CN108898591A - Methods of marking and device, electronic equipment, the readable storage medium storing program for executing of picture quality - Google Patents
Methods of marking and device, electronic equipment, the readable storage medium storing program for executing of picture quality Download PDFInfo
- Publication number
- CN108898591A CN108898591A CN201810654626.2A CN201810654626A CN108898591A CN 108898591 A CN108898591 A CN 108898591A CN 201810654626 A CN201810654626 A CN 201810654626A CN 108898591 A CN108898591 A CN 108898591A
- Authority
- CN
- China
- Prior art keywords
- probability distribution
- input picture
- score value
- mlp
- picture
- 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
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30168—Image quality inspection
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Biophysics (AREA)
- Evolutionary Computation (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Biomedical Technology (AREA)
- Quality & Reliability (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Image Analysis (AREA)
Abstract
The disclosure is directed to a kind of methods of marking of picture quality and device, electronic equipment, readable storage medium storing program for executing.The method includes：Determine input picture；Determine probability distribution of the input picture in scoring range；The score value of the input picture is determined according to the probability distribution.As it can be seen that by calculating the available scoring preference to public users to input picture of probability distribution in the present embodiment；The demand that user can be met according to the score value that the scoring preference of public users determines facilitates user preferably to manage image according to score value, promotes usage experience.
Description
Technical field
This disclosure relates to the methods of marking and device of technical field of image processing more particularly to a kind of picture quality, electronics
Equipment, readable storage medium storing program for executing.
Background technique
With the raising of camera quality on smart phone, user tends to smart phone shooting figure anywhere or anytime
Picture is conducive to that user is helped to keep beautiful moment here.
Due to the limited storage space of existing smart phone, user needs periodically to deposit to hard disk or delete pipe image shifting
Reason.For deleting management, during deleting, user needs to browse all images, then will be unwanted according to hobby
Image-erasing.But if amount of images is excessive, such as 1000 or more, then user's excessive time can be expended.It is certain when deleting
After degree, user can hesitate between multiple images, can not determine the image of deletion.
Summary of the invention
The disclosure provides the methods of marking and device, electronic equipment, readable storage medium storing program for executing of a kind of picture quality, to solve phase
Deficiency in the technology of pass.
According to the first aspect of the embodiments of the present disclosure, a kind of methods of marking of picture quality is provided, including：
Determine input picture；
Determine probability distribution of the input picture in scoring range；
The score value of the input picture is determined according to the probability distribution.
Optionally it is determined that probability distribution of the input picture in scoring range includes：
The input picture is pre-processed, standard picture is obtained；
The standard picture is inputted into the Feature Selection Model, the standard drawing is determined by the Feature Selection Model
The feature vector of picture；
Described eigenvector is inputted into multi-layer perception (MLP), determines the input picture described by the multi-layer perception (MLP)
Probability distribution in the range that scores.
Optionally, before the feature vector that the standard picture is determined by the Feature Selection Model, the method is also
Including：
Obtain N sample images and the respective artificial Marking Probability distribution of the N sample images；N is positive integer；
The N sample images are sequentially input into multi-layer perception (MLP), until the loss function in the multi-layer perception (MLP) is received
Multi-layer perception (MLP) described in deconditioning when holding back, in the multi-layer perception (MLP) after being trained；The loss function is based on described artificial
The probability distribution of Marking Probability distribution and multi-layer perception (MLP) output judges whether to restrain.
Optionally, the score value for determining the input picture according to the probability distribution includes：
Probability value of the input picture in the scoring range at each score value is determined according to the probability distribution；
The score value of the input picture is determined based on each score value and the probability value.
According to the second aspect of an embodiment of the present disclosure, a kind of scoring apparatus of picture quality is provided, including：
Input picture determining module, for determining input picture；
Probability distribution determining module, for determining probability distribution of the input picture in scoring range；
Score value determining module, for determining the score value of the input picture according to the probability distribution.
Optionally, the probability distribution determining module includes：
Standard picture acquiring unit obtains standard picture for pre-processing to the input picture；
Feature vector determination unit passes through the feature for the standard picture to be inputted the Feature Selection Model
Extract the feature vector that model determines the standard picture；
Probability distribution determination unit passes through the multi-layer perception (MLP) for described eigenvector to be inputted multi-layer perception (MLP)
Determine probability distribution of the input picture in the scoring range.
Optionally, described device further includes：
Probability distribution obtains module, general for obtaining N sample images and the N respective artificial marks of sample image
Rate distribution；N is positive integer；
Characteristic model training module, for the N sample images to be sequentially input multi-layer perception (MLP), until the multilayer
Multi-layer perception (MLP) described in deconditioning when loss function convergence in perceptron, in the multi-layer perception (MLP) after being trained；It is described
Loss function judges whether to restrain based on the probability distribution of the artificial Marking Probability distribution and multi-layer perception (MLP) output.
Optionally, the score value determining module includes：
Probability value determination unit, for determining that the input picture is each in the scoring range according to the probability distribution
Probability value at score value；
Score value determination unit, for determining the scoring of the input picture based on each score value and the probability value
Value.
According to the third aspect of an embodiment of the present disclosure, a kind of electronic equipment is provided, including：
Processor；
For storing the memory of the processor-executable instruction；
Wherein, the processor is configured to executing the executable instruction in the memory to realize described in first aspect
The step of method.
According to a fourth aspect of embodiments of the present disclosure, a kind of computer readable storage medium is provided, calculating is stored thereon with
Machine program, when which is executed by processor the step of realization first aspect the method.
The technical scheme provided by this disclosed embodiment can include the following benefits：
By determining probability distribution of the input picture in scoring range in the embodiment of the present disclosure, and based on probability distribution and
Each score value calculates the score value of input picture.As it can be seen that available to masses by calculating probability distribution in the present embodiment
Scoring preference of the user to input picture；The need of user can be met according to the score value that the scoring preference of public users determines
It asks, user is facilitated preferably to manage image according to score value, promote usage experience.
It should be understood that above general description and following detailed description be only it is exemplary and explanatory, not
The disclosure can be limited.
Detailed description of the invention
The drawings herein are incorporated into the specification and forms part of this specification, and shows the implementation for meeting the disclosure
Example, and together with specification for explaining the principles of this disclosure.
Fig. 1 is a kind of flow diagram of the methods of marking of picture quality shown according to an exemplary embodiment；
Fig. 2 is a kind of flow diagram of the methods of marking of the picture quality shown according to another exemplary embodiment；
Fig. 3 is a kind of application scenario diagram shown according to an exemplary embodiment；
Fig. 4 is a kind of application scenario diagram shown according to another exemplary embodiment；
A kind of Fig. 5~Fig. 8 block diagram of the scoring apparatus of picture quality shown according to an exemplary embodiment；
Fig. 9 is the block diagram of a kind of electronic equipment shown according to an exemplary embodiment.
Specific embodiment
Example embodiments are described in detail here, and the example is illustrated in the accompanying drawings.Following description is related to
When attached drawing, unless otherwise indicated, the same numbers in different drawings indicate the same or similar elements.Following exemplary embodiment
Described in embodiment do not represent all implementations consistent with this disclosure.On the contrary, they be only with it is such as appended
The consistent device example of some aspects be described in detail in claims, the disclosure.
Currently, user tends to shoot image anywhere or anytime with smart phone, to keep beautiful moment here.Due to existing
The limited storage space of smart phone, user need periodically to deposit to hard disk or delete management image shifting.It is to delete management
Example, during deleting, user needs to browse all images, is then deleted unwanted picture according to hobby.But if
Amount of images is excessive, such as 1000 or more, then can expend user's excessive time.After deleting to a certain extent, user can be
It hesitates between multiple images, can not determine the image of deletion.
In order to solve the above technical problems, Fig. 1 is basis the embodiment of the invention provides a kind of methods of marking of picture quality
A kind of methods of marking of picture quality shown in one exemplary embodiment.It will be appreciated that scoring provided in an embodiment of the present invention
Method can be applied to the electronic equipments such as computer, mobile phone, tablet computer, for convenience of describing, with intelligent hand in subsequent embodiment
It is described for machine.Referring to Fig. 1, a kind of methods of marking of picture quality includes step 101, step 102 and step 103：
101, determine input picture.
In the present embodiment, the processor of smart phone can obtain input picture automatically, such as when user selects " picture library "
When, it is meant that user needs to browse image therein, and under this scene, processor can successively obtain the image stored in " picture library "
As input picture.Certainly, processor can also detect the selection operation of user, using the corresponding image of selection operation as input
Image.
102, determine probability distribution of the input picture in scoring range.
In the present embodiment, processor can call preset algorithm to extract the feature vector of input picture, according to feature to
Amount can determine probability distribution of the input picture in scoring range.Since subsequent implementation regular meeting is described in detail, herein
First it is not illustrated.
Wherein, scoring range refers to that the maximum value of input picture score value and minimum value constitute scoring section, such as scores
It may range from 1~10 or 1~100 etc..In one embodiment, scoring range is selected as 1~10.Probability distribution refers to use
In the probabilistic law of each score value in statement scoring range.
103, the score value of the input picture is determined according to the probability distribution.
In the present embodiment, processor can determine probability value of the input picture at each score value according to probability distribution,
The score value of input picture can be determined according to each score value and its corresponding probability value later.
So far, available that input is schemed to public users by determining the probability distribution of input picture in the present embodiment
The scoring preference of picture, i.e., the scoring preference in the present embodiment according to public users determine that the score value of input picture can more accord with
The scoring demand for sharing family facilitates user preferably to manage image according to score value, promotes usage experience.
Fig. 2 is a kind of flow diagram of the methods of marking of picture quality shown according to an exemplary embodiment.Referring to
Fig. 2, a kind of methods of marking of picture quality include step 201~step 203：
201, determine input picture.
Step 201 and the specific method of step 101 are consistent with principle, and detailed description please refers to the correlation of Fig. 1 and step 101
Content, details are not described herein again.
202, determine probability distribution of the input picture in scoring range.
In the present embodiment, the mode of processor acquisition probability distribution, at least may include mode one and mode two.It is right below
Mode one and mode two are described：
Mode one
Firstly, processor can call preset deep-neural-network, the deep-neural-network is trained to be finished.
In view of the computing resource of smart phone, above-mentioned deep-neural-network can using MobileNet network,
The light-weighted networks such as ShuffleNet network, SqueezeNet network, so as to save the computing resource of smart phone.When
So, when the equipment such as electronic equipment PC, server, processor can call other neural networks, such as CNN neural network, BP
Neural network etc., equally can achieve the effect for extracting the feature vector of input picture, and corresponding scheme falls into the application's
Protection scope.
Secondly, input picture is input to deep-neural-network by processor, it is defeated to extract this by the deep-neural-network
Enter the feature of image, and output (the corresponding step 2021) in the form of feature vector.
It will be appreciated that before input picture is input to deep-neural-network by processor, it can be according to deep-neural-network
Requirement input picture is pre-processed.By taking size adjusting as an example, processor can will be defeated according to the demand of deep-neural-network
The size adjusting for entering image is to be sized, such as being sized to be 224*224 (unit pixel), to obtain standard drawing
Picture.Preprocessor standard picture is input in deep-neural-network, be able to ascend in this way deep-neural-network extract feature
The efficiency of vector.Technical staff can adjust pretreated component content according to concrete scene, such as pretreatment can also include
Normalized (subtracting mean value, except variance) etc., scheme adjusted equally falls into the protection scope of the application.
Again, feature vector is input in multi-layer perception (MLP) (MLP) by processor, can be determined by Logic Regression Models
Probability distribution (corresponding step 2022) of the input picture in scoring range out.
It will be appreciated that processor may call upon other neural networks, such as BP neural network, to substitute Multilayer Perception
Machine, so that neural network determines that it corresponds to the probability value of classification according to feature vector, then by multiple classification and its probability value
It is combined into the probability distribution of scoring range, the scheme of this step equally may be implemented.Technical staff can select according to concrete scene
It selects suitable model, algorithm etc. and determines probability distribution, be not limited thereto.
It should be noted that will acquire probability distribution in the present embodiment is divided into two steps to handle, advantage is：Processing
Device can individually train deep-neural-network or multi-layer perception (MLP), will extract feature vector process and recognition feature vector process solution
Coupling both makes the training of corresponding network without directly connection, can be improved deep-neural-network extract the accuracy of feature vector with
And improve the accuracy of multi-layer perception (MLP) recognition feature vector.
In one embodiment, to get public users to the scoring preference of input picture, deep-neural-network is being trained
Or when multi-layer perception (MLP), N (N is positive integer) sample image can be used, the value of N can be according to the essence of deep-neural-network
Degree is adjusted, such as 1000,2000 or 10000 etc., is not limited thereto.Wherein every sample image may include corresponding to
The distribution of artificial Marking Probability, which is to count at least M public users to carry out manually sample image
What the score value (such as 1~10 point) of mark obtained.Such as M value 100, mark score value are 1,2,3,4,5,6,7,8,9,10
Number of users be respectively：0,5,8,9,9,10,12,15,16,10 and 6, then the corresponding artificial Marking Probability of the sample image divides
Cloth P={ 0,0.05,0.08,0.09,0.09,0.10,0.10,0.12,0.15,0.16,0.10,0.06 }.
Processor opens sample images training multi-layer perception (MLP) processes using N：
First, the artificial Marking Probability of N sample images and every sample image is distributed to form image training by processor
Collection.
Second, every sample image in training set of images is separately input to multi-layer perception (MLP) by processor, passes through multilayer sense
Know that machine determines the probability distribution in scoring range.
Third, the probability point that processor is exported according to the artificial Marking Probability distribution of every sample image and multi-layer perception (MLP)
Cloth determines whether loss function restrains, and second step is returned when not restraining, and terminates to train in convergence.
Wherein, loss function is as follows：
Wherein
Wherein, i indicates i-th of score value in value range；K indicates k-th of score value in value range；Pi expression is based on
The probability value of score value i obtained from the determined probability distribution of multi-layer perception (MLP)；It indicates according to multi-layer perception (MLP)
The probability distribution determined and the integrated value calculated；Pi ' is indicated based on the general of score value i obtained from the distribution of artificial Marking Probability
Rate value；It indicates according to the distribution of artificial Marking Probability and calculated integrated value.
In the present embodiment during training multi-layer perception (MLP), all score values manually marked are taken full advantage of, it can
Promote the service efficiency of sample image.Also, the output valve of trained multi-layer perception (MLP) is probability distribution, i.e., multiple score values
And probability value rather than a score value, the scoring preference of different user in public users can be simulated, to make input picture
Final score value more close to the scoring preference of public users, i.e. the accuracy of score value is higher.
Mode two
Firstly, processor can call preset neural network, the neural network is trained to be finished.
Then, input picture is input to preset neural network by processor, and obtaining its output data is scoring range
Probability distribution.Such as processor can be exported using the corresponding data of full articulamentum in neural network as a probability distribution.
Wherein, processor is using the serial number of tie point each in full articulamentum as score value, and the data of each tie point are as probability value, in this way
The probability distribution that available score value and corresponding probability value are formed, is shown below：
Wherein, i indicates that i-th of score value in scoring range, pi indicate the corresponding probability value of i-th of score value, work as i=
When 10, the sum of corresponding probability value of 10 score values is equal to 1；P indicates probability distribution.
203, the score value of the input picture is determined according to the probability distribution.
In the present embodiment, processor can determine that input picture in scoring range at each score value according to probability distribution
Probability value (corresponding step 2031), such as 1 point (0.05) (1 point is score value, and 0.05 is probability value), 2 points (0.06), 3 points
(0.08) ..., 8 points (0.30), 9 points (0.25) and 10 points (0.20).
Later, processor (corresponds to step 2032), such as according to the score value of each score value and probability value calculating input image
Under：
Wherein, i indicates that i-th of score value in scoring range, pi indicate that the corresponding probability value of i-th of score value, s indicate
The final score value of input picture.
204, score value is showed simultaneously with input picture.
In the present embodiment, processor exports the score value of input picture, shows score value simultaneously with input picture.Exhibition
Now mode may include：Score value and input picture are side by side or score value is superimposed upon on input picture.Referring to Fig. 3, processing
Score value is superimposed on input picture by device in a manner of pop-up.
Certainly, processor can also classify the image in " picture library " according to score value, classification results as shown in figure 4, this
Sample user directly can carry out bulk management to each classification, such as delete the lower multiple images that score, to improve management
Efficiency.
So far, multi-layer perception (MLP) is trained by using great amount of samples image in the present embodiment, due to every sample
Image is manually marked by multiple users, therefore processor determines that the probability distribution of input picture is able to reflect out public use
The score value of input picture can more meet to be determined at family according to the scoring preference of public users to the scoring preference of input picture
The scoring demand of user, such user can preferably manage image according to score value, improve the efficiency of management and promotion uses body
It tests.
Fig. 5 is a kind of block diagram of the scoring apparatus of picture quality shown according to an exemplary embodiment.Referring to Fig. 5, one
The scoring apparatus 500 of kind of picture quality includes：
Input picture determining module 501, for determining input picture；
Probability distribution determining module 502, for determining probability distribution of the input picture in scoring range；
Score value determining module 503, for determining the score value of the input picture according to the probability distribution.
As it can be seen that available inclined to the scoring of input picture to public users by calculating probability distribution in the present embodiment
It is good；The demand that user can be met according to the score value that the scoring preference of public users is determined, facilitates user according to score value
Preferably management image promotes usage experience.
Fig. 6 is a kind of block diagram of the scoring apparatus of picture quality shown according to an exemplary embodiment.Referring to Fig. 6,
On the basis of scoring apparatus 500 shown in Fig. 5, probability distribution determining module 502 includes：
Standard picture acquiring unit 601 obtains standard picture for pre-processing to the input picture；
Feature vector determination unit 602 passes through the spy for the standard picture to be inputted the Feature Selection Model
Sign extracts the feature vector that model determines the standard picture；
Probability distribution determination unit 603 passes through the Multilayer Perception for described eigenvector to be inputted multi-layer perception (MLP)
Machine determines probability distribution of the input picture in the scoring range.
As it can be seen that the feature vector for extracting input picture and recognition feature vector are divided into two steps in the present embodiment, it can
Model and multi-layer perception (MLP) are extracted with independent training characteristics, can be improved the utilization rate of sample image, and improve feature extraction
Model determines the accuracy of feature vector and improves the accuracy that multi-layer perception (MLP) determines probability distribution.
Fig. 7 is a kind of block diagram of the scoring apparatus of picture quality shown according to an exemplary embodiment.Referring to Fig. 7,
On the basis of scoring apparatus shown in Fig. 6, further include：
Probability distribution obtains module 701, for obtaining N sample images and the N respective artificial marks of sample image
Infuse probability distribution；N is positive integer；
Characteristic model training module 702, for the N sample images to be sequentially input multi-layer perception (MLP), until described
Multi-layer perception (MLP) described in deconditioning when loss function convergence in multi-layer perception (MLP), in the multi-layer perception (MLP) after being trained；
The loss function judges whether to receive based on the probability distribution of the artificial Marking Probability distribution and multi-layer perception (MLP) output
It holds back.
As it can be seen that all score values manually marked are taken full advantage of during training multi-layer perception (MLP) in the present embodiment,
It is able to ascend the service efficiency of sample image.Also, the output valve of trained multi-layer perception (MLP) is probability distribution, i.e., multiple to comment
Score value and probability value rather than a score value, can simulate the scoring preference of different user in public users, to make to input
For the final score value of image more close to the scoring preference of public users, i.e. the accuracy of score value is higher.
Fig. 8 is a kind of block diagram of the scoring apparatus of picture quality shown according to an exemplary embodiment.Referring to Fig. 8,
On the basis of scoring apparatus 500 shown in Fig. 5, the score value determining module 503 includes：
Probability value determination unit 801, for determining the input picture in the scoring range according to the probability distribution
Probability value at interior each score value；
Score value determination unit 802, for determining the input picture based on each score value and the probability value
Score value.
By probability distribution acquisition probability value in the present embodiment, score value is then determined according to probability value, it not only can be square
Just technical staff's training multi-layer perception (MLP), also facilitates user to manage image according to score value.
Fig. 9 is the block diagram of a kind of electronic equipment shown according to an exemplary embodiment.For example, electronic equipment 900 can be with
It is mobile phone, computer, digital broadcasting terminal, messaging device, game console, tablet device, Medical Devices, body-building
Equipment, personal digital assistant etc..
Referring to Fig. 9, electronic equipment 900 may include following one or more components：Processing component 902, memory 904,
Power supply module 906, multimedia component 908, audio component 910, the interface 912 of input/output (I/O), sensor module 914,
And communication component 916.Wherein, memory 904 is used to store the executable instruction of processing component 902.Processing component 902 is from depositing
Reservoir 904 reads instruction to realize：
Determine input picture；
Determine probability distribution of the input picture in scoring range；
The score value of the input picture is determined according to the probability distribution.
The integrated operation of the usual control device 900 of processing component 902, such as with display, telephone call, data communication, phase
Machine operation and record operate associated operation.Processing component 902 may include that one or more processors 920 refer to execute
It enables.In addition, processing component 902 may include one or more modules, convenient for the friendship between processing component 902 and other assemblies
Mutually.For example, processing component 902 may include multi-media module, to facilitate between multimedia component 908 and processing component 902
Interaction.
Memory 904 is configured as storing various types of data to support the operation in device 900.These data are shown
Example includes the instruction of any application or method for operating on device 900, contact data, and telephone book data disappears
Breath, picture, video etc..Memory 904 can be by any kind of volatibility or non-volatile memory device or their group
It closes and realizes, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM) is erasable to compile
Journey read-only memory (EPROM), programmable read only memory (PROM), read-only memory (ROM), magnetic memory, flash
Device, disk or CD.
Power supply module 906 provides electric power for the various assemblies of device 900.Power supply module 906 may include power management system
System, one or more power supplys and other with for device 900 generate, manage, and distribute the associated component of electric power.
Multimedia component 908 includes the screen of one output interface of offer between described device 900 and user.One
In a little embodiments, screen may include liquid crystal display (LCD) and touch panel (TP).If screen includes touch panel, screen
Curtain may be implemented as touch screen, to receive input signal from the user.Touch panel includes one or more touch sensings
Device is to sense the gesture on touch, slide, and touch panel.The touch sensor can not only sense touch or sliding action
Boundary, but also detect duration and pressure associated with the touch or slide operation.In some embodiments, more matchmakers
Body component 908 includes a front camera and/or rear camera.When device 900 is in operation mode, such as screening-mode or
When video mode, front camera and/or rear camera can receive external multi-medium data.Each front camera and
Rear camera can be a fixed optical lens system or have focusing and optical zoom capabilities.
Audio component 910 is configured as output and/or input audio signal.For example, audio component 910 includes a Mike
Wind (MIC), when device 900 is in operation mode, when such as call mode, recording mode, and voice recognition mode, microphone is matched
It is set to reception external audio signal.The received audio signal can be further stored in memory 904 or via communication set
Part 916 is sent.In some embodiments, audio component 910 further includes a loudspeaker, is used for output audio signal.
I/O interface 912 provides interface between processing component 902 and peripheral interface module, and above-mentioned peripheral interface module can
To be keyboard, click wheel, button etc..These buttons may include, but are not limited to：Home button, volume button, start button and lock
Determine button.
Sensor module 914 includes one or more sensors, and the state for providing various aspects for device 900 is commented
Estimate.For example, sensor module 914 can detecte the state that opens/closes of device 900, and the relative positioning of component, for example, it is described
Component is the display and keypad of device 900, and sensor module 914 can be with 900 1 components of detection device 900 or device
Position change, the existence or non-existence that user contacts with device 900,900 orientation of device or acceleration/deceleration and device 900
Temperature change.Sensor module 914 may include proximity sensor, be configured to detect without any physical contact
Presence of nearby objects.Sensor module 914 can also include optical sensor, such as CMOS or ccd image sensor, at
As being used in application.In some embodiments, which can also include acceleration transducer, gyro sensors
Device, Magnetic Sensor, pressure sensor or temperature sensor.
Communication component 916 is configured to facilitate the communication of wired or wireless way between device 900 and other equipment.Device
900 can access the wireless network based on communication standard, such as WiFi, 2G or 3G or their combination.In an exemplary implementation
In example, communication component 916 receives broadcast singal or broadcast related information from external broadcasting management system via broadcast channel.
In one exemplary embodiment, the communication component 916 further includes near-field communication (NFC) module, to promote short range communication.Example
Such as, NFC module can be based on radio frequency identification (RFID) technology, Infrared Data Association (IrDA) technology, ultra wide band (UWB) technology,
Bluetooth (BT) technology and other technologies are realized.
In the exemplary embodiment, device 900 can be believed by one or more application specific integrated circuit (ASIC), number
Number processor (DSP), digital signal processing appts (DSPD), programmable logic device (PLD), field programmable gate array
(FPGA), controller, microcontroller, microprocessor or other electronic components are realized.
In the exemplary embodiment, a kind of non-transitorycomputer readable storage medium including instruction, example are additionally provided
It such as include the memory 904 of instruction, above-metioned instruction can be executed by the processor 920 of device 900, to realize FIG. 1 to FIG. 4 institute diagram
The step of methods of marking of image quality amount.For example, the non-transitorycomputer readable storage medium can be ROM, arbitrary access
Memory (RAM), CD-ROM, tape, floppy disk and optical data storage devices etc..
Those skilled in the art will readily occur to its of the disclosure after considering specification and practicing disclosure disclosed herein
Its embodiment.This application is intended to cover any variations, uses, or adaptations of the disclosure, these modifications, purposes or
Person's adaptive change follows the general principles of this disclosure and including the undocumented common knowledge in the art of the disclosure
Or conventional techniques.The description and examples are only to be considered as illustrative, and the true scope and spirit of the disclosure are by following
Claim is pointed out.
It should be understood that the present disclosure is not limited to the precise structures that have been described above and shown in the drawings, and
And various modifications and changes may be made without departing from the scope thereof.The scope of the present disclosure is only limited by the accompanying claims.
Claims (10)
1. a kind of methods of marking of picture quality, which is characterized in that the method includes：
Determine input picture；
Determine probability distribution of the input picture in scoring range；
The score value of the input picture is determined according to the probability distribution.
2. methods of marking according to claim 1, which is characterized in that determine that the input picture is general in scoring range
Rate is distributed：
The input picture is pre-processed, standard picture is obtained；
The standard picture is inputted into the Feature Selection Model, the standard picture is determined by the Feature Selection Model
Feature vector；
Described eigenvector is inputted into multi-layer perception (MLP), determines the input picture in the scoring by the multi-layer perception (MLP)
Probability distribution in range.
3. methods of marking according to claim 2, which is characterized in that determine the standard by the Feature Selection Model
Before the feature vector of image, the method also includes：
Obtain N sample images and the respective artificial Marking Probability distribution of the N sample images；N is positive integer；
The N sample images are sequentially input into multi-layer perception (MLP), until when the loss function in the multi-layer perception (MLP) is restrained
Multi-layer perception (MLP) described in deconditioning, in the multi-layer perception (MLP) after being trained；The loss function is based on the artificial mark
Probability distribution and the probability distribution of multi-layer perception (MLP) output judge whether to restrain.
4. methods of marking according to claim 1, which is characterized in that determine the input picture according to the probability distribution
Score value include：
Probability value of the input picture in the scoring range at each score value is determined according to the probability distribution；
The score value of the input picture is determined based on each score value and the probability value.
5. a kind of scoring apparatus of picture quality, which is characterized in that described device includes：
Input picture determining module, for determining input picture；
Probability distribution determining module, for determining probability distribution of the input picture in scoring range；
Score value determining module, for determining the score value of the input picture according to the probability distribution.
6. scoring apparatus according to claim 5, which is characterized in that the probability distribution determining module includes：
Standard picture acquiring unit obtains standard picture for pre-processing to the input picture；
Feature vector determination unit passes through the feature extraction for the standard picture to be inputted the Feature Selection Model
Model determines the feature vector of the standard picture；
Probability distribution determination unit is determined for described eigenvector to be inputted multi-layer perception (MLP) by the multi-layer perception (MLP)
Probability distribution of the input picture in the scoring range.
7. scoring apparatus according to claim 6, which is characterized in that described device further includes：
Probability distribution obtains module, for obtaining N sample images and the N respective artificial Marking Probabilities of sample image point
Cloth；N is positive integer；
Characteristic model training module, for the N sample images to be sequentially input multi-layer perception (MLP), until the Multilayer Perception
Multi-layer perception (MLP) described in deconditioning when loss function convergence in machine, in the multi-layer perception (MLP) after being trained；The loss
Function judges whether to restrain based on the probability distribution of the artificial Marking Probability distribution and multi-layer perception (MLP) output.
8. scoring apparatus according to claim 5, which is characterized in that the score value determining module includes：
Probability value determination unit, for determining that the input picture respectively scores in the scoring range according to the probability distribution
Probability value at value；
Score value determination unit, for determining the score value of the input picture based on each score value and the probability value.
9. a kind of electronic equipment, which is characterized in that the electronic equipment includes：
Processor；
For storing the memory of the processor-executable instruction；
Wherein, the processor is configured to executing the executable instruction in the memory to realize that Claims 1 to 4 is any
The step of item the method.
10. a kind of readable storage medium storing program for executing, is stored thereon with computer program, which is characterized in that when the program is executed by processor
The step of realizing any one of Claims 1 to 4 the method.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810654626.2A CN108898591A (en) | 2018-06-22 | 2018-06-22 | Methods of marking and device, electronic equipment, the readable storage medium storing program for executing of picture quality |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810654626.2A CN108898591A (en) | 2018-06-22 | 2018-06-22 | Methods of marking and device, electronic equipment, the readable storage medium storing program for executing of picture quality |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108898591A true CN108898591A (en) | 2018-11-27 |
Family
ID=64345632
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810654626.2A Pending CN108898591A (en) | 2018-06-22 | 2018-06-22 | Methods of marking and device, electronic equipment, the readable storage medium storing program for executing of picture quality |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108898591A (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109639973A (en) * | 2018-12-21 | 2019-04-16 | 中国科学院自动化研究所南京人工智能芯片创新研究院 | Shoot image methods of marking, scoring apparatus, electronic equipment and storage medium |
| CN110505398A (en) * | 2019-07-16 | 2019-11-26 | 北京三快在线科技有限公司 | A kind of image processing method, device, electronic equipment and storage medium |
| CN110610479A (en) * | 2019-07-31 | 2019-12-24 | 华为技术有限公司 | Object scoring method and device |
| CN111179382A (en) * | 2020-01-02 | 2020-05-19 | 广东博智林机器人有限公司 | Image typesetting method, device, medium and electronic equipment |
| CN111797258A (en) * | 2020-07-10 | 2020-10-20 | 携程计算机技术（上海）有限公司 | Image pushing method, system, equipment and storage medium based on aesthetic evaluation |
| CN113408528A (en) * | 2021-06-24 | 2021-09-17 | 数贸科技（北京）有限公司 | Commodity image quality identification method and device, computing equipment and storage medium |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106155593A (en) * | 2016-08-01 | 2016-11-23 | 惠州Tcl移动通信有限公司 | A kind of method and system deleting photo based on shooting quality |
| CN106570110A (en) * | 2016-10-25 | 2017-04-19 | 北京小米移动软件有限公司 | De-overlapping processing method and apparatus of image |
| CN107203558A (en) * | 2016-03-17 | 2017-09-26 | 腾讯科技（深圳）有限公司 | Object recommendation method and apparatus, recommendation information treating method and apparatus |
| CN107545309A (en) * | 2016-06-23 | 2018-01-05 | 西门子保健有限责任公司 | Scored using the picture quality of depth generation machine learning model |
| CN107610123A (en) * | 2017-10-11 | 2018-01-19 | 中共中央办公厅电子科技学院 | A kind of image aesthetic quality evaluation method based on depth convolutional neural networks |
| CN107680077A (en) * | 2017-08-29 | 2018-02-09 | 南京航空航天大学 | A kind of non-reference picture quality appraisement method based on multistage Gradient Features |
-
2018
- 2018-06-22 CN CN201810654626.2A patent/CN108898591A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107203558A (en) * | 2016-03-17 | 2017-09-26 | 腾讯科技（深圳）有限公司 | Object recommendation method and apparatus, recommendation information treating method and apparatus |
| CN107545309A (en) * | 2016-06-23 | 2018-01-05 | 西门子保健有限责任公司 | Scored using the picture quality of depth generation machine learning model |
| CN106155593A (en) * | 2016-08-01 | 2016-11-23 | 惠州Tcl移动通信有限公司 | A kind of method and system deleting photo based on shooting quality |
| CN106570110A (en) * | 2016-10-25 | 2017-04-19 | 北京小米移动软件有限公司 | De-overlapping processing method and apparatus of image |
| CN107680077A (en) * | 2017-08-29 | 2018-02-09 | 南京航空航天大学 | A kind of non-reference picture quality appraisement method based on multistage Gradient Features |
| CN107610123A (en) * | 2017-10-11 | 2018-01-19 | 中共中央办公厅电子科技学院 | A kind of image aesthetic quality evaluation method based on depth convolutional neural networks |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109639973A (en) * | 2018-12-21 | 2019-04-16 | 中国科学院自动化研究所南京人工智能芯片创新研究院 | Shoot image methods of marking, scoring apparatus, electronic equipment and storage medium |
| CN110505398A (en) * | 2019-07-16 | 2019-11-26 | 北京三快在线科技有限公司 | A kind of image processing method, device, electronic equipment and storage medium |
| CN110610479A (en) * | 2019-07-31 | 2019-12-24 | 华为技术有限公司 | Object scoring method and device |
| CN110610479B (en) * | 2019-07-31 | 2024-05-03 | 花瓣云科技有限公司 | Object scoring method and device |
| CN111179382A (en) * | 2020-01-02 | 2020-05-19 | 广东博智林机器人有限公司 | Image typesetting method, device, medium and electronic equipment |
| CN111797258A (en) * | 2020-07-10 | 2020-10-20 | 携程计算机技术（上海）有限公司 | Image pushing method, system, equipment and storage medium based on aesthetic evaluation |
| CN111797258B (en) * | 2020-07-10 | 2023-11-24 | 携程计算机技术（上海）有限公司 | Image pushing method, system, equipment and storage medium based on aesthetic evaluation |
| CN113408528A (en) * | 2021-06-24 | 2021-09-17 | 数贸科技（北京）有限公司 | Commodity image quality identification method and device, computing equipment and storage medium |
| CN113408528B (en) * | 2021-06-24 | 2024-02-23 | 数贸科技（北京）有限公司 | Quality recognition method and device for commodity image, computing equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108898591A (en) | Methods of marking and device, electronic equipment, the readable storage medium storing program for executing of picture quality | |
| KR101906827B1 (en) | Apparatus and method for taking a picture continously | |
| CN104700353B (en) | Image filters generation method and device | |
| CN105184313B (en) | Disaggregated model construction method and device | |
| CN110188236A (en) | A kind of recommended method of music, apparatus and system | |
| CN104125396A (en) | Image shooting method and device | |
| CN105468767B (en) | Card information acquisition methods and device | |
| CN107239535A (en) | Similar pictures search method and device | |
| CN104580886B (en) | Filming control method and device | |
| CN107193983A (en) | Image search method and device | |
| CN105357425B (en) | Image capturing method and device | |
| CN107766820A (en) | Image classification method and device | |
| CN108921178B (en) | Method and device for obtaining image blur degree classification and electronic equipment | |
| CN105528078B (en) | The method and device of controlling electronic devices | |
| CN109145679A (en) | A kind of method, apparatus and system issuing warning information | |
| CN106250921A (en) | Image processing method and device | |
| CN109819288A (en) | Determination method, apparatus, electronic equipment and the storage medium of advertisement dispensing video | |
| CN108108671A (en) | Description of product information acquisition method and device | |
| CN108462833A (en) | Image pickup method, device and computer readable storage medium | |
| CN105139033A (en) | Classifier construction method and device and image processing method and device | |
| CN106303198A (en) | Photographing information acquisition methods and device | |
| CN105335714A (en) | Photograph processing method, device and apparatus | |
| CN109145878B (en) | Image extraction method and device | |
| CN107895041A (en) | Screening-mode method to set up, device and storage medium | |
| CN110297929A (en) | Image matching method, device, electronic equipment and storage medium |
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 |