CN108198177A - Image acquiring method, device, terminal and storage medium - Google Patents
Image acquiring method, device, terminal and storage medium Download PDFInfo
- Publication number
- CN108198177A CN108198177A CN201711484291.6A CN201711484291A CN108198177A CN 108198177 A CN108198177 A CN 108198177A CN 201711484291 A CN201711484291 A CN 201711484291A CN 108198177 A CN108198177 A CN 108198177A
- Authority
- CN
- China
- Prior art keywords
- image
- picture frame
- frame
- picture
- video data
- 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
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N23/00—Cameras or camera modules comprising electronic image sensors; Control thereof
- H04N23/80—Camera processing pipelines; Components thereof
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N5/00—Details of television systems
- H04N5/222—Studio circuitry; Studio devices; Studio equipment
- H04N5/262—Studio circuits, e.g. for mixing, switching-over, change of character of image, other special effects ; Cameras specially adapted for the electronic generation of special effects
- H04N5/265—Mixing
-
- 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
- G06T2207/10—Image acquisition modality
- G06T2207/10016—Video; Image sequence
-
- 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
- G06T2207/10—Image acquisition modality
- G06T2207/10024—Color image
-
- 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
- G06T2207/20—Special algorithmic details
- G06T2207/20081—Training; Learning
-
- 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
- G06T2207/20—Special algorithmic details
- G06T2207/20084—Artificial neural networks [ANN]
-
- 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
-
- 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
- G06T2207/30196—Human being; Person
- G06T2207/30201—Face
Landscapes
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Signal Processing (AREA)
- Quality & Reliability (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Image Analysis (AREA)
Abstract
This application discloses a kind of image acquiring method, device, terminal and storage mediums, belong to image processing field.The method includes：Obtain the target video data of shooting；By the picture frame of the target video data, image Rating Model trained in advance is inputted respectively, obtains the corresponding scoring of each picture frame；According to highest first picture frame that scores in the picture frame of the target video data, picture of taking pictures is determined.The application from target video data by filtering out the highest picture frame of scoring, and picture of taking pictures is determined according to the picture frame, due to without user's manual screening, server can select the highest picture frame of scoring from target video data automatically, therefore it solves the problems, such as that the self-timer efficiency of user is low, has achieved the effect that improve the self-timer efficiency of user.
Description
Technical field
The invention relates to image processing field, more particularly to a kind of image acquiring method, device, terminal and storage
Medium.
Background technology
As mobile phone, tablet computer etc. have the terminal rapid proliferation of camera, self-timer has become one and extremely lives
The thing of change.
In actual scene, the gimmick of taking pictures of user is incorrect, and the photo effect for leading to self-timer is poor.For example self-timer is too near
Seem and be bold, for self-timer angle not to causing to seem that very face is fat to people, photo daylighting is excessively bright or excessively dark etc..
Once the photo of self-timer is poor, user then needs self-timer again, this undoubtedly reduces the self-timer efficiency of user.
Invention content
The embodiment of the present application provides a kind of image acquiring method, device, terminal and storage medium, can be used for solving to use
The problem of self-timer efficiency at family is low.The technical solution is as follows：
In a first aspect, a kind of image acquiring method is provided, the method includes：
Obtain the target video data of shooting；
By the picture frame of the target video data, image Rating Model trained in advance is inputted respectively, obtains each figure
As the corresponding scoring of frame；
According to highest first picture frame that scores in the picture frame of the target video data, picture of taking pictures is determined.
Second aspect, provides a kind of image acquiring device, and described device includes：
First acquisition module, for obtaining the target video data of shooting；
Input module, for by the picture frame of the target video data, inputting image scoring mould trained in advance respectively
Type obtains the corresponding scoring of each picture frame；
Determining module for highest first picture frame that scores in the picture frame according to the target video data, determines
It takes pictures picture.
The third aspect, provides a kind of terminal, and the terminal includes processor, memory, is stored in the memory
At least one instruction, described instruction are loaded by the processor and are performed to realize image acquisition side as described in relation to the first aspect
Method.
Fourth aspect provides a kind of computer readable storage medium, at least one finger is stored in the storage medium
It enables, described instruction is loaded by processor and performed to realize image acquiring method as described in relation to the first aspect.
The advantageous effect that technical solution provided by the embodiments of the present application is brought is：
By filtering out the highest picture frame of scoring from target video data, and figure of taking pictures is determined according to the picture frame
Piece, since without user's manual screening, server can select the highest picture frame of scoring from target video data automatically, because
This solves the problems, such as that the self-timer efficiency of user is low, has achieved the effect that improve the self-timer efficiency of user.
Description of the drawings
In order to illustrate more clearly of the technical solution in the embodiment of the present application, make required in being described below to embodiment
Attached drawing is briefly described, it should be apparent that, the accompanying drawings in the following description is only some embodiments of the present application, for
For those of ordinary skill in the art, without creative efforts, other are can also be obtained according to these attached drawings
Attached drawing.
Fig. 1 is the structure diagram of server 100 that one exemplary embodiment of the application is provided；
Fig. 2A is the flow chart of image acquiring method that one exemplary embodiment of the application provides；
Fig. 2 B are the flow charts for the training process for applying for the image Rating Model that an exemplary embodiment provides；
Fig. 3 A are the flow charts of the image acquiring method of the application another exemplary embodiment offer；
Fig. 3 B are the flow charts of the training process of image classification model that one exemplary embodiment of the application provides；
Fig. 4 is the flow chart for the image acquiring method that the application further exemplary embodiment provides；
Fig. 5 is the structure diagram of image acquiring device that one exemplary embodiment of the application provides.
Specific embodiment
Purpose, technical scheme and advantage to make the application are clearer, below in conjunction with attached drawing to the application embodiment party
Formula is described in further detail.
In the following description when referring to the accompanying drawings, unless otherwise indicated, the same numbers in different attached drawings represent same or similar
Element.Embodiment described in following exemplary embodiment does not represent all embodiment party consistent with the application
Formula.On the contrary, they are only the device consistent with some aspects being described in detail in such as the appended claims, the application and side
The example of method.
In the description of the present invention, it is to be understood that term " first ", " second " etc. are only used for description purpose, without
It is understood that indicate or implying relative importance.In the description of the present invention, it should be noted that unless otherwise specific regulation
And restriction, term " connected ", " connection " should be interpreted broadly, for example, it may be fixedly connected or be detachably connected,
Or it is integrally connected；Can be mechanical connection or electrical connection；It can be directly connected, intermediary can also be passed through
It is indirectly connected.For the ordinary skill in the art, the tool of above-mentioned term in the present invention can be understood with concrete condition
Body meaning.In addition, in the description of the present invention, unless otherwise indicated, " multiple " refer to two or more."and/or" is retouched
The incidence relation of affiliated partner is stated, expression may have three kinds of relationships, for example, A and/or B, can represent：Individualism A, together
When there are A and B, these three situations of individualism B.It is a kind of relationship of "or" that character "/", which typicallys represent forward-backward correlation object,.
First, to this application involves to noun be introduced.
Image Rating Model：It is a kind of mathematical model for the quality score for being used to determine image according to the input data.
Optionally, picture appraisal model includes but not limited to：Convolutional neural networks (Convolutional Neural
Network, CNN) model, deep neural network (Deep Neural Network, DNN) model, Recognition with Recurrent Neural Network
(Recurrent Neural Networks, RNN) model, insertion (embedding) model, gradient promote decision tree
(Gradient Boosting Decision Tree, GBDT) model, logistic regression (Logistic Regression, LR) mould
At least one of type.
DNN models are a kind of deep learning frames.DNN models include input layer, at least one layer of hidden layer (or middle layer)
And output layer.Optionally, input layer, at least one layer of hidden layer (or middle layer) and output layer include at least one neuron,
Neuron is used to handle the data received.Optionally, the quantity of the neuron between different layers can be identical；Or
Person can also be different.
RNN models are a kind of neural networks with feedback arrangement.In RNN models, the output of neuron can be under
One timestamp is applied directly to itself, that is, input of the i-th layer of neuron at the m moment, in addition to (i-1) layer neuron this when
Outside the output at quarter, its own output at (m-1) moment is further included.
Embedding models are shown based on entity and relationship distribution vector table, by the relationship in each triple example
Regard the translation from entity head to entity tail as.Wherein, triple example includes main body, relationship, object, and triple example can be with table
It is shown as (main body, relationship, object)；Main body is entity head, and object is entity tail.Such as：The father of Xiao Zhang is big, then passes through three
Tuple example is expressed as (Xiao Zhang, father are big to open).
GBDT models are a kind of decision Tree algorithms of iteration, which is made of more decision trees, and the result of all trees is tired out
It adds up as final result.Each node of decision tree can obtain a predicted value, and by taking the age as an example, predicted value is belongs to
The average value at owner's age of age corresponding node.
LR models refer on the basis of linear regression, apply mechanically the model that a logical function is established.
In actual scene, the gimmick of taking pictures of user is incorrect, and the photo effect for leading to self-timer is poor.Once the photograph of self-timer
Piece is poor, and user then needs self-timer again, this undoubtedly reduces the self-timer efficiency of user.For this purpose, this application provides a kind of figures
As acquisition methods, device, terminal and storage medium, to solve the problems, such as above-mentioned the relevant technologies.The skill that the application provides
In art scheme, by filtering out the highest picture frame of scoring from target video data, and figure of taking pictures is determined according to the picture frame
Piece, since without user's manual screening, server can select the highest picture frame of scoring from target video data automatically, because
This improves the self-timer efficiency of user, is illustrated below using schematical embodiment.
Before the embodiment of the present application is explained, first the application scenarios of the embodiment of the present application are illustrated.
Fig. 1 shows the structure diagram of server 100 that one exemplary embodiment of the application is provided.
Image Rating Model is stored in server 100.Optionally, which is using sample image pair
The model that CNN is trained.
Optionally, which includes one or more such as lower component：Processor 110 and memory 120.
Processor 110 can include one or more processing core.Processor 110 utilizes various interfaces and connection
Various pieces in entire elevator dispatching equipment, by running or performing the instruction being stored in memory 120, program, code
Collection or instruction set and calling are stored in the data in memory 120, perform the various functions of elevator dispatching equipment and processing number
According to.Optionally, processor 110 Digital Signal Processing (Digital Signal Processing, DSP) may be used, scene can
Program gate array (Field-Programmable Gate Array, FPGA), programmable logic array (Programmable
Logic Array, PLA) at least one of example, in hardware realize.Processor 110 can integrating central processor (Central
Processing Unit, CPU) and one or more of modem etc. combination.Wherein, the main processing operation systems of CPU
System and application program etc.；Modem wirelessly communicates for handling.It is understood that above-mentioned modem can not also
It is integrated into processor 110, is realized separately through chip piece.
Optionally, realize that following each embodiment of the method provides when processor 110 performs the program instruction in memory 120
Image acquiring method.
Memory 120 can include random access memory (Random Access Memory, RAM), can also include read-only
Memory (Read-Only Memory).Optionally, which includes non-transient computer-readable medium (non-
transitory computer-readable storage medium).Memory 120 can be used for store instruction, program, generation
Code, code set or instruction set.Memory 120 may include storing program area and storage data field, wherein, storing program area can store
Be used to implement operating system instruction, at least one function instruction, be used to implement the finger of following each embodiment of the method
Enable etc.；Storage data field can store data arrived involved in following each embodiment of the method etc..
A is please referred to Fig.2, it illustrates the flow charts of image acquiring method that one exemplary embodiment of the application provides.
The present embodiment is illustrated with the image acquiring method applied to implementation environment shown in figure 1.The image acquiring method packet
It includes：
Step 201, the target video data of shooting is obtained.
Optionally, server obtains the target video data for including target object that terminal is sent.Corresponding, which regards
Frequency evidence is video data of the terminal by preposition photography.Wherein, the image category of target object include personage, animal,
At least one of still life and landscape.
Step 202, by the picture frame of target video data, image Rating Model trained in advance is inputted respectively, is obtained every
The corresponding scoring of a picture frame.
Image Rating Model is using sample image and the model being trained with sample image scoring to CNN, is used
In the quality score for calculating image.
Optionally, image Rating Model is stored in server, which is according at least one trained sample
What this training obtained, each training sample includes：Sample image and sample image scoring.
Wherein, the training process of image Rating Model can refer to the associated description in the following examples, not be situated between first herein
It continues.
Corresponding score of each picture frame is used to indicate the picture quality of the image, picture quality include image fidelity and
Image intelligibility.Wherein, image fidelity is the departure degree between the image and real image of extraction, and image intelligibility is behaved
Or machine is drawn into the degree of characteristic information from image.
Optionally, quality score is used to indicate the picture quality of image, that is, is used to indicate composition ratio, the color pair of image
Than degree, color saturation and light and shade contrast.For example, the quality score of image is higher, then it represents that the picture quality of the image is got over
Good, i.e., the effect corresponding to the composition ratio of image, color contrast, color saturation and light and shade contrast is better.
Step 203, according to highest first picture frame that scores in the picture frame of target video data, picture of taking pictures is determined.
Since quality score is used to indicate the picture quality of image, the highest picture frame that scores is target video data
In picture frame, the best picture frame of picture quality, therefore, the determining picture of taking pictures determined according to the first picture frame, compared to it
The picture quality of determining picture of taking pictures that his picture frame is determined is more preferable.
In conclusion the embodiment of the present application from target video data by filtering out the highest picture frame of scoring, and root
Picture of taking pictures is determined according to the picture frame, since without user's manual screening, server can be chosen from target video data automatically
Go out the highest picture frame that scores, therefore solve the problems, such as that the self-timer efficiency of user is low, reach the self-timer effect for improving user
The effect of rate.
It should be noted that before step 201, server needs that image is trained to obtain image Rating Model.
B is please referred to Fig.2, it illustrates the training process of image Rating Model that one exemplary embodiment of the application provides
Flow chart.Optionally, the training process of image Rating Model includes but not limited to following steps：
Step 204, multiple training samples are obtained.
Wherein, each training sample includes sample image and sample image scores.
Optionally, training sample is obtained from terminal or is obtained from other servers, and sample image scores by artificially determining.
Step 205, it is inputted sample image as training, sample image scoring is as output reference value, to initial pictures
Disaggregated model is trained, the image Rating Model after being trained.
It scores for the sample image at least one training sample and sample image, sample graph is extracted from sample image
As feature, sample image feature is inputted into initial pictures disaggregated model, obtains training result.
Optionally, server extracts sample image feature according to pre-set image Processing Algorithm from sample image.Wherein,
Pre-set image Processing Algorithm is perceives hash algorithm (Perceptual hash algorithm, pHash algorithm).Server leads to
It crosses pHash algorithms and calculates the corresponding perception cryptographic Hash of sample image, the perception cryptographic Hash being calculated is determined as sample image
Feature.
Optionally, initial pictures disaggregated model be according to Establishment of Neural Model, such as：Initial pictures disaggregated model
It is a kind of foundation in CNN models, DNN models and RNN models.
Schematically, for each training sample, terminal creates the corresponding inputoutput pair of the training sample, input and output
To input parameter be the corresponding sample image feature of sample image in the training sample, output parameter is in the training sample
Sample image scores；Inputoutput pair is inputted initial pictures disaggregated model by server, obtains training result.
For example, sample image is characterized as " sample image feature 1 ", sample image scores as " sample image scoring 1 ", terminal
The inputoutput pair of establishment is：(sample image feature 1)->(sample image scoring 1)；Wherein, (sample image feature 1) is defeated
Enter parameter, (sample image scoring 1) is output parameter.
Optionally, inputoutput pair is represented by feature vector.
Image Rating Model is obtained based on above-mentioned training, please refers to Fig.3 A, another exemplary reality it illustrates the application
The flow chart of the image acquiring method of example offer is provided.The present embodiment is applied to implementation shown in figure 1 with the image acquiring method
Environment illustrates.The image acquiring method includes：
Step 301, the target video data of shooting is obtained.
Step 302, the second picture frame of predeterminated position in target video data is obtained, by the second picture frame input picture point
Class model obtains the image category of the second picture frame.
Since the image category of target object in target video data is included in personage, animal, still life and landscape at least
One kind, the quality score standard of the corresponding image of different image categories is different, therefore according to different image category selections not
Same image Rating Model, may be such that the quality score for the image being calculated based on image Rating Model is more accurate.
For the more brief target video data of shooting duration of video (such as video data in 10s), target video
Target object in data does not usually change, therefore a frame picture frame (the second picture frame) is extracted from target video data,
By the second picture frame input picture disaggregated model, the image category of the second picture frame is obtained, and by the image class of the second picture frame
Image category not as picture frames all in target video data.
It should be noted that predeterminated position refers to the second picture frame in the position of target video data, which can
For the 5th frame, the 10th frame, the 15th frame, the present embodiment does not limit the concrete numerical value of predeterminated position.
In a kind of mode in the cards, step 304 can be replaced by：According to preset video processnig algorithms, to obtaining
The target video data got is analyzed, and the type identification of target object in target video data is calculated, according to calculating
Obtained type identification determines to be used for unique mark image category with the corresponding image category of the type mark, the type mark.
Step 303, according to pre-stored image category and the correspondence of image Rating Model, the second picture frame is determined
The corresponding target image Rating Model of image type, each picture frame in target video data is inputted into target image respectively
Rating Model obtains the corresponding scoring of each picture frame.
Wherein, the correspondence of image category and image Rating Model is as shown in Table 1.In Table 1, image category is
When " personage ", corresponding image Rating Model is " image Rating Model 1 "；When image category is " animal ", corresponding image is commented
Sub-model is " image Rating Model 2 "；When image category is " still life ", corresponding image Rating Model is " image Rating Model
3”；When image category is " landscape ", corresponding image Rating Model is " image Rating Model 4 ".
Table one
| Image category | Image Rating Model |
| Personage | Image Rating Model 1 |
| Animal | Image Rating Model 2 |
| Still life | Image Rating Model 3 |
| Landscape | Image Rating Model 4 |
Schematically, the default correspondence provided based on table one, when the image category for the target object that terminal is got
During for " personage ", image Rating Model " image Rating Model 1 " corresponding with the image category " personage " of target object is obtained.
Step 304, according to highest first picture frame that scores in the picture frame of target video data, picture of taking pictures is determined.
It should be noted that step 301 is similar with step 201 in the present embodiment, step 304 is similar with step 203, step
301 and step 304 specific descriptions can refer to step 201 and step 203 respectively, details are not described herein.
In conclusion the embodiment of the present application from target video data by filtering out the highest picture frame of scoring, and root
Picture of taking pictures is determined according to the picture frame, since without user's manual screening, server can be chosen from target video data automatically
Go out the highest picture frame that scores, therefore solve the problems, such as that the self-timer efficiency of user is low, reach the self-timer effect for improving user
The effect of rate.
In the present embodiment, different image Rating Models is selected according to different image categories, may be such that and commented based on image
The quality score for the image that sub-model is calculated is more accurate.
It should be noted that before step 301, server needs that image is trained to obtain image classification model.
B is please referred to Fig.3, it illustrates the training process of image classification model that one exemplary embodiment of the application provides
Flow chart.Optionally, the training process of image Rating Model includes but not limited to following steps：
Step 305, multiple training samples are obtained.
Wherein, each training sample includes sample image and sample image classification.
Optionally, training sample is obtained from terminal or is obtained from other servers, and sample image classification is by artificially determining.
Step 306, it is inputted sample image as training, sample image classification is as output reference value, to initial pictures
Disaggregated model is trained, the image classification model after being trained.
For the sample image at least one training sample and sample image classification, sample graph is extracted from sample image
As feature, sample image feature is inputted into initial pictures disaggregated model, obtains training result.
Optionally, server extracts sample image feature according to pre-set image Processing Algorithm from sample image.Wherein,
Pre-set image Processing Algorithm is perceives hash algorithm (Perceptual hash algorithm, pHash algorithm).Server leads to
It crosses pHash algorithms and calculates the corresponding perception cryptographic Hash of sample image, the perception cryptographic Hash being calculated is determined as sample image
Feature.
Optionally, initial pictures disaggregated model be according to Establishment of Neural Model, such as：Initial pictures disaggregated model
It is a kind of foundation in CNN models, DNN models and RNN models.
Schematically, for each training sample, terminal creates the corresponding inputoutput pair of the training sample, input and output
To input parameter be the corresponding sample image feature of sample image in the training sample, output parameter is in the training sample
Sample image classification；Inputoutput pair is inputted initial pictures disaggregated model by server, obtains training result.
For example, sample image is characterized as " sample image feature 1 ", sample image classification is " sample image classification 1 ", terminal
The inputoutput pair of establishment is：(sample image feature 1)->(sample image classification 1)；Wherein, (sample image feature 1) is defeated
Enter parameter, (sample image classification 1) is output parameter.
Optionally, inputoutput pair is represented by feature vector.
It please refers to Fig.4, it illustrates the flow charts for the image acquiring method that the application further exemplary embodiment provides.
The present embodiment is illustrated with the image acquiring method applied to implementation environment shown in figure 1.The image acquiring method packet
It includes：
Step 401, the target video data of shooting is obtained.
Step 402, by the picture frame of target video data, image Rating Model trained in advance is inputted respectively, is obtained every
The corresponding scoring of a picture frame.
Step 403, highest first picture frame of scoring is determined.
Step 404, the first picture frame, the preceding m frames picture frame of the first picture frame and the rear n frames image of the first picture frame are extracted
Frame.
Since the corresponding scoring of picture frame is according to the comprehensive score of all pixels block in the picture frame, the figure is not represented
As each block of pixels is higher than the scoring of the block of pixels of same position in other picture frames in frame, for the figure for the picture that ensures to take pictures
Image quality amount after highest first picture frame of scoring is extracted, can extract the preceding m frames picture frame of first picture frame together
With the rear n frames picture frame of the first picture frame, preceding m frames picture frame and the first picture frame to the first picture frame, the first picture frame
N frames picture frame carries out image synthesis afterwards, obtains the image of taking pictures of high quality.
It should be noted that m and n are positive integer, m and n numerical value can be the same or different, and the present embodiment is simultaneously unlimited
Determine the concrete numerical value of m and n.
Since the quality of synthesized picture of taking pictures when two image differences are larger, can be influenced, therefore in order to ensure to close
Obtain taking pictures after the quality of picture, it is preferred that m 1, n 1.I.e. server is determining highest first picture frame that scores
Afterwards, the first picture frame, the preceding 1 frame picture frame of the first picture frame and the rear 1 frame picture frame of the first picture frame are extracted.
It is a kind of it is special in the case of, the first picture frame is first frame in target video data or is target video number
Last frame in, i.e., if server can not extract the preceding m frames picture frame of the first picture frame, not to the preceding m of the first picture frame
Frame picture frame extracts, and equally, server can not extract the rear n frames picture frame of the first picture frame, then not to the first picture frame
Rear n frames picture frame extract.
Step 405, the rear n frames picture frame of the preceding m frames picture frame to the first picture frame, the first picture frame and the first picture frame
Image synthesis is carried out, obtains picture of taking pictures.
Optionally, in the rear n frames picture frame of the first picture frame, the preceding m frames picture frame of the first picture frame and the first picture frame
In, by the highest block of pixels of clarity in the block of pixels of same position in different images frame, it is combined, obtains picture of taking pictures.
Optionally, in the rear n frames picture frame of the first picture frame, the preceding m frames picture frame of the first picture frame and the first picture frame
In, by the highest block of pixels of color saturation in the block of pixels of same position in different images frame, it is combined, obtains figure of taking pictures
Piece.
It should be noted that in the present embodiment step 401 to step 402 and step 201 to step 202 similar, step 401
Step 201 is can refer to step 202 to step 402 specific descriptions, and details are not described herein.
In conclusion the embodiment of the present application from target video data by filtering out the highest picture frame of scoring, and root
Picture of taking pictures is determined according to the picture frame, since without user's manual screening, server can be chosen from target video data automatically
Go out the highest picture frame that scores, therefore solve the problems, such as that the self-timer efficiency of user is low, reach the self-timer effect for improving user
The effect of rate.
In the present embodiment, the rear n frames figure of preceding m frames picture frame and the first picture frame to the first picture frame, the first picture frame
As frame progress image synthesis, so as to obtain the image of taking pictures of high quality.
Following is the application device embodiment, can be used for performing the application embodiment of the method.For the application device reality
The details not disclosed in example is applied, please refers to the application embodiment of the method.
Fig. 5 is please referred to, it illustrates the structural representations of image acquiring device that one exemplary embodiment of the application provides
Figure.The image acquiring device can be by special hardware circuit, alternatively, software and hardware is implemented in combination with as the terminal in Fig. 1
All or part of, which includes：First acquisition module 501, input module 502 and determining module 503.
First acquisition module 501, for obtaining the target video data of shooting；
Input module 502, for by the picture frame of target video data, inputting image scoring mould trained in advance respectively
Type obtains the corresponding scoring of each picture frame；
Determining module 503 for highest first picture frame that scores in the picture frame according to target video data, determines to clap
Photograph and picture.
In the alternative embodiment provided based on embodiment illustrated in fig. 5, the input module 502, including：First input
Unit and the second input unit.
First input unit, for obtaining the second picture frame of predeterminated position in target video data, by the second picture frame
Input picture disaggregated model obtains the image category of the second picture frame；
Second input unit for the correspondence according to pre-stored image category and image Rating Model, determines
The corresponding target image Rating Model of image type of second picture frame, each picture frame difference in target video data is defeated
Enter target image Rating Model, obtain the corresponding scoring of each picture frame.
In the alternative embodiment provided based on embodiment illustrated in fig. 5, which further includes：Second acquisition module and
First training module.
Second acquisition module, for obtaining multiple training samples, wherein, each training sample includes sample image and sample
Image category；
First training module, for being inputted sample image as training, sample image classification is right as output reference value
Initial pictures disaggregated model is trained, the image classification model after being trained.
In the alternative embodiment provided based on embodiment illustrated in fig. 5, the determining module 503, including：It determines single
Member, extraction unit and synthesis unit.
Determination unit, for determining highest first picture frame of scoring；
Extraction unit, for extracting the first picture frame, the preceding m frames picture frame of the first picture frame and the rear n of the first picture frame
Frame picture frame；
Synthesis unit, for the preceding m frames picture frame and the rear n frames of the first picture frame to the first picture frame, the first picture frame
Picture frame carries out image synthesis, obtains picture of taking pictures.
In the alternative embodiment provided based on embodiment illustrated in fig. 5, the synthesis unit, including：
First combination subelement, in the preceding m frames picture frame of the first picture frame, the first picture frame and the first picture frame
Afterwards in n frames picture frame, by the highest block of pixels of clarity in the block of pixels of same position in different images frame, it is combined, obtains
To picture of taking pictures；
Second combination subelement, in the preceding m frames picture frame of the first picture frame, the first picture frame and the first picture frame
Afterwards in n frames picture frame, by the highest block of pixels of color saturation in the block of pixels of same position in different images frame, group is carried out
It closes, obtains picture of taking pictures.
In the alternative embodiment provided based on embodiment illustrated in fig. 5, which further includes：Third acquisition module and
Second training module.
Third acquisition module, for obtaining multiple training samples, wherein, each training sample includes sample image and sample
Image scores；
Second training module, for being inputted sample image as training, sample image scoring is right as output reference value
Initial pictures disaggregated model is trained, the image Rating Model after being trained.
Correlative detail can be combined with reference to figure 2A to embodiment of the method shown in Fig. 4.Wherein, the first acquisition module 501 is also used
It is any other implicit or disclosed with the relevant function of receiving step in above method embodiment in realizing；Input module 502 is also
It is used to implement any other implicit or disclosed and relevant function of monitoring step in above method embodiment；Determining module 503
It is additionally operable to realize any other implicit or disclosed and relevant function of monitoring step in above method embodiment.
It should be noted that the device that above-described embodiment provides, when realizing its function, only with above-mentioned each function module
It divides and for example, in practical application, can be completed as needed and by above-mentioned function distribution by different function modules,
The internal structure of equipment is divided into different function modules, to complete all or part of function described above.In addition,
The apparatus and method embodiment that above-described embodiment provides belongs to same design, and specific implementation process refers to embodiment of the method, this
In repeat no more.
The application also provides a kind of computer-readable medium, is stored thereon with program instruction, and the program instruction is by processor
The image acquiring method that above-mentioned each embodiment of the method provides is realized during execution.
Present invention also provides it is a kind of comprising instruction computer program product, when run on a computer so that
Computer performs the image acquiring method described in above-mentioned each embodiment of the method.
Above-mentioned the embodiment of the present application serial number is for illustration only, does not represent the quality of embodiment.
One of ordinary skill in the art will appreciate that it is completely or partially walked in the document handling method of realization above-described embodiment
Suddenly it can be completed by hardware, relevant hardware can also be instructed to complete by program, the program can be stored in
In a kind of computer readable storage medium, storage medium mentioned above can be read-only memory, disk or CD etc..More than
Described is only the preferred embodiment of the application, all within spirit herein and principle not to limit the application, is made
Any modification, equivalent substitution, improvement and etc., should be included within the protection domain of the application.
Claims (9)
1. a kind of image acquiring method, which is characterized in that the method includes：
Obtain the target video data of shooting；
By the picture frame of the target video data, image Rating Model trained in advance is inputted respectively, obtains each picture frame
Corresponding scoring；
According to highest first picture frame that scores in the picture frame of the target video data, picture of taking pictures is determined.
2. according to the method described in claim 1, it is characterized in that, the picture frame by the target video data, difference
Input image Rating Model trained in advance, obtains the corresponding scoring of each picture frame, including：
The second picture frame of predeterminated position in the target video data is obtained, by the second picture frame input picture classification mould
Type obtains the image category of second picture frame；
According to pre-stored image category and the correspondence of image Rating Model, the image class of second picture frame is determined
Each picture frame in the target video data is inputted the target image by the corresponding target image Rating Model of type respectively
Rating Model obtains the corresponding scoring of each picture frame.
3. according to the method described in claim 2, it is characterized in that, the method further includes：
Multiple training samples are obtained, wherein, each training sample includes sample image and sample image classification；
It is inputted the sample image as training, the sample image classification classifies to initial pictures as output reference value
Model is trained, the described image disaggregated model after being trained.
4. it according to the method described in claim 1, it is characterized in that, is commented in the picture frame according to the target video data
Divide highest first picture frame, determine picture of taking pictures, including：
Determine highest first picture frame of scoring；
Extract described first image frame, the preceding m frames picture frame of described first image frame and the rear n frames image of described first image frame
Frame；
The rear n frames picture frame of preceding m frames picture frame and described first image frame to described first image frame, described first image frame
Image synthesis is carried out, obtains picture of taking pictures.
It is 5. according to the method described in claim 4, it is characterized in that, described to described first image frame, described first image frame
Preceding m frames picture frame and described first image frame rear n frames picture frame carry out image synthesis, obtain picture of taking pictures, including：
In the rear n frames picture frame of described first image frame, the preceding m frames picture frame of described first image frame and described first image frame
In, by the highest block of pixels of clarity in the block of pixels of same position in different images frame, it is combined, obtains picture of taking pictures；
Alternatively,
In the rear n frames picture frame of described first image frame, the preceding m frames picture frame of described first image frame and described first image frame
In, by the highest block of pixels of saturation degree in the block of pixels of same position in different images frame, it is combined, obtains picture of taking pictures.
6. according to the method any in claim 1-5, which is characterized in that the method further includes：
Multiple training samples are obtained, wherein, each training sample includes sample image and sample image scores；
It is inputted the sample image as training, the sample image scoring classifies to initial pictures as output reference value
Model is trained, the described image Rating Model after being trained.
7. a kind of image acquiring device, which is characterized in that described device includes：
First acquisition module, for obtaining the target video data of shooting；
Input module for by the picture frame of the target video data, inputting image Rating Model trained in advance respectively, obtains
To the corresponding scoring of each picture frame；
Determining module for highest first picture frame that scores in the picture frame according to the target video data, determines to take pictures
Picture.
8. a kind of terminal, which is characterized in that the terminal includes processor, memory, and at least one is stored in the memory
Item instructs, and described instruction is loaded by the processor and performed to realize the image acquisition side as described in claim 1 to 6 is any
Method.
9. a kind of computer readable storage medium, which is characterized in that at least one instruction is stored in the storage medium, it is described
Instruction is loaded by processor and is performed to realize the image acquiring method as described in claim 1 to 6 is any.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711484291.6A CN108198177A (en) | 2017-12-29 | 2017-12-29 | Image acquiring method, device, terminal and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711484291.6A CN108198177A (en) | 2017-12-29 | 2017-12-29 | Image acquiring method, device, terminal and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108198177A true CN108198177A (en) | 2018-06-22 |
Family
ID=62586727
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711484291.6A Pending CN108198177A (en) | 2017-12-29 | 2017-12-29 | Image acquiring method, device, terminal and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108198177A (en) |
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109409235A (en) * | 2018-09-27 | 2019-03-01 | Oppo广东移动通信有限公司 | Image-recognizing method and device, electronic equipment, computer readable storage medium |
| CN110390263A (en) * | 2019-06-17 | 2019-10-29 | 宁波江丰智能科技有限公司 | A kind of method of video image processing and system |
| CN110659614A (en) * | 2019-09-25 | 2020-01-07 | Oppo广东移动通信有限公司 | Video sampling method, device, equipment and storage medium |
| WO2020116983A1 (en) | 2018-12-07 | 2020-06-11 | Samsung Electronics Co., Ltd. | Electronic apparatus, controlling method of electronic apparatus, and computer readable medium |
| CN111935479A (en) * | 2020-07-30 | 2020-11-13 | 浙江大华技术股份有限公司 | Target image determination method and device, computer equipment and storage medium |
| CN112055847A (en) * | 2018-09-18 | 2020-12-08 | 谷歌有限责任公司 | Method and system for processing image |
| CN112188075A (en) * | 2019-07-05 | 2021-01-05 | 杭州海康威视数字技术股份有限公司 | Snapshot, image processing device and image processing method |
| CN112492297A (en) * | 2020-11-17 | 2021-03-12 | 腾讯科技（深圳）有限公司 | Video processing method and related equipment |
| CN112771612A (en) * | 2019-09-06 | 2021-05-07 | 华为技术有限公司 | Method and device for shooting image |
| CN113259601A (en) * | 2020-02-11 | 2021-08-13 | 北京字节跳动网络技术有限公司 | Video processing method and device, readable medium and electronic equipment |
| CN113365023A (en) * | 2021-04-22 | 2021-09-07 | 北京房江湖科技有限公司 | Holder visual angle correction method and device, electronic equipment and storage medium |
| CN115802146A (en) * | 2021-09-07 | 2023-03-14 | 荣耀终端有限公司 | Method for snapping image in video and electronic equipment |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20150071547A1 (en) * | 2013-09-09 | 2015-03-12 | Apple Inc. | Automated Selection Of Keeper Images From A Burst Photo Captured Set |
| CN105631457A (en) * | 2015-12-17 | 2016-06-01 | 小米科技有限责任公司 | Method and device for selecting picture |
| CN105913400A (en) * | 2016-05-03 | 2016-08-31 | 成都索贝数码科技股份有限公司 | Device for obtaining high-quality and real-time beautiful image |
| 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 |
| CN107153838A (en) * | 2017-04-19 | 2017-09-12 | 中国电子科技集团公司电子科学研究院 | A kind of photo automatic grading method and device |
| CN107230192A (en) * | 2017-05-31 | 2017-10-03 | 广东欧珀移动通信有限公司 | Image processing method, device, computer-readable recording medium and mobile terminal |
-
2017
- 2017-12-29 CN CN201711484291.6A patent/CN108198177A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20150071547A1 (en) * | 2013-09-09 | 2015-03-12 | Apple Inc. | Automated Selection Of Keeper Images From A Burst Photo Captured Set |
| CN105631457A (en) * | 2015-12-17 | 2016-06-01 | 小米科技有限责任公司 | Method and device for selecting picture |
| CN105913400A (en) * | 2016-05-03 | 2016-08-31 | 成都索贝数码科技股份有限公司 | Device for obtaining high-quality and real-time beautiful image |
| 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 |
| CN107153838A (en) * | 2017-04-19 | 2017-09-12 | 中国电子科技集团公司电子科学研究院 | A kind of photo automatic grading method and device |
| CN107230192A (en) * | 2017-05-31 | 2017-10-03 | 广东欧珀移动通信有限公司 | Image processing method, device, computer-readable recording medium and mobile terminal |
Cited By (19)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US11947591B2 (en) | 2018-09-18 | 2024-04-02 | Google Llc | Methods and systems for processing imagery |
| CN112055847B (en) * | 2018-09-18 | 2024-03-12 | 谷歌有限责任公司 | Method and system for processing image |
| CN112055847A (en) * | 2018-09-18 | 2020-12-08 | 谷歌有限责任公司 | Method and system for processing image |
| CN109409235A (en) * | 2018-09-27 | 2019-03-01 | Oppo广东移动通信有限公司 | Image-recognizing method and device, electronic equipment, computer readable storage medium |
| EP3821372A4 (en) * | 2018-12-07 | 2021-09-15 | Samsung Electronics Co., Ltd. | Electronic apparatus, controlling method of electronic apparatus, and computer readable medium |
| WO2020116983A1 (en) | 2018-12-07 | 2020-06-11 | Samsung Electronics Co., Ltd. | Electronic apparatus, controlling method of electronic apparatus, and computer readable medium |
| US11558545B2 (en) | 2018-12-07 | 2023-01-17 | Samsung Electronics Co., Ltd. | Electronic apparatus, controlling method of electronic apparatus, and computer readable medium |
| CN110390263A (en) * | 2019-06-17 | 2019-10-29 | 宁波江丰智能科技有限公司 | A kind of method of video image processing and system |
| CN112188075A (en) * | 2019-07-05 | 2021-01-05 | 杭州海康威视数字技术股份有限公司 | Snapshot, image processing device and image processing method |
| CN112188075B (en) * | 2019-07-05 | 2023-04-18 | 杭州海康威视数字技术股份有限公司 | Snapshot, image processing device and image processing method |
| CN112771612A (en) * | 2019-09-06 | 2021-05-07 | 华为技术有限公司 | Method and device for shooting image |
| CN110659614A (en) * | 2019-09-25 | 2020-01-07 | Oppo广东移动通信有限公司 | Video sampling method, device, equipment and storage medium |
| CN113259601A (en) * | 2020-02-11 | 2021-08-13 | 北京字节跳动网络技术有限公司 | Video processing method and device, readable medium and electronic equipment |
| US11996124B2 (en) | 2020-02-11 | 2024-05-28 | Beijing Bytedance Network Technology Co., Ltd. | Video processing method, apparatus, readable medium and electronic device |
| CN111935479A (en) * | 2020-07-30 | 2020-11-13 | 浙江大华技术股份有限公司 | Target image determination method and device, computer equipment and storage medium |
| CN112492297A (en) * | 2020-11-17 | 2021-03-12 | 腾讯科技（深圳）有限公司 | Video processing method and related equipment |
| CN113365023A (en) * | 2021-04-22 | 2021-09-07 | 北京房江湖科技有限公司 | Holder visual angle correction method and device, electronic equipment and storage medium |
| CN115802146A (en) * | 2021-09-07 | 2023-03-14 | 荣耀终端有限公司 | Method for snapping image in video and electronic equipment |
| CN115802146B (en) * | 2021-09-07 | 2024-04-02 | 荣耀终端有限公司 | Method for capturing images in video and electronic equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108198177A (en) | Image acquiring method, device, terminal and storage medium | |
| CN108234870B (en) | Image processing method, device, terminal and storage medium | |
| CN106778928B (en) | Image processing method and device | |
| CN103617432B (en) | A kind of scene recognition method and device | |
| CN111597938B (en) | Living body detection and model training method and device | |
| JP2017531950A (en) | Method and apparatus for constructing a shooting template database and providing shooting recommendation information | |
| CN108492294B (en) | Method and device for evaluating harmony degree of image colors | |
| CN110929569A (en) | Face recognition method, device, equipment and storage medium | |
| CN108810406B (en) | Portrait light effect processing method, device, terminal and computer readable storage medium | |
| CN111985281B (en) | Image generation model generation method and device and image generation method and device | |
| CN109559362B (en) | Image subject face replacing method and device | |
| CN109785400B (en) | Silhouette image manufacturing method and device, electronic equipment and storage medium | |
| CN107153838A (en) | A kind of photo automatic grading method and device | |
| CN111127309B (en) | Portrait style migration model training method, portrait style migration method and device | |
| CN107172354A (en) | Method for processing video frequency, device, electronic equipment and storage medium | |
| JP2021531571A (en) | Certificate image extraction method and terminal equipment | |
| CN108156385A (en) | Image acquiring method and image acquiring device | |
| CN108200337A (en) | It takes pictures method, apparatus, terminal and the storage medium of processing | |
| CN107943811A (en) | The dissemination method and device of content | |
| CN112633221A (en) | Face direction detection method and related device | |
| CN111080746A (en) | Image processing method, image processing device, electronic equipment and storage medium | |
| CN109829364A (en) | A kind of expression recognition method, device and recommended method, device | |
| CN106815803A (en) | The processing method and processing device of picture | |
| CN109636867B (en) | Image processing method and device and electronic equipment | |
| CN114360018A (en) | Rendering method and device of three-dimensional facial expression, storage medium and electronic device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20180622 |
|
| RJ01 | Rejection of invention patent application after publication |