CN108198177A - Image acquisition method, device, terminal and storage medium - Google Patents

Abstract

The application discloses an image acquisition method, an image acquisition device, a terminal and a storage medium, and belongs to the field of image processing. The method comprises the following steps: acquiring shot target video data; respectively inputting the image frames of the target video data into a pre-trained image scoring model to obtain a score corresponding to each image frame; and determining a photographed picture according to the first image frame with the highest score in the image frames of the target video data. According to the method and the device, the image frame with the highest score is screened from the target video data, the picture to be shot is determined according to the image frame, and the server can automatically select the image frame with the highest score from the target video data without manual screening by a user, so that the problem that the self-shooting efficiency of the user is low is solved, and the effect of improving the self-shooting efficiency of the user is achieved.

Description

Image acquisition method, device, terminal and storage medium

technical field

The embodiments of the present application relate to the field of image processing, and in particular to an image acquisition method, device, terminal, and storage medium.

Background technique

With the rapid popularization of terminals with cameras such as mobile phones and tablet computers, taking selfies has become an extremely daily thing.

In the actual scene, the user's camera technique is not correct, resulting in poor self-portrait photos. For example, if you take a selfie too close, your face will look big, if you take a selfie from the wrong angle, it will make your face appear fat, or if the light in the photo is too bright or too dark, etc.

Once the selfie photo is poor, the user needs to take a selfie again, which undoubtedly reduces the user's selfie efficiency.

Contents of the invention

Embodiments of the present application provide an image acquisition method, device, terminal, and storage medium, which can be used to solve the problem of low efficiency of a user's selfie. Described technical scheme is as follows:

In a first aspect, an image acquisition method is provided, the method comprising:

Obtain the captured target video data;

The image frames of the target video data are respectively input into the pre-trained image scoring model to obtain the corresponding scoring of each image frame;

A photographed picture is determined according to the first image frame with the highest score among the image frames of the target video data.

In a second aspect, an image acquisition device is provided, the device comprising:

The first obtaining module is used to obtain the target video data of shooting;

The input module is used to input the image frames of the target video data into the pre-trained image scoring model respectively to obtain the corresponding scoring of each image frame;

The determination module is configured to determine the photographed picture according to the first image frame with the highest score among the image frames of the target video data.

In a third aspect, a terminal is provided, the terminal includes a processor and a memory, at least one instruction is stored in the memory, and the instruction is loaded and executed by the processor to realize the image as described in the first aspect Get method.

In a fourth aspect, a computer-readable storage medium is provided, wherein at least one instruction is stored in the storage medium, and the instruction is loaded and executed by a processor to implement the image acquisition method as described in the first aspect.

The beneficial effects brought by the technical solutions provided by the embodiments of the present application are:

By filtering out the image frame with the highest score from the target video data, and determining the picture to be taken according to the image frame, the server will automatically select the image frame with the highest score from the target video data without manual filtering by the user, thus solving the problem of the user The problem of low self-timer efficiency has achieved the effect of improving the user's Selfie efficiency.

Description of drawings

In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings that need to be used in the description of the embodiments will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present application. For those skilled in the art, other drawings can also be obtained based on these drawings without creative effort.

FIG. 1 is a schematic structural diagram of a server 100 provided by an exemplary embodiment of the present application;

FIG. 2A is a flowchart of an image acquisition method provided by an exemplary embodiment of the present application;

Fig. 2B is a flow chart of the training process of the image scoring model provided by an exemplary embodiment of the application;

FIG. 3A is a flowchart of an image acquisition method provided by another exemplary embodiment of the present application;

Fig. 3B is a flowchart of the training process of the image classification model provided by an exemplary embodiment of the present application;

FIG. 4 is a flow chart of an image acquisition method provided in another exemplary embodiment of the present application;

Fig. 5 is a schematic structural diagram of an image acquisition device provided by an exemplary embodiment of the present application.

Detailed ways

In order to make the purpose, technical solution and advantages of the present application clearer, the implementation manners of the present application will be further described in detail below in conjunction with the accompanying drawings.

When the following description refers to the accompanying drawings, the same numerals in different drawings refer to the same or similar elements unless otherwise indicated. The implementations described in the following exemplary embodiments do not represent all implementations consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with aspects of the present application as recited in the appended claims.

In the description of the present invention, it should be understood that the terms "first", "second" and so on are used for descriptive purposes only, and cannot be interpreted as indicating or implying relative importance. In the description of the present invention, it should be noted that unless otherwise specified and limited, the terms "connected" and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral Ground connection; it can be mechanical connection or electrical connection; it can be direct connection or indirect connection through an intermediary. Those of ordinary skill in the art can understand the specific meanings of the above terms in the present invention in specific situations. In addition, in the description of the present invention, unless otherwise specified, "plurality" means two or more. "And/or" describes the association relationship of associated objects, indicating that there may be three types of relationships, for example, A and/or B may indicate: A exists alone, A and B exist simultaneously, and B exists independently. The character "/" generally indicates that the contextual objects are an "or" relationship.

First, the nouns involved in this application are introduced.

Image scoring model: is a mathematical model used to determine the quality score of an image based on input data.

Optionally, the image evaluation model includes but is not limited to: convolutional neural network (Convolutional Neural Network, CNN) model, deep neural network (Deep Neural Network, DNN) model, recurrent neural network (Recurrent Neural Networks, RNN) model, embedding ( embedding) model, gradient boosting decision tree (Gradient Boosting Decision Tree, GBDT) model, logistic regression (Logistic Regression, LR) model at least one.

The DNN model is a deep learning framework. The DNN model includes an input layer, at least one hidden layer (or middle layer) and an output layer. Optionally, the input layer, at least one hidden layer (or middle layer) and the output layer all include at least one neuron, and the neuron is used to process the received data. Optionally, the number of neurons in different layers may be the same; or may also be different.

The RNN model is a neural network with a feedback structure. In the RNN model, the output of a neuron can directly act on itself at the next time stamp, that is, the input of the i-th layer neuron at the m moment, in addition to the output of the (i-1) layer neuron at this moment, also includes Its own output at (m-1) time.

The embedding model is based on a distributed vector representation of entities and relationships, and regards the relationship in each triplet instance as a translation from the entity head to the entity tail. Wherein, the triplet instance includes subject, relation, and object, and the triplet instance can be expressed as (subject, relation, object); the subject is the entity head, and the object is the entity tail. For example: Xiao Zhang's father is Da Zhang, then it is expressed as (Xiao Zhang, father, Da Zhang) through a triplet instance.

The GBDT model is an iterative decision tree algorithm, which consists of multiple decision trees, and the results of all trees are added up as the final result. Each node of the decision tree will get a predicted value. Taking age as an example, the predicted value is the average age of all people belonging to the node corresponding to the age.

The LR model refers to a model established by applying a logistic function on the basis of linear regression.

In the actual scene, the user's camera technique is not correct, resulting in poor self-portrait photos. Once the selfie photo is poor, the user needs to take a selfie again, which undoubtedly reduces the user's selfie efficiency. To this end, the present application provides an image acquisition method, device, terminal and storage medium to solve the problems existing in the above-mentioned related technologies. In the technical solution provided by this application, by selecting the image frame with the highest score from the target video data, and determining the photographed picture according to the image frame, since the user does not need to manually filter, the server will automatically select the highest score from the target video data image frame, thus improving the Selfie efficiency of the user, which will be described below using a schematic embodiment.

Before explaining the embodiment of the present application, the application scenario of the embodiment of the present application will be described first. Fig. 1 shows a schematic structural diagram of a server 100 provided by an exemplary embodiment of the present application.

An image scoring model is stored in the server 100 . Optionally, the image scoring model is a model obtained by using sample images to train CNN.

Optionally, the server 100 includes one or more of the following components: a processor 110 and a memory 120 .

Processor 110 may include one or more processing cores. The processor 110 uses various interfaces and lines to connect various parts of the entire elevator dispatching equipment, and executes or executes instructions, programs, code sets or instruction sets stored in the memory 120, and calls data stored in the memory 120 to execute Various functions and processing data of elevator dispatching equipment. Optionally, the processor 110 may use at least one of Digital Signal Processing (Digital Signal Processing, DSP), Field-Programmable Gate Array (Field-Programmable Gate Array, FPGA), and Programmable Logic Array (ProgrammableLogic Array, PLA). implemented in the form of hardware. The processor 110 may integrate one or a combination of a central processing unit (Central Processing Unit, CPU), a modem, and the like. Among them, the CPU mainly deals with the operating system and application programs, etc.; the modem is used to deal with wireless communication. It can be understood that, the above-mentioned modem may not be integrated into the processor 110, but implemented by a single chip.

Optionally, when the processor 110 executes the program instructions in the memory 120, the image acquisition methods provided in the following method embodiments are implemented.

The memory 120 may include a random access memory (Random Access Memory, RAM), and may also include a read-only memory (Read-Only Memory). Optionally, the storage 120 includes a non-transitory computer-readable storage medium (non-transitory computer-readable storage medium). The memory 120 may be used to store instructions, programs, codes, sets of codes, or sets of instructions. The memory 120 may include a program storage area and a data storage area, wherein the program storage area may store instructions for implementing an operating system, instructions for at least one function, instructions for realizing the following various method embodiments, etc.; the storage data area Data and the like involved in the following method embodiments may be stored.

Please refer to FIG. 2A , which shows a flowchart of an image acquisition method provided by an exemplary embodiment of the present application. This embodiment is described by taking the image acquisition method applied to the implementation environment shown in FIG. 1 as an example. The image acquisition methods include:

Step 201, acquire the captured target video data.

Optionally, the server acquires the target video data including the target object sent by the terminal. Correspondingly, the target video data is video data captured by the terminal through a front camera. Wherein, the image category of the target object includes at least one of person, animal, still life and landscape.

In step 202, the image frames of the target video data are respectively input into the pre-trained image scoring model to obtain the corresponding score of each image frame.

The image scoring model is a model obtained by training CNN with sample images and sample image scores, and is used to calculate the quality score of the image.

Optionally, an image scoring model is stored in the server, and the image scoring model is trained according to at least one training sample, and each training sample includes: a sample image and a score of the sample image.

For the training process of the image scoring model, reference may be made to the relevant descriptions in the following embodiments, which will not be introduced here.

The score corresponding to each image frame is used to indicate the image quality of the image, and the image quality includes image fidelity and image intelligibility. Among them, image fidelity is the degree of deviation between the extracted image and the actual image, and image intelligibility is the degree to which people or machines can extract feature information from images.

Optionally, the quality score is used to indicate the image quality of the image, that is, to indicate the composition ratio, color contrast, color saturation, and light-dark contrast of the image. For example, the higher the quality score of the image, the better the image quality of the image, that is, the better the effect corresponding to the composition ratio, color contrast, color saturation and light-dark contrast of the image.

Step 203: Determine the photographed picture according to the first image frame with the highest score among the image frames of the target video data.

Since the quality score is used to indicate the image quality of the image, the image frame with the highest score is the image frame with the best image quality among the image frames of the target video data. The image quality of the photographed picture determined by other image frames is better.

To sum up, in the embodiment of the present application, the image frame with the highest score is selected from the target video data, and the photographed picture is determined according to the image frame. Since the user does not need to manually filter, the server will automatically select the image frame with the highest score from the target video data. image frames, thus solving the problem of low self-portrait efficiency of the user, and achieving the effect of improving the self-portrait efficiency of the user.

It should be noted that before step 201, the server needs to train images to obtain an image scoring model.

Please refer to FIG. 2B , which shows a flowchart of a training process of an image scoring model provided by an exemplary embodiment of the present application. Optionally, the training process of the image scoring model includes but not limited to the following steps:

Step 204, acquiring multiple training samples.

Wherein, each training sample includes a sample image and a score of the sample image.

Optionally, the training samples are obtained from the terminal or other servers, and the scores of the sample images are manually determined.

Step 205, using the sample image as a training input and the sample image score as an output reference value to train the initial image classification model to obtain a trained image score model.

For the sample image and the sample image score in at least one training sample, extract sample image features from the sample image, input the sample image feature into the initial image classification model, and obtain the training result.

Optionally, the server extracts sample image features from sample images according to a preset image processing algorithm. Wherein, the preset image processing algorithm is a perceptual hash algorithm (Perceptual hash algorithm, pHash algorithm). The server calculates the perceptual hash value corresponding to the sample image through the pHash algorithm, and determines the calculated perceptual hash value as the feature of the sample image.

Optionally, the initial image classification model is established according to a neural network model, for example: the initial image classification model is established according to one of a CNN model, a DNN model and an RNN model.

Schematically, for each training sample, the terminal creates an input-output pair corresponding to the training sample, the input parameter of the input-output pair is the sample image feature corresponding to the sample image in the training sample, and the output parameter is the sample image in the training sample Scoring; the server inputs the input and output to the initial image classification model to obtain the training result.

For example, the sample image feature is "sample image feature 1", the sample image score is "sample image score 1", and the input-output pair created by the terminal is: (sample image feature 1) -> (sample image score 1); where ( The sample image feature 1) is an input parameter, and (sample image score 1) is an output parameter.

Optionally, input-output pairs are represented by feature vectors.

An image scoring model is obtained based on the above training, please refer to FIG. 3A , which shows a flow chart of an image acquisition method provided by another exemplary embodiment of the present application. This embodiment is described by taking the image acquisition method applied to the implementation environment shown in FIG. 1 as an example. The image acquisition methods include:

Step 301, acquire the captured target video data.

Step 302, acquiring a second image frame at a preset position in the target video data, and inputting the second image frame into an image classification model to obtain an image category of the second image frame.

Since the image category of the target object in the target video data includes at least one of people, animals, still life and scenery, the quality scoring standards of images corresponding to different image categories are different, so different image scoring models are selected according to different image categories, The image quality score calculated based on the image scoring model can be made more accurate.

For the shorter target video data (such as video data within 10s) of the shooting time, the target object in the target video data usually does not change, so a frame of image frame (second image frame) is extracted from the target video data , input the second image frame into the image classification model to obtain the image category of the second image frame, and use the image category of the second image frame as the image category of all image frames in the target video data.

It should be noted that the preset position refers to the position of the second image frame in the target video data, and the preset position can be the 5th frame, the 10th frame, and the 15th frame, and the specific value of the preset position is not specified in this embodiment. Do limited.

In a possible implementation manner, step 304 may be replaced by: analyzing the acquired target video data according to a preset video processing algorithm, and calculating the type identifier of the target object in the target video data, and according to the calculated The type identifier determines the image category corresponding to the type identifier, and the type identifier is used to uniquely identify the image category.

Step 303: Determine the target image scoring model corresponding to the image type of the second image frame according to the pre-stored correspondence between the image category and the image scoring model, input each image frame in the target video data into the target image scoring model, and obtain each Scores corresponding to image frames.

Among them, the corresponding relationship between the image category and the image scoring model is shown in Table 1. In Table 1, when the image category is "person", the corresponding image scoring model is "image scoring model 1"; when the image category is "animal", the corresponding image scoring model is "image scoring model 2"; the image category is For "still life", the corresponding image scoring model is "image scoring model 3"; when the image category is "landscape", the corresponding image scoring model is "image scoring model 4".

Table I

image category Image Scoring Model figure Image Scoring Model 1 animal Image Scoring Model 2 still life Image Scoring Model 3 landscape Image Scoring Model 4

Schematically, based on the preset correspondence provided in Table 1, when the image category of the target object acquired by the terminal is "person", the image scoring model "image scoring model 1" corresponding to the image category of the target object "person" is obtained ".

Step 304: Determine the photographed picture according to the first image frame with the highest score among the image frames of the target video data.

It should be noted that in this embodiment, step 301 is similar to step 201, and step 304 is similar to step 203. For specific descriptions of step 301 and step 304, please refer to step 201 and step 203 respectively, and details are not repeated here.

To sum up, in the embodiment of the present application, the image frame with the highest score is selected from the target video data, and the photographed picture is determined according to the image frame. Since the user does not need to manually filter, the server will automatically select the image frame with the highest score from the target video data. image frames, thus solving the problem of low self-portrait efficiency of the user, and achieving the effect of improving the self-portrait efficiency of the user.

In this embodiment, selecting different image scoring models according to different image categories can make the image quality score calculated based on the image scoring model more accurate.

It should be noted that before step 301, the server needs to train images to obtain an image classification model.

Please refer to FIG. 3B , which shows a flowchart of a training process of an image classification model provided by an exemplary embodiment of the present application. Optionally, the training process of the image scoring model includes but not limited to the following steps:

Step 305, acquiring multiple training samples.

Wherein, each training sample includes a sample image and a category of the sample image.

Optionally, the training samples are obtained from the terminal or other servers, and the category of the sample images is manually determined.

Step 306, using the sample image as a training input and the category of the sample image as an output reference value to train the initial image classification model to obtain a trained image classification model.

For the sample image and the sample image category in at least one training sample, sample image features are extracted from the sample image, and the sample image features are input into the initial image classification model to obtain a training result.

Optionally, the server extracts sample image features from sample images according to a preset image processing algorithm. Wherein, the preset image processing algorithm is a perceptual hash algorithm (Perceptual hash algorithm, pHash algorithm). The server calculates the perceptual hash value corresponding to the sample image through the pHash algorithm, and determines the calculated perceptual hash value as the feature of the sample image.

Optionally, the initial image classification model is established according to a neural network model, for example: the initial image classification model is established according to one of a CNN model, a DNN model and an RNN model.

Schematically, for each training sample, the terminal creates an input-output pair corresponding to the training sample, the input parameter of the input-output pair is the sample image feature corresponding to the sample image in the training sample, and the output parameter is the sample image in the training sample category; the server inputs the input and output pairs into the initial image classification model to obtain the training result.

For example, the sample image feature is "sample image feature 1", the sample image category is "sample image category 1", and the input-output pair created by the terminal is: (sample image feature 1) -> (sample image category 1); among them, ( Sample image feature 1) is an input parameter, and (sample image category 1) is an output parameter.

Optionally, input-output pairs are represented by feature vectors.

Please refer to FIG. 4 , which shows a flow chart of an image acquisition method provided in another exemplary embodiment of the present application. This embodiment is described by taking the image acquisition method applied to the implementation environment shown in FIG. 1 as an example. The image acquisition methods include:

Step 401, acquire the captured target video data.

In step 402, the image frames of the target video data are respectively input into the pre-trained image scoring model to obtain the corresponding score of each image frame.

Step 403, determine the first image frame with the highest score.

Step 404, extracting the first image frame, the first m image frames of the first image frame, and the last n image frames of the first image frame.

Since the score corresponding to the image frame is based on the comprehensive score of all pixel blocks in the image frame, it does not mean that each pixel block in the image frame has a higher score than the pixel blocks in the same position in other image frames. quality, after extracting the first image frame with the highest score, the first m image frames of the first image frame and the last n image frames of the first image frame can be extracted together, for the first image frame, the first Image synthesis is performed on the first m image frames of the image frame and the last n image frames of the first image frame to obtain a high-quality photographed image.

It should be noted that both m and n are positive integers, and the values of m and n may be the same or different, and this embodiment does not limit the specific values of m and n.

Since the quality of the synthesized photographic picture will be affected when the difference between the two images is large, in order to ensure the quality of the synthesized photographic picture, preferably, m is 1 and n is 1. That is, after determining the first image frame with the highest score, the server extracts the first image frame, an image frame before the first image frame, and an image frame after the first image frame.

In a special case, the first image frame is the first frame in the target video data, or the last frame in the target video data, that is, if the server cannot extract the first m image frames of the first image frame, then the first image frame is not The first m image frames of an image frame are extracted. Similarly, if the server cannot extract the last n image frames of the first image frame, the next n image frames of the first image frame are not extracted.

Step 405 , performing image synthesis on the first image frame, the first m image frames of the first image frame, and the last n image frames of the first image frame, to obtain a photographed picture.

Optionally, in the first image frame, the first m image frames of the first image frame, and the last n image frames of the first image frame, the pixel block with the highest definition among the pixel blocks at the same position in different image frames , to combine them to get the picture taken.

Optionally, in the first image frame, the first m image frames of the first image frame, and the last n image frames of the first image frame, the pixel with the highest color saturation in the pixel blocks at the same position in different image frames Blocks are combined to obtain a photographed picture.

It should be noted that, in this embodiment, steps 401 to 402 are similar to steps 201 to 202. For specific descriptions of steps 401 to 402, reference may be made to steps 201 to 202, which will not be repeated here.

To sum up, in the embodiment of the present application, the image frame with the highest score is selected from the target video data, and the photographed picture is determined according to the image frame. Since the user does not need to manually filter, the server will automatically select the image frame with the highest score from the target video data. image frames, thus solving the problem of low self-portrait efficiency of the user, and achieving the effect of improving the self-portrait efficiency of the user.

In this embodiment, image synthesis is performed on the first image frame, the first m image frames of the first image frame, and the last n image frames of the first image frame, so as to obtain a high-quality photographed image.

The following are device embodiments of the present application, which can be used to implement the method embodiments of the present application. For details not disclosed in the device embodiments of the present application, please refer to the method embodiments of the present application.

Please refer to FIG. 5 , which shows a schematic structural diagram of an image acquisition device provided by an exemplary embodiment of the present application. The image acquisition device can be implemented as all or part of the terminal in FIG. 1 through a dedicated hardware circuit, or a combination of software and hardware. The image acquisition device includes: a first acquisition module 501 , an input module 502 and a determination module 503 .

The first obtaining module 501 is used to obtain the target video data of shooting;

The input module 502 is used to input the image frames of the target video data into the image scoring model trained in advance respectively, so as to obtain the corresponding scoring of each image frame;

The determination module 503 is configured to determine a photographed picture according to the first image frame with the highest score among the image frames of the target video data.

In an optional embodiment provided based on the embodiment shown in FIG. 5 , the input module 502 includes: a first input unit and a second input unit.

The first input unit is used to obtain a second image frame at a preset position in the target video data, and input the second image frame into the image classification model to obtain the image category of the second image frame;

The second input unit is used to determine the target image scoring model corresponding to the image type of the second image frame according to the corresponding relationship between the pre-stored image category and the image scoring model, and input each image frame in the target video data into the target image scoring model respectively model to get the score corresponding to each image frame.

In an optional embodiment provided based on the embodiment shown in FIG. 5 , the device further includes: a second acquiring module and a first training module.

The second obtaining module is used to obtain a plurality of training samples, wherein each training sample includes a sample image and a sample image category;

The first training module is used to use the sample image as a training input and the category of the sample image as an output reference value to train the initial image classification model to obtain a trained image classification model.

In an optional embodiment provided based on the embodiment shown in FIG. 5 , the determining module 503 includes: a determining unit, an extracting unit, and a combining unit.

a determining unit, configured to determine the first image frame with the highest score;

An extraction unit, configured to extract the first image frame, the first m image frames of the first image frame, and the last n image frames of the first image frame;

The synthesis unit is configured to perform image synthesis on the first image frame, the first m image frames of the first image frame, and the last n image frames of the first image frame to obtain a photographed picture.

In an optional embodiment provided based on the embodiment shown in FIG. 5, the synthesis unit includes:

The first combination subunit is used to combine the sharpness of pixel blocks at the same position in different image frames in the first image frame, the first m image frames of the first image frame, and the last n image frames of the first image frame The highest pixel blocks are combined to obtain a photographed picture;

The second combination subunit is used to saturate the colors in the pixel blocks at the same position in different image frames in the first image frame, the first m image frames of the first image frame, and the last n image frames of the first image frame The pixel blocks with the highest degree of accuracy are combined to obtain a photographed picture.

In an optional embodiment provided based on the embodiment shown in FIG. 5 , the device further includes: a third acquiring module and a second training module.

The third obtaining module is used to obtain a plurality of training samples, wherein each training sample includes a sample image and a score of the sample image;

The second training module is used to train the initial image classification model by using the sample image as a training input and the sample image score as an output reference value to obtain a trained image score model.

Relevant details may be combined with reference to the method embodiments shown in FIG. 2A to FIG. 4 . Wherein, the first acquisition module 501 is also used to implement any other implicit or disclosed functions related to the receiving step in the above method embodiments; the input module 502 is also used to implement any other implicit or disclosed functions related to the receiving step in the above method embodiments. Functions related to the monitoring step; the determining module 503 is also configured to implement any other implicit or disclosed functions related to the monitoring step in the above method embodiments.

It should be noted that, when realizing the functions of the device provided by the above-mentioned embodiments, the division of the above-mentioned functional modules is used as an example for illustration. In practical applications, the above-mentioned function allocation can be completed by different functional modules according to the needs. The internal structure of the device is divided into different functional modules to complete all or part of the functions described above. In addition, the device and the method embodiment provided by the above embodiment belong to the same idea, and the specific implementation process thereof is detailed in the method embodiment, and will not be repeated here.

The present application also provides a computer-readable medium, on which program instructions are stored, and when the program instructions are executed by a processor, the image acquisition methods provided in the foregoing method embodiments are implemented.

The present application also provides a computer program product containing instructions, which, when run on a computer, causes the computer to execute the image acquisition method described in each method embodiment above.

The serial numbers of the above embodiments of the present application are for description only, and do not represent the advantages and disadvantages of the embodiments.

Those of ordinary skill in the art can understand that all or part of the steps in the file processing method of the above-mentioned embodiments can be completed by hardware, and can also be completed by instructing related hardware through a program, and the program can be stored in a computer-readable storage Among the media, the storage medium mentioned above may be a read-only memory, a magnetic disk or an optical disk, and the like. The above descriptions are only preferred embodiments of the application, and are not intended to limit the application. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the application shall be included in the protection of the application. within range.

Claims (9)

1. An image acquisition method, characterized in that the method comprises: Obtain the captured target video data; The image frames of the target video data are respectively input into the pre-trained image scoring model to obtain the corresponding scoring of each image frame; A photographed picture is determined according to the first image frame with the highest score among the image frames of the target video data. 2. The method according to claim 1, wherein the image frames of the target video data are respectively input into a pre-trained image scoring model to obtain the corresponding scoring of each image frame, including: Obtaining a second image frame at a preset position in the target video data, inputting the second image frame into an image classification model to obtain an image category of the second image frame; According to the correspondence between the pre-stored image category and the image scoring model, determine the target image scoring model corresponding to the image type of the second image frame, and input each image frame in the target video data into the target image scoring model respectively. , to get the score corresponding to each image frame. 3. The method according to claim 2, wherein the method further comprises: Obtaining a plurality of training samples, wherein each training sample includes a sample image and a sample image category; The sample image is used as a training input, and the category of the sample image is used as an output reference value to train an initial image classification model to obtain the trained image classification model. 4. The method according to claim 1, wherein the first image frame with the highest score in the image frame according to the target video data, determining the photographed picture includes: determining the first image frame with the highest score; extracting the first image frame, the first m image frames of the first image frame, and the last n image frames of the first image frame; performing image synthesis on the first image frame, the first m image frames of the first image frame, and the last n image frames of the first image frame to obtain a photographed picture. 5. The method according to claim 4, wherein the pair of the first image frame, the first m image frames of the first image frame and the last n image frames of the first image frame Perform image synthesis to obtain photographed pictures, including: In the first image frame, the first m image frames of the first image frame, and the last n image frames of the first image frame, the highest definition among the pixel blocks at the same position in different image frames Pixel blocks are combined to obtain a photographed picture; or, In the first image frame, the first m image frames of the first image frame, and the last n image frames of the first image frame, the pixel block with the highest saturation in the same position in different image frames Pixel blocks are combined to obtain a photographed picture. 6. The method according to any one of claims 1-5, wherein the method further comprises: Obtaining a plurality of training samples, wherein each training sample includes a sample image and a score of the sample image; The sample image is used as a training input, and the sample image score is used as an output reference value to train an initial image classification model to obtain the trained image score model. 7. An image acquisition device, characterized in that the device comprises: The first obtaining module is used to obtain the target video data of shooting; The input module is used to input the image frames of the target video data into the pre-trained image scoring model respectively to obtain the corresponding scoring of each image frame; The determination module is configured to determine the photographed picture according to the first image frame with the highest score among the image frames of the target video data. 8. A terminal, characterized in that the terminal comprises a processor and a memory, and at least one instruction is stored in the memory, and the instruction is loaded and executed by the processor to implement any one of claims 1 to 6. The image acquisition method described. 9. A computer-readable storage medium, characterized in that at least one instruction is stored in the storage medium, and the instruction is loaded and executed by a processor to implement the image acquisition method according to any one of claims 1 to 6 .