis system can be automatically adjusted according to the scene type and classification confidence level of realtime image data, is solved the problems, such as the wrong report of video quality diagnosis and is failed to report.
Description
Technical field
The invention belongs to video quality diagnostic techniques field more particularly to a kind of video quality diagnosis based on scene classification
Method and device.
Background technology
Video quality diagnosis in, can usually carry out the diagnosis of some checkup items, for example, to video colour cast, clarity,
Brightness exception blank screen, shakes, freezes, the diagnosis of the checkup items such as noise, when carrying out the diagnosis of these checkup items, can be related to
To warning sensitivity.The warning sensitivity refers to judge colour cast occur in video image, clarity, brightness exception, blank screen, tremble
It moves, freeze, the critical indicator of the abnormal conditions such as noise.When the warning sensitivity is higher, easier triggering alarm, then to described
Occur in video image colour cast, clarity, brightness exception, blank screen, shake, freeze, the misinformation probability of the abnormal conditions such as noise increases
Greatly;Conversely, when the warning sensitivity is lower, triggering alarm is less susceptible to, then to occurring colour cast, clear in the video image
Degree, brightness exception, blank screen, shakes, freeze, the miss probability of the abnormal conditions such as noise increases.
In the prior art, although by the setting interface opening of the warning sensitivity to user, while to each diagnosis item
Mesh setting acquiescence warning sensitivity, still, identical warning sensitivity can not be suitable for all site environments, therefore, still without
Method solves the problems, such as to report by mistake and fail to report in video quality diagnosis.
Invention content
In view of this, an embodiment of the present invention provides a kind of methods and dress of the video quality diagnosis based on scene classification
It puts, to solve the problems, such as not solving the wrong report of video quality diagnosis in the prior art and fail to report.
The first aspect of the embodiment of the present invention provides a kind of method of the video quality diagnosis based on scene classification, packet
It includes:
Obtain realtime image data;
Classified using trained scene classifier to the scene of the realtime image data, obtained and described real-time
The corresponding scene type of image data and classification confidence level;
If the classification confidence level is greater than or equal to predetermined threshold value, according to the scene type, checkup item and diagnosis
The correspondence of project sensitivity three, obtains the checkup item corresponding with the scene type and checkup item is sensitive
Degree;
The checkup item of adjustment video quality diagnosis system reaches the checkup item sensitivity, described so as to carry out
The video quality diagnosis of realtime image data.
The second aspect of the embodiment of the present invention provides a kind of device of the video quality diagnosis based on scene classification, packet
It includes:
Image acquisition unit, for obtaining realtime image data;
Taxon, for being classified using scene classifier to the scene of the realtime image data, acquisition and institute
State the corresponding scene type of realtime image data and classification confidence level;
Sensitivity acquiring unit, if being greater than or equal to predetermined threshold value for the classification confidence level, according to the scene
The correspondence of classification, checkup item and checkup item sensitivity three obtains the diagnosis corresponding with the scene type
Project and checkup item sensitivity;
Adjustment unit, it is sensitive that the checkup item for adjusting video quality diagnosis system reaches the checkup item
Degree, so as to carry out the diagnosis of the video quality of the realtime image data.
The third aspect of the embodiment of the present invention provides a kind of device of the video quality diagnosis based on scene classification, packet
It includes:Including memory, processor and the computer program that can be run in the memory and on the processor is stored in,
The step of processor realizes the above method when performing the computer program.
The fourth aspect of the embodiment of the present invention provides a kind of computer readable storage medium, the computer-readable storage
Media storage has computer program, when the computer program is executed by processor the step of the realization above method
In the embodiment of the present invention, the scene of the realtime image data is carried out by using trained scene classifier
Classification obtains corresponding with realtime image data scene type and classification confidence level and corresponding with the scene type
The checkup item and checkup item sensitivity;It is greater than or equal to predetermined threshold value in the classification confidence level, adjusts video matter
Amount diagnostic system the checkup item reach the checkup item sensitivity so that the video quality diagnosis system it is described
Checkup item sensitivity can be automatically adjusted according to the scene type and classification confidence level of realtime image data, solved and regarded
The wrong report of frequency quality diagnosis and fail to report problem.
Description of the drawings
It to describe the technical solutions in the embodiments of the present invention more clearly, below will be to embodiment or description of the prior art
Needed in attached drawing be briefly described, it should be apparent that, the accompanying drawings in the following description be only the present invention some
Embodiment, for those of ordinary skill in the art, without having to pay creative labor, can also be according to these
Attached drawing obtains other attached drawings.
Fig. 1 is that a kind of method of video quality diagnosis based on scene classification provided in an embodiment of the present invention realizes flow
Figure;
Fig. 2 is the realization flow diagram of Training scene grader provided in an embodiment of the present invention;
Fig. 3 is another realization flow diagram of Training scene grader provided in an embodiment of the present invention;
Fig. 4 is the realization of the method S102 of video quality diagnosis based on scene classification provided in an embodiment of the present invention a kind of
Flow diagram;
Fig. 5 is a kind of structural representation of the device of video quality diagnosis based on scene classification provided in an embodiment of the present invention
Figure;
Fig. 6 is that a kind of another structure of device of video quality diagnosis based on scene classification provided in an embodiment of the present invention is shown
It is intended to.
Specific embodiment
In being described below, in order to illustrate rather than in order to limit, it is proposed that such as tool of particular system structure, technology etc
Body details, to understand thoroughly the embodiment of the present invention.However, it will be clear to one skilled in the art that there is no these specifically
The present invention can also be realized in the other embodiments of details.In other situations, it omits to well-known system, device, electricity
Road and the detailed description of method, in case unnecessary details interferes description of the invention.
In the embodiment of the present invention, the scene of the realtime image data is carried out by using trained scene classifier
Classification obtains corresponding with realtime image data scene type and classification confidence level and corresponding with the scene type
The checkup item and checkup item sensitivity;It is greater than or equal to predetermined threshold value in the classification confidence level, adjusts video matter
Amount diagnostic system the checkup item reach the checkup item sensitivity so that the video quality diagnosis system it is described
Checkup item sensitivity can be automatically adjusted according to the scene type and classification confidence level of realtime image data, solved and regarded
The wrong report of frequency quality diagnosis and fail to report problem.
In order to illustrate technical solutions according to the invention, illustrated below by specific embodiment.
As Fig. 1 shows that the method that a kind of video quality based on scene classification provided in an embodiment of the present invention diagnoses is realized
Flow chart, this method include:S101 to S104.
In S101, realtime image data is obtained.
Video quality diagnosis is widely used in a variety of application fields, for example, the applications such as navigation picture acquisition, disaster inspection
Scene is required to diagnose video quality, so as to improve the precision of target detection, reduces wrong report or the probability failed to report.
Wherein, the premise that realtime image data is video quality diagnosis is obtained, in some embodiments of the present invention mode,
Real time image data acquisition can be carried out by camera, then the realtime image data of acquisition is inputted into trained scene
In grader, scene classification is carried out to it.
Optionally, before the acquisition realtime image data, it is also necessary to Training scene grader.
In an embodiment of the present invention, the scene classifier can be support vector cassification model or be other
Disaggregated model.The present invention is not especially limited the concrete structure of disaggregated model, suitable for the model of classification under the prior art
For the present invention.
As shown in Fig. 2, the Training scene grader includes:S201 to S203.
In S201, multiple sample image datas and sample scene corresponding with each sample image data are obtained
Classification and sample class confidence level.
Specifically, the sample image data includes the image under various scenes, for example, temple, lawn, snow scenes, people's row
The image data of the scenes such as road, car lane, street, house, field, forest, market, school, office building, will be described in acquisition
Sample image data carry out category label, that is to say, that each sample image data is carried out the label of scene type with
And the label of classification confidence level so that each sample image data has corresponding scene type and classification confidence level.
Wherein, the scene type includes temple, lawn, snow scenes, pavement, car lane, street, house, field, gloomy
The scene types such as woods, market, school, office building, it will be appreciated by those skilled in the art that the scene type can include tens
Kind even hundreds of kinds of classifications, can be determined according to sample image data, be merely illustrative herein, not indicate that as to this hair
The limitation of bright protection domain.
The classification confidence level refers to that the sample image data belongs to the probability value of a certain scene type.
For example, the sample image data is it is possible that 100 percent cannot belong to a certain scene type, therefore, to described
When sample image data divides scene type, the corresponding classification confidence level of the scene type is also added.For another example, certain piece image
In not only exist and belong to the probability of the first scene type, but also there is the probability for belonging to the second scene type or even also exist and belong to the
The probability of three scene types etc. scene type, and the sum of probability is 1, then directly cannot carry out scene point to the image at this time
Class.When the image 100 percent belongs to a certain scene type, then the corresponding classification confidence level of described image is 1.
That is, each sample image data can only there are one corresponding scene type and classification confidences
Degree, it is possibility to have multiple corresponding scene types and classification confidence level, still, the sum of described classification confidence level are 1.
In S202, the image feature information descriptor of the sample image data is extracted.
After the good scene type of the sample image data of label and classification confidence level, you can extract the sample image number
According to image feature information descriptor;Utilize the sample scene type, sample class confidence level and described image characteristic information
Descriptor is trained scene classifier, and obtains the trained scene classifier.
Wherein, described image characteristic information descriptor is including the use of Scale invariant features transform (Scale-invariant
Feature transform, SIFT) obtained image SIFT feature value and utilize Linear back projection algorithm (Local
Binary Patterns, LBP) obtained LBP operators, color histogram and gradient information.
SIFT feature value is used to detecting and describing the locality characteristic in image, is included in space scale and finds extreme value
Point, and extract position, scale and rotational invariants.
LBP operators are effective texture description operators, for measuring and extracting the texture information of image local, are had to illumination
There is invariance.
In S203, using the sample scene type, sample class confidence level and described image characteristic information descriptor,
Scene classifier is trained, obtains the trained scene classifier.
Optionally, as shown in figure 3, after step S201, S2011 is further included to S2012.
In S2011, the sample image data with the identical sample scene type is regarded as described
One group of diagnosis object of frequency quality diagnosis system, diagnoses object for described in every group, adjusts the video quality diagnosis system
Checkup item sensitivity corresponding with checkup item obtains diagnosis index corresponding with the checkup item sensitivity.
For example, all sample image datas for belonging to " temple " scene type are examined as the video quality
One group of diagnosis object of disconnected system, the checkup item corresponding with checkup item for adjusting the video quality diagnosis system are sensitive
Degree, obtains diagnosis index corresponding with the checkup item sensitivity.
Wherein, the checkup item include colour cast, clarity, brightness exception, blank screen, shakes, freeze, noise etc. diagnoses item
Mesh.The diagnosis index includes rate of false alarm and rate of failing to report.The rate of false alarm refer to video image do not occur colour cast in checkup item,
Clarity, brightness exception, blank screen, shake, freeze, noise is when abnormal conditions, the probability alarmed;The rate of failing to report refers to
Video image occur colour cast in checkup item, clarity, brightness exception, blank screen, shake, freeze, noise is when abnormal conditions, not
The probability alarmed.
In S2012, object is diagnosed for described in every group, by checkup item sensitivity corresponding with best diagnosis index
As checkup item sensitivity corresponding with the checkup item of the sample scene type, so as to establish the scene type, examine
Disconnected project and the correspondence of checkup item sensitivity three.
It is described obtain multiple sample image datas and sample scene type corresponding with each sample image data and
After sample class confidence level, for object is diagnosed described in every group, repeatedly adjust the video quality diagnosis system with diagnosis
The corresponding checkup item sensitivity of project, you can obtain multigroup diagnosis index corresponding with the checkup item sensitivity.
In S102, classified using trained scene classifier to the scene of the realtime image data, obtained
Scene type corresponding with the realtime image data and classification confidence level.
In an embodiment of the present invention, by inputting the reality into the trained support vector cassification model
When the corresponding image feature information descriptor of image data, you can obtain the corresponding scene type of the realtime image data and class
Other confidence level.
Wherein, the scene type includes temple, lawn, snow scenes, pavement, car lane, street, house, field, gloomy
The scene types such as woods, market, school, office building, it will be appreciated by those skilled in the art that the scene type can include tens
Kind even hundreds of kinds of classifications, can be determined according to sample image data, be merely illustrative herein, not indicate that as to this hair
The limitation of bright protection domain.
The classification confidence level refers to that the realtime image data belongs to the probability value of a certain scene type.
For example, the realtime image data is it is possible that 100 percent cannot belong to a certain scene type, therefore, to described
When realtime image data divides scene type, the corresponding classification confidence level of the scene type is also added.For another example, certain piece image
In not only exist and belong to the probability of the first scene type, but also there is the probability for belonging to the second scene type or even also exist and belong to the
The probability of three scene types etc. scene type, and the sum of probability is 1, then directly cannot carry out scene point to the image at this time
Class.When the image 100 percent belongs to a certain scene type, then the corresponding classification confidence level of described image is 1.
Optionally, as shown in figure 4, it is described using trained scene classifier to the scene of the realtime image data into
Row classification, obtains scene type corresponding with the realtime image data and classification confidence level, including step S401 to S403.
In S401, the image feature information descriptor of the realtime image data is extracted.
Wherein, described image characteristic information descriptor including the use of the obtained image SIFT feature values of SIFT and utilizes
LBP operators, color histogram and the gradient information that LBP is obtained.
SIFT feature value is used to detecting and describing the locality characteristic in image, is included in space scale and finds extreme value
Point, and extract position, scale and rotational invariants.
LBP operators are effective texture description operators, for measuring and extracting the texture information of image local, are had to illumination
There is invariance.
In S402, described image characteristic information descriptor is inputted into the trained scene classifier, you can output
The scene type of the realtime image data and its corresponding classification confidence level.
In S403, the classification confidence level is ranked up according to size order, by the maximum classification confidence level
And its corresponding scene type is as the scene type of the realtime image data and classification confidence level.
That is, the scene type of the realtime image data is the maximum corresponding scene class of the classification confidence level
Not, i.e., the scene type that described realtime image data most possibly belongs to.
S103:If the classification confidence level be greater than or equal to predetermined threshold value, according to the scene type, checkup item,
And the correspondence of checkup item sensitivity three, obtain the checkup item corresponding with the scene type and checkup item
Sensitivity.
Wherein, the predetermined threshold value is obtained according to practical experience, for example, the predetermined threshold value is equal to 0.5.
In embodiments of the present invention, the classification confidence level is greater than or equal to the predetermined threshold value, then it represents that described real-time
The possibility that image data belongs to the corresponding scene type of category confidence level is higher, therefore, can according to the scene type,
Checkup item and the correspondence of checkup item sensitivity three obtain the checkup item corresponding with the scene type
And checkup item sensitivity, then adjust the checkup item of video quality diagnosis system and reach the checkup item sensitivity,
So as to carry out the diagnosis of the video quality of the realtime image data.
S104:The checkup item of adjustment video quality diagnosis system reaches the checkup item sensitivity, thus into
The video quality diagnosis of the row realtime image data.
That is, each checkup item has corresponding checkup item sensitivity.It is colour cast in checkup item, clear
It spends, brightness exception, blank screen, shakes, freezes, noise has corresponding checkup item sensitivity respectively.
Optionally, it is described to be classified using trained scene classifier to the scene of the realtime image data, it obtains
After obtaining scene type corresponding with the realtime image data and classification confidence level, further include:If the classification confidence level is small
In predetermined threshold value, then the checkup item sensitivity of the video quality diagnosis system is not adjusted, directly using default parameters,
Carry out video quality diagnosis.
That is, the classification confidence level is less than the predetermined threshold value, then it represents that the realtime image data belongs to this
The possibility of the corresponding scene type of classification confidence level is relatively low, therefore, can not have to adjust the video quality diagnosis system
The checkup item sensitivity directly using default parameters, carries out video quality diagnosis.
Optionally, the checkup item of the adjustment video quality diagnosis system reaches the checkup item sensitivity,
After carrying out the diagnosis of the video quality of the realtime image data, further include:If the described of the realtime image data is examined
The characteristic value of disconnected project meets preset threshold condition, then exports the alarm control signal of the checkup item, so as to control alarm
Device sends out alarm.
If for example, the checkup item is colour cast, the realtime image data is divided into n block sub-image datas, is obtained every
The color component information of block sub-image data, judges whether the color component information of every piece of sub-image data is more than threshold value 1;It is if big
In threshold value 1, then the block sub-image data is colour cast image block;If the sum of colour cast image block is more than threshold value 2, the reality is judged
When image data entirety colour cast, export colour cast alarm control signal, so as to which warning device be controlled to send out alarm.
For another example, if the checkup item is clarity, the gradient information of the realtime image data is calculated;If the gradient
The value of information is less than predetermined threshold value, then judges that the realtime image data clarity is inadequate, exports clarity alarm control signal, from
And warning device is controlled to send out alarm.
Those skilled in the art know, are merely illustrative description herein, are not interpreted as the concrete restriction to the present invention.
It should be understood that the size of the serial number of each step is not meant to the priority of execution sequence, each process in above-described embodiment
Execution sequence should determine that the implementation process without coping with the embodiment of the present invention forms any limit with its function and internal logic
It is fixed.
Fig. 5 shows a kind of device 500 of video quality diagnosis based on scene classification provided in an embodiment of the present invention
Structure diagram, the device include:Image acquisition unit 501, taxon 502, sensitivity acquiring unit 503 and adjustment unit
504,
Image acquisition unit 501, for obtaining realtime image data;
Taxon 502, for being classified using scene classifier to the scene of the realtime image data, obtain with
The corresponding scene type of the realtime image data and classification confidence level;
Sensitivity acquiring unit 503, if being greater than or equal to predetermined threshold value for the classification confidence level, according to the field
The correspondence of scape classification, checkup item and checkup item sensitivity three, obtain it is corresponding with the scene type described in examine
Disconnected project and checkup item sensitivity;
Adjustment unit 504, the checkup item for adjusting video quality diagnosis system reach the checkup item spirit
Sensitivity, so as to carry out the diagnosis of the video quality of the realtime image data.
Further, described device 500 further include sample acquisition unit, extraction unit and training unit.
Wherein, the sample acquisition unit, for obtain multiple sample image datas and with each sample image number
According to corresponding sample scene type and sample class confidence level;
Extraction unit, for extracting the image feature information descriptor of the sample image data;
Training unit, for being retouched using the sample scene type, sample class confidence level and described image characteristic information
Symbol is stated, scene classifier is trained, obtains the trained scene classifier.
Further, described device 500 further include diagnosis unit and establish unit.
Diagnosis unit, for the sample graph with the identical sample scene type for obtaining the acquiring unit
As data are respectively as one group of diagnosis object of the video quality diagnosis system, for object is diagnosed described in every group, institute is adjusted
The checkup item sensitivity corresponding with checkup item of video quality diagnosis system is stated, is obtained and the checkup item sensitivity pair
The diagnosis index answered;
Unit is established, for diagnosing object for described in every group, by checkup item corresponding with best diagnosis index spirit
Sensitivity is as checkup item sensitivity corresponding with the checkup item of the sample scene type, so as to establish the scene class
Not, checkup item and the correspondence of checkup item sensitivity three.
It should be noted that for convenience and simplicity of description, a kind of video quality based on scene classification of foregoing description
The specific work process of the device 500 of diagnosis, can be no longer excessive superfluous herein referring to figs. 1 to the corresponding process of Fig. 4 the methods
It states.
Fig. 6 is another structure of device of a kind of video quality diagnosis based on scene classification that one embodiment of the invention provides
Schematic diagram.As shown in fig. 6, the device 6 of the diagnosis of the video quality based on scene classification of the embodiment includes:Processor 60 is deposited
Reservoir 61 and the computer program 62 that can be run in the memory 61 and on the processor 60 is stored in, such as based on
The video quality diagnostic program of scene classification.The processor 60 realizes above-mentioned each be based on when performing the computer program 62
Step in the video quality diagnosing method embodiment of scene classification, such as step 101 shown in FIG. 1 is to 104.It is alternatively, described
Processor 60 realizes the function of each module/unit in above-mentioned each device embodiment, such as Fig. 5 when performing the computer program 62
The function of shown unit 51 to 54.
Illustratively, the computer program 62 can be divided into one or more module/units, it is one or
Multiple module/units are stored in the memory 61, and are performed by the processor 60, to complete the present invention.Described one
A or multiple module/units can be the series of computation machine program instruction section that can complete specific function, which is used for
Implementation procedure of the computer program 62 in the device 6 of the video quality diagnosis based on scene classification is described.For example,
It is (empty that the computer program 62 can be divided into image acquisition unit, taxon, sensitivity acquiring unit, adjustment unit
Intend the module in device), each unit concrete function is as follows:
Image acquisition unit, for obtaining realtime image data;
Taxon, for being classified using scene classifier to the scene of the realtime image data, acquisition and institute
State the corresponding scene type of realtime image data and classification confidence level;
Sensitivity acquiring unit, if being greater than or equal to predetermined threshold value for the classification confidence level, according to the scene
The correspondence of classification, checkup item and checkup item sensitivity three obtains the diagnosis corresponding with the scene type
Project and checkup item sensitivity;
Adjustment unit, it is sensitive that the checkup item for adjusting video quality diagnosis system reaches the checkup item
Degree, so as to carry out the diagnosis of the video quality of the realtime image data.
The device 6 of the video quality diagnosis based on scene classification can be desktop PC, notebook, palm electricity
The computing devices such as brain and cloud server.The video quality diagnostic device based on scene classification may include, but be not limited only to,
Processor 60, memory 61.It will be understood by those skilled in the art that Fig. 6 is only based on the video quality diagnosis of scene classification
Device 6 example, do not form to based on scene classification video quality diagnose device 6 restriction, can include than figure
Show that more or fewer components either combine certain components or different components, such as the video based on scene classification
Quality diagnosis device can also include input-output equipment, network access equipment, bus etc..
Alleged processor 60 can be central processing unit (Central Processing Unit, CPU), can also be
Other general processors, digital signal processor (Digital Signal Processor, DSP), application-specific integrated circuit
(Application Specific Integrated Circuit, ASIC), ready-made programmable gate array (Field-
Programmable Gate Array, FPGA) either other programmable logic device, discrete gate or transistor logic,
Discrete hardware components etc..General processor can be microprocessor or the processor can also be any conventional processor
Deng.
The memory 61 can be the storage inside list of the device 6 of the video quality diagnosis based on scene classification
Member, such as the hard disk or memory of the device 6 of the diagnosis of the video quality based on scene classification.The memory 61 can also be described
The External memory equipment of the device 6 of video quality diagnosis based on scene classification, such as the video matter based on scene classification
Measure the plug-in type hard disk being equipped on the device 6 of diagnosis, intelligent memory card (Smart Media Card, SMC), secure digital
(Secure Digital, SD) blocks, flash card (Flash Card) etc..Further, the memory 61 can also both include
The internal storage unit of the device 6 of the video quality diagnosis based on scene classification also includes External memory equipment.It is described to deposit
Reservoir 61 is used to store its needed for the device that the computer program and the video quality based on scene classification diagnose
His program and data.The memory 61 can be also used for temporarily storing the data that has exported or will export.
It is apparent to those skilled in the art that for convenience of description and succinctly, only with above-mentioned each work(
Can unit, module division progress for example, in practical application, can be as needed and by above-mentioned function distribution by different
Functional unit, module are completed, i.e., the internal structure of described device are divided into different functional units or module, more than completion
The all or part of function of description.Each functional unit, module in embodiment can be integrated in a processing unit, also may be used
To be that each unit is individually physically present, can also two or more units integrate in a unit, it is above-mentioned integrated
The form that hardware had both may be used in unit is realized, can also be realized in the form of SFU software functional unit.In addition, each function list
Member, the specific name of module are not limited to the protection domain of the application also only to facilitate mutually distinguish.Above system
The specific work process of middle unit, module can refer to the corresponding process in preceding method embodiment, and details are not described herein.
In the above-described embodiments, it all emphasizes particularly on different fields to the description of each embodiment, is not described in detail or remembers in some embodiment
The part of load may refer to the associated description of other embodiments.
Those of ordinary skill in the art may realize that each exemplary lists described with reference to the embodiments described herein
Member and algorithm steps can be realized with the combination of electronic hardware or computer software and electronic hardware.These functions are actually
It is performed with hardware or software mode, specific application and design constraint depending on technical solution.Professional technician
Described function can be realized using distinct methods to each specific application, but this realization is it is not considered that exceed
The scope of the present invention.
In embodiment provided by the present invention, it should be understood that disclosed device/terminal device and method, it can be with
It realizes by another way.For example, device described above/terminal device embodiment is only schematical, for example, institute
The division of module or unit is stated, only a kind of division of logic function can have other dividing mode in actual implementation, such as
Multiple units or component may be combined or can be integrated into another system or some features can be ignored or does not perform.Separately
A bit, shown or discussed mutual coupling or direct-coupling or communication connection can be by some interfaces, device
Or the INDIRECT COUPLING of unit or communication connection, can be electrical, machinery or other forms.
The unit illustrated as separating component may or may not be physically separate, be shown as unit
The component shown may or may not be physical unit, you can be located at a place or can also be distributed to multiple
In network element.Some or all of unit therein can be selected according to the actual needs to realize the mesh of this embodiment scheme
's.
In addition, each functional unit in each embodiment of the present invention can be integrated in a processing unit, it can also
That each unit is individually physically present, can also two or more units integrate in a unit.Above-mentioned integrated list
The form that hardware had both may be used in member is realized, can also be realized in the form of SFU software functional unit.
If the integrated module/unit realized in the form of SFU software functional unit and be independent product sale or
In use, it can be stored in a computer read/write memory medium.Based on such understanding, the present invention realizes above-mentioned implementation
All or part of flow in example method, can also instruct relevant hardware to complete, the meter by computer program
Calculation machine program can be stored in a computer readable storage medium, the computer program when being executed by processor, it can be achieved that on
The step of stating each embodiment of the method.Wherein, the computer program includes computer program code, the computer program generation
Code can be source code form, object identification code form, executable file or certain intermediate forms etc..The computer-readable medium
It can include:Any entity of the computer program code or device, recording medium, USB flash disk, mobile hard disk, magnetic can be carried
Dish, CD, computer storage, read-only memory (Read-Only Memory, ROM), random access memory (Random
Access Memory, RAM), electric carrier signal, telecommunication signal and software distribution medium etc..It should be noted that the meter
The content that calculation machine readable medium includes can carry out appropriate increase and decrease according to legislation in jurisdiction and the requirement of patent practice,
Such as in certain jurisdictions, according to legislation and patent practice, computer-readable medium does not include electric carrier signal and telecommunications
Signal.
Embodiment described above is merely illustrative of the technical solution of the present invention, rather than its limitations;Although with reference to aforementioned reality
Example is applied the present invention is described in detail, it will be understood by those of ordinary skill in the art that:It still can be to aforementioned each
Technical solution recorded in embodiment modifies or carries out equivalent replacement to which part technical characteristic;And these are changed
Or replace, the spirit and scope for various embodiments of the present invention technical solution that it does not separate the essence of the corresponding technical solution should all
It is included within protection scope of the present invention.
Claims (10)
- A kind of 1. method of the video quality diagnosis based on scene classification, which is characterized in that including:Obtain realtime image data;Classified using trained scene classifier to the scene of the realtime image data, obtained and the realtime graphic The corresponding scene type of data and classification confidence level;If the classification confidence level is greater than or equal to predetermined threshold value, according to the scene type, checkup item and checkup item The correspondence of sensitivity three obtains the checkup item corresponding with the scene type and checkup item sensitivity;The checkup item of adjustment video quality diagnosis system reaches the checkup item sensitivity, described real-time so as to carry out The video quality diagnosis of image data.
- 2. the method as described in claim 1, which is characterized in that before the acquisition realtime image data, further include:Obtain multiple sample image datas and sample scene type corresponding with each sample image data and sample class Confidence level;Extract the image feature information descriptor of the sample image data;Using the sample scene type, sample class confidence level and described image characteristic information descriptor, to scene classifier It is trained, obtains the trained scene classifier.
- 3. method as claimed in claim 2, which is characterized in that it is described obtain multiple sample image datas and with it is each described After the corresponding sample scene type of sample image data and sample class confidence level, further include:Using the sample image data with the identical sample scene type as the video quality diagnosis system One group of diagnosis object, diagnose object for described in every group, adjust the corresponding with checkup item of the video quality diagnosis system Checkup item sensitivity, obtain diagnosis index corresponding with the checkup item sensitivity;Diagnose object for described in every group, using checkup item sensitivity corresponding with best diagnosis index as with the sample The corresponding checkup item sensitivity of checkup item of scene type, so as to establish the scene type, checkup item and diagnosis item The correspondence of mesh sensitivity three.
- 4. the method as described in claim 1, which is characterized in that described to utilize trained scene classifier to the real-time figure As the scene of data is classified, scene type corresponding with the realtime image data and classification confidence level are obtained, including:Extract the image feature information descriptor of the realtime image data;Described image characteristic information descriptor is inputted into the trained scene classifier, exports the realtime image data Scene type and its corresponding classification confidence level;The classification confidence level is ranked up according to size order, by the maximum classification confidence level and its corresponding scene Classification is as the scene type of the realtime image data and classification confidence level.
- 5. the method as described in claim 1, which is characterized in that described to utilize trained scene classifier to the real-time figure As the scene of data is classified, after obtaining scene type corresponding with the realtime image data and classification confidence level, and also Including:If the classification confidence level is less than predetermined threshold value, the checkup item spirit of the video quality diagnosis system is not adjusted Sensitivity directly using default parameters, carries out video quality diagnosis.
- 6. the method as described in claim 1, which is characterized in that the checkup item of the adjustment video quality diagnosis system Reach the checkup item sensitivity, after carrying out the diagnosis of the video quality of the realtime image data, further include:If the characteristic value of the checkup item of the realtime image data meets preset threshold condition, the diagnosis item is exported Purpose alarm control signal, so as to which warning device be controlled to send out alarm.
- 7. a kind of device of the video quality diagnosis based on scene classification, which is characterized in that including:Image acquisition unit, for obtaining realtime image data;Taxon for being classified using scene classifier to the scene of the realtime image data, is obtained and the reality When the corresponding scene type of image data and classification confidence level;Sensitivity acquiring unit, if for the classification confidence level be greater than or equal to predetermined threshold value, according to the scene type, Checkup item and the correspondence of checkup item sensitivity three obtain the checkup item corresponding with the scene type And checkup item sensitivity;Adjustment unit, the checkup item for adjusting video quality diagnosis system reach the checkup item sensitivity, from And carry out the video quality diagnosis of the realtime image data.
- 8. device as claimed in claim 7, which is characterized in that described device, further include sample acquisition unit, extraction unit and Training unit;The sample acquisition unit, for obtaining multiple sample image datas and corresponding with each sample image data Sample scene type and sample class confidence level;Extraction unit, for extracting the image feature information descriptor of the sample image data;Training unit, for utilizing the sample scene type, sample class confidence level and described image characteristic information descriptor, Scene classifier is trained, obtains the trained scene classifier.
- 9. a kind of device of the video quality diagnosis based on scene classification, including memory, processor and is stored in described deposit In reservoir and the computer program that can run on the processor, which is characterized in that the processor performs the computer It is realized during program such as the step of any one of claim 1 to 6 the method.
- 10. a kind of computer readable storage medium, the computer-readable recording medium storage has computer program, and feature exists In when the computer program is executed by processor the step of realization such as any one of claim 1 to 6 the method.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201711298596.8A CN108182444A (en) | 2017-12-08 | 2017-12-08 | The method and device of video quality diagnosis based on scene classification |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201711298596.8A CN108182444A (en) | 2017-12-08 | 2017-12-08 | The method and device of video quality diagnosis based on scene classification |
Publications (1)
Publication Number | Publication Date |
---|---|
CN108182444A true CN108182444A (en) | 2018-06-19 |
Family
ID=62545833
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201711298596.8A Pending CN108182444A (en) | 2017-12-08 | 2017-12-08 | The method and device of video quality diagnosis based on scene classification |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108182444A (en) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109446887A (en) * | 2018-09-10 | 2019-03-08 | 易诚高科(大连)科技有限公司 | It is a kind of for picture quality subjectivity evaluation and test image scene generation method is described |
CN110827364A (en) * | 2018-08-07 | 2020-02-21 | 北京优酷科技有限公司 | Method and device for detecting green screen image |
CN112949390A (en) * | 2021-01-28 | 2021-06-11 | 浙江大华技术股份有限公司 | Event detection method and device based on video quality |
CN114727093A (en) * | 2022-03-18 | 2022-07-08 | 百度在线网络技术(北京)有限公司 | Data analysis method and device, electronic equipment and computer storage medium |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104780361A (en) * | 2015-03-27 | 2015-07-15 | 南京邮电大学 | Quality evaluation method for urban video monitoring system |
CN105163111A (en) * | 2015-09-10 | 2015-12-16 | 宁波大学 | Method for evaluating visual comfort of three-dimensional image based on classification of scene modes |
-
2017
- 2017-12-08 CN CN201711298596.8A patent/CN108182444A/en active Pending
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104780361A (en) * | 2015-03-27 | 2015-07-15 | 南京邮电大学 | Quality evaluation method for urban video monitoring system |
CN105163111A (en) * | 2015-09-10 | 2015-12-16 | 宁波大学 | Method for evaluating visual comfort of three-dimensional image based on classification of scene modes |
Non-Patent Citations (4)
Title |
---|
AJAY RAGHAVAN, BHASKAR SAHA: "Algorithm design for automated transportation photo enforcement camera image and video quality diagnostic check modules", 《VIDEO SURVEILLANCE AND TRANSPORTATION IMAGING APPLICATIONS》 * |
FUXIANG LU, XIAOKANG YANG, RUI ZHANG, SONGYU YU: "Scene categorization based on heterogeneous features", 《SOCIETY OF PHOTO-OPTICAL INSTRUMENTATION ENGINEERS(SPIE)》 * |
李德仁: "《空间信息系统的集成与实现》", 31 March 2002, 武汉大学出版社 * |
邵军: "基于场景分类的图像质量评价", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110827364A (en) * | 2018-08-07 | 2020-02-21 | 北京优酷科技有限公司 | Method and device for detecting green screen image |
CN109446887A (en) * | 2018-09-10 | 2019-03-08 | 易诚高科(大连)科技有限公司 | It is a kind of for picture quality subjectivity evaluation and test image scene generation method is described |
CN109446887B (en) * | 2018-09-10 | 2022-03-25 | 易诚高科(大连)科技有限公司 | Image scene description generation method for subjective evaluation of image quality |
CN112949390A (en) * | 2021-01-28 | 2021-06-11 | 浙江大华技术股份有限公司 | Event detection method and device based on video quality |
CN112949390B (en) * | 2021-01-28 | 2024-03-15 | 浙江大华技术股份有限公司 | Event detection method and device based on video quality |
CN114727093A (en) * | 2022-03-18 | 2022-07-08 | 百度在线网络技术(北京)有限公司 | Data analysis method and device, electronic equipment and computer storage medium |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN104268498B (en) | A kind of recognition methods of Quick Response Code and terminal | |
CN108182444A (en) | The method and device of video quality diagnosis based on scene classification | |
CN108280477B (en) | Method and apparatus for clustering images | |
CN109002820B (en) | License plate recognition method and device and related equipment | |
CN105446988B (en) | The method and apparatus for predicting classification | |
CN107742290A (en) | Plant disease identifies method for early warning and device | |
CN110533654A (en) | The method for detecting abnormality and device of components | |
CN112348765A (en) | Data enhancement method and device, computer readable storage medium and terminal equipment | |
CN112581438B (en) | Slice image recognition method and device, storage medium and electronic equipment | |
CN109117773A (en) | A kind of characteristics of image point detecting method, terminal device and storage medium | |
CN108805859A (en) | A kind of image detecting method, image detection device and terminal device | |
CN109544862A (en) | Activity recognition method, apparatus, storage medium and equipment based on smart home | |
CN108205580A (en) | A kind of image search method, device and computer readable storage medium | |
CN110378254B (en) | Method and system for identifying vehicle damage image modification trace, electronic device and storage medium | |
CN113065609B (en) | Image classification method, device, electronic equipment and readable storage medium | |
CN112463859B (en) | User data processing method and server based on big data and business analysis | |
CN112183212A (en) | Weed identification method and device, terminal equipment and readable storage medium | |
CN108229658A (en) | The implementation method and device of object detector based on finite sample | |
CN110163109A (en) | A kind of lane line mask method and device | |
CN113608916A (en) | Fault diagnosis method and device, electronic equipment and storage medium | |
CN110287767A (en) | Can attack protection biopsy method, device, computer equipment and storage medium | |
CN108764139A (en) | A kind of method for detecting human face, mobile terminal and computer readable storage medium | |
CN107909569A (en) | One kind flower screen detection method, flower screen detection device and electronic equipment | |
CN109376079A (en) | The test method and server that interface calls | |
CN112966687B (en) | Image segmentation model training method and device and communication 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 |
Application publication date: 20180619 |
|
RJ01 | Rejection of invention patent application after publication |