Visually rich document information extraction method, system and medium based on layout relevance

Abstract

The present invention discloses a method, system and medium for extracting visually rich document information based on layout relevance, wherein the method comprises: obtaining a visually rich document image, and annotating the visually rich document image to obtain a data set; constructing a document information extraction model for layout relevance, and using the data set to train the document information extraction model; obtaining a visually rich document image, and inputting the visually rich document image into the trained document information extraction model to obtain an information extraction result; wherein layout relevance refers to the positional relationship between static fields and dynamic fields, static fields are fields with fixed text in the same template, and dynamic fields are fields in the same template that change according to actual content. The present invention provides a scheme for extracting visually rich document information using document layout relevance, which can achieve high-precision visually rich document information extraction tasks with small sample data, and can be widely used in the field of visual information extraction.

Description

Technical Field

[0001] The present invention relates to the field of visual information extraction, and in particular to a method, system and medium for extracting visual rich document information based on layout relevance. Background Art

[0002] Visually rich document images are everywhere in people's daily lives. For example, cards, bills, contracts, etc. are all included in the category of visually rich documents. These images play an important role in all aspects of people's lives. For example, identity cards, bank cards, invoices, etc. record some important information. The common feature of these images is that their semantic hierarchy is not only determined by the semantics of the text itself, but also related to the font color, size, position, layout, etc. of the text. The usual text detection and recognition technology can only transcribe all the text above into an electronic version. But this is far from enough. In daily work and life, we need to know not only the text on these pictures, but also the information behind these words. Therefore, information extraction for visually rich documents is particularly important. For example, daily automatic filling in information, automatic correction, automatic resume extraction, etc. are inseparable from information extraction technology. The purpose of extracting key information from visually rich documents is to extract the named entities that people are concerned about from the existing OCR results of these pictures to form key-value pairs.

[0003] In recent years, with the development of deep learning technology, deep learning-based OCR text detection and recognition technology has made great progress, and mature products have been put into all aspects of people's daily life. At the same time, in addition to knowing what the text in the document image is in the scene, people are more concerned about the meaning behind these texts, such as: date, name, amount, etc. Therefore, the visual rich document information extraction method based on deep learning has gradually become a hot topic of research and attention. Based on methods such as BiLSTM-CRF, related studies have introduced graph convolutional networks on this basis to integrate layout and other features into the document graph structure as a reference. There are also methods to construct raster images and fuse word embedding features in the channel dimension. These methods have achieved good performance through training and testing under labeled data sets. However, there are also some problems. For example, existing methods often continue the idea of ​​solving problems in natural language processing, flattening two-dimensional text information with spatial layout relationships into a sequence model. In this process, the relationship between the layout is ignored and each field is separated for analysis. This actually ignores the existing spatial relationship of visual rich documents. In addition, most of the existing methods are designed for specific scenarios to train and test models. As a result, the model is only applicable to fixed scenarios and the generalization performance is greatly reduced. The above difficulties have seriously hindered the development of information extraction technology. Summary of the invention

[0004] In order to solve at least one of the technical problems existing in the prior art to a certain extent, the present invention aims to provide a method, system and medium for extracting visual rich document information based on layout relevance.

[0005] The technical solution adopted by the present invention is:

[0006] A method for extracting information from visually rich documents based on layout relevance comprises the following steps:

[0007] Obtain visually rich document images, and annotate the visually rich document images to obtain a data set;

[0008] Build a document information extraction model for layout relevance and use the data set to train the document information extraction model;

[0009] Obtain a visually rich document image, input the visually rich document image into the trained document information extraction model, and obtain an information extraction result;

[0010] Among them, layout relevance refers to the positional relationship between static fields and dynamic fields. Static fields are fields with fixed text in the same template, and dynamic fields are fields in the same template that change according to actual content.

[0011] Furthermore, the training process of the document information extraction model includes:

[0012] Different semantics in the field are embedded with unique 1024-dimensional vectors, and the positional features of the field, such as the relative position of the center point coordinates, are quantified into numbers. The processed semantics and positional features are input together as input features;

[0013] For each category in the preset model, calculate the mean of all input features in this category as the category center of each category;

[0014] Classification is done by measuring the distance between the sample and the class center.

[0015] Furthermore, the document information extraction model processes the input visually rich document image as follows:

[0016] Get each individual field in the visual rich document image as a node in the graph;

[0017] Get the edge connection relationship between nodes; where the edge connection relationship is {|X i-j |,|Y i-j |, W i / W j , H i / H j},|X i-j | represents the distance between two field nodes on the horizontal axis, |Yi-j | represents the distance between two field nodes on the ordinate, W i / W j Indicates the ratio of the widths of the two rectangular boxes corresponding to the two field nodes, H i / H j Indicates the ratio of the heights of the two rectangular boxes corresponding to the two field nodes;

[0018] Get the connection relationship between all static fields and dynamic fields, and get the matching relationship between static fields and dynamic fields based on the connection relationship.

[0019] Furthermore, the matching relationship is obtained in the following manner:

[0020] Get the matching probability value between the preset field and all fields;

[0021] Selecting a matching relationship with a matching probability value greater than a threshold value indicates that there is a matching relationship between the two fields.

[0022] Furthermore, the matching relationship includes a one-to-one matching relationship, a one-to-many matching relationship, and a many-to-one matching relationship;

[0023] When a one-to-many matching relationship or a many-to-one matching relationship occurs, the optimal probability matching is obtained according to the probability value.

[0024] Further, obtaining the optimal probability match according to the probability value includes:

[0025] Get the probability value R of the classification result of each dynamic field;

[0026] Sort the probability value R of each dynamic field about the classification result from large to small;

[0027] Traverse element i in the set of probability values ​​R;

[0028] Add the top three probability values ​​of elements i to the set Q;

[0029] Traverse all elements j except element i in the set of probability values ​​R;

[0030] If the cumulative probability sum of the current probability value is greater than the cumulative probability sum of the existing probabilities in the set, then the current probability value is added to the set Q and the old value is removed;

[0031] Update the set Q.

[0032] Furthermore, the visually rich document image includes a ticket image, an invoice image, a certificate image, and an ID image.

[0033] Another technical solution adopted by the present invention is:

[0034] A visual rich document information extraction system based on layout relevance, comprising:

[0035] A data acquisition module is used to acquire visually rich document images and annotate the visually rich document images to obtain a data set;

[0036] Model training module, used to build a document information extraction model for layout relevance, and use the data set to train the document information extraction model;

[0037] An information extraction module is used to obtain visually rich document images, input the visually rich document images into the trained document information extraction model, and obtain information extraction results;

[0038] Among them, layout relevance refers to the positional relationship between static fields and dynamic fields. Static fields are fields with fixed text in the same template, and dynamic fields are fields in the same template that change according to actual content.

[0039] Another technical solution adopted by the present invention is:

[0040] A visual rich document information extraction system based on layout relevance, comprising:

[0041] at least one processor;

[0042] at least one memory for storing at least one program;

[0043] When the at least one program is executed by the at least one processor, the at least one processor implements the above method.

[0044] Another technical solution adopted by the present invention is:

[0045] A computer-readable storage medium stores a program executable by a processor, wherein the program executable by the processor is used to execute the method described above when executed by the processor.

[0046] The beneficial effect of the present invention is that the present invention provides a scheme for extracting visual rich document information by utilizing the relevance of document layout, which can realize high-precision visual rich document information extraction tasks with small sample data, and has a positive effect on information extraction work in scenarios such as bank cards, bills, contract documents, etc. with similar layout information. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the embodiments of the present invention or the drawings of related technical solutions in the prior art are introduced below. It should be understood that the drawings introduced below are only for the convenience of clearly describing some embodiments of the technical solutions of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative work.

[0048] Figure 1 It is a flowchart of the steps of a method for extracting visually rich document information based on layout relevance in an embodiment of the present invention;

[0049] Figure 2 is a schematic diagram of sampling of a visual rich document information extraction data set used in an embodiment of the present invention;

[0050] Figure 3 is a model diagram of the information extraction and matching method in an embodiment of the present invention;

[0051] Figure 4 It is a schematic diagram of the information extraction and matching effect in an embodiment of the present invention. DETAILED DESCRIPTION

[0052] The embodiments of the present invention are described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and are not to be construed as limitations of the present invention. For the step numbers in the following embodiments, they are only provided for the convenience of explanation, and the order between the steps is not limited in any way, and the execution order of each step in the embodiment can be adaptively adjusted according to the understanding of those skilled in the art.

[0053] In the description of the present invention, it should be understood that descriptions involving orientations, such as up, down, front, back, left, right, etc., and orientations or positional relationships indicated are based on the orientations or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as a limitation on the present invention.

[0054] In the description of the present invention, "several" means one or more, "more" means more than two, "greater than", "less than", "exceed" etc. are understood as not including the number itself, and "above", "below", "within" etc. are understood as including the number itself. If there is a description of "first" or "second", it is only used for the purpose of distinguishing the technical features, and cannot be understood as indicating or implying the relative importance or implicitly indicating the number of the indicated technical features or implicitly indicating the order of the indicated technical features.

[0055] In the description of the present invention, unless otherwise clearly defined, terms such as setting, installing, connecting, etc. should be understood in a broad sense, and technicians in the relevant technical field can reasonably determine the specific meanings of the above terms in the present invention based on the specific content of the technical solution.

[0056] like Figure 1 As shown, this embodiment provides a method for extracting visually rich document information based on layout relevance, comprising the following steps:

[0057] S1. Obtain visually rich document images, and annotate the visually rich document images to obtain a data set.

[0058] S2. Build a document information extraction model for layout relevance and use the data set to train the document information extraction model.

[0059] S3. Obtain a visually rich document image, input the visually rich document image into the trained document information extraction model, and obtain an information extraction result.

[0060] Among them, layout relevance refers to the positional relationship between static fields and dynamic fields. Static fields are fields with fixed text in the same template, and dynamic fields are fields in the same template that change according to actual content.

[0061] The following is combined with Figure 2-4 A detailed explanation of the training and use of the document information extraction model.

[0062] First, we need to collect some visual rich document datasets to be extracted. The visual rich document datasets implemented in this paper cover samples with complex layout and simple layout, which are collected from EATEN, SROIE, CORD, wirldreceipt and other datasets, respectively. They have 1000, 300, 400, 300, a total of 2000 image data, and then annotate these 2000 image data, mainly including the content of the text, the location of the text, and the three elements of its named entity attributes. The pictures sampled in this dataset are as follows: Figure 2 As shown, it can be found that these images are related in layout, and the attributes of some fields can be determined by the semantics of their adjacent fields. According to the layout in this dataset, it can be divided into several categories of information. The specific statistical categories are shown in Table 1 below:

[0063] Table 1 Distribution statistics of visual rich document dataset

[0064] EATEN SROIE CORD wirldreceipt Ticket Type Ticket Business Card bill bill bill Sample size 1000 300 400 300 Number of layouts 4 4 2 2

[0065] Among them, the model diagram of the visual rich document information extraction model is as follows Figure 3As shown in the figure, the query set is the set of samples that need to be extracted, and the support set refers to the set of templates used for reference.

[0066] Training the model for extracting information from visual rich documents: All the 2,000 images obtained are randomly divided into training and test sets at a ratio of 4:1. That is, the training data has 1,600 images and the test data set has 400 images. According to the layout type, it is divided into 12 types, of which the training set contains 8 types and the test set contains 4 types; it should be noted here that the specific type can be determined according to the actual usage. The fields are defined into two types. The first type is called static fields. These fields are usually unchanging texts, such as "date", "name", "amount", etc., which are used to explain other fields. These fields are fixedly distributed in the image and have a certain layout distribution similarity. The other type is called dynamic field, which will be adjusted according to the specific situation. For example, everyone's name, date, etc. will be automatically adjusted. The overall idea of ​​the method is to use the ideas of prototype network and matching network in few-sample learning. The difference is that compared with previous methods, it is mainly aimed at modeling the features of images. Here, the main modeling is the positional relationship between the static fields and dynamic fields mentioned above, as well as their own semantic features.

[0067] Using the idea of ​​document graph, each independent field is regarded as a node in the graph, and the edge connection relationship between each node is defined as the spatial position relationship between two fields. Take node i and node j as an example. The connection relationship between them is {|X i-j |,|Y i-j |, W i / W j , H i / H j}, respectively representing the horizontal and vertical coordinates between the center points and the width-to-height ratio. Regarding the semantic features of the field, that is, the features of the node, the semantic representation of the sentence meaning extracted by the large-scale pre-trained language model is adopted, and the representation is represented by a 768-dimensional feature. Specifically, for each word in the vocabulary, there is and only one unique 768-dimensional vector to represent the similarity between words. Regarding feature extraction, first calculate the edge and node features between all static fields and dynamic fields, organize them according to the relationship of "node-edge-node" triples, and calculate the feature value of the category center according to the number of categories of the dynamic field. Specifically, calculate the average value of the sum of the feature values ​​between the static fields under all the connection relationships of a dynamic field. Then, use the attention mechanism to merge the feature values ​​of each field to be queried with the above-calculated category center and send them to the two-layer multilayer perceptron. Based on the idea of ​​matching network based on meta-learning, the classification probability value of the number of dynamic field categories × the total number of categories is obtained, and the gap between the cross entropy loss function and the target value is calculated, and the model parameters of deep learning are updated through gradient feedback.

[0068] It is worth mentioning that in order to make the entire parameter calculation differentiable, this method uses the DD-ILP solver. Finally, the accuracy of the parameters of the designed model is shown in Table 2:

[0069] Table 2. Accuracy of the proposed visual rich document information extraction method

[0070] F1-score(%) Method(input: node features) 78.69 Method (input: node features, edge features) 83.65

[0071] In addition, the number of matching combinations obtained may be many-to-one, one-to-many, etc. At this time, it is necessary to design additional corresponding rules for filtering. To this end, in order to solve this problem, this embodiment designs the following algorithm:

[0072]

[0073] After training the model, we input a visually rich document image and extracted the information of the document layout using a few samples of visually rich document information. The final result is shown in the attached figure. Figure 4 shown.

[0074] In summary, this embodiment solves the problem of how to learn a migration relationship of layout features based on the correlation of text layouts between samples under small sample data. This makes it possible to achieve high-precision visual rich document information extraction tasks by only annotating a small number of samples. This has a positive effect on information extraction work in scenarios such as bank cards, bills, and contract documents with similar layout information. It has the advantages of strong generalization, high precision, and fast processing speed.

[0075] This embodiment also provides a visual rich document information extraction system based on layout relevance, including:

[0076] A data acquisition module is used to acquire visually rich document images and annotate the visually rich document images to obtain a data set;

[0077] Model training module, used to build a document information extraction model for layout relevance, and use the data set to train the document information extraction model;

[0078] An information extraction module is used to obtain visually rich document images, input the visually rich document images into the trained document information extraction model, and obtain information extraction results;

[0079] Among them, layout relevance refers to the positional relationship between static fields and dynamic fields. Static fields are fields with fixed text in the same template, and dynamic fields are fields in the same template that change according to actual content.

[0080] A visual rich document information extraction system based on layout relevance in this embodiment can execute a visual rich document information extraction method based on layout relevance provided by the method embodiment of the present invention, can execute any combination of implementation steps of the method embodiment, and has the corresponding functions and beneficial effects of the method.

[0081] This embodiment also provides a visual rich document information extraction system based on layout relevance, including:

[0082] at least one processor;

[0083] at least one memory for storing at least one program;

[0084] When the at least one program is executed by the at least one processor, the at least one processor implements Figure 1 The method shown.

[0085] A visual rich document information extraction system based on layout relevance in this embodiment can execute a visual rich document information extraction method based on layout relevance provided by the method embodiment of the present invention, can execute any combination of implementation steps of the method embodiment, and has the corresponding functions and beneficial effects of the method.

[0086] The present application also discloses a computer program product or a computer program, which includes a computer instruction stored in a computer-readable storage medium. A processor of a computer device can read the computer instruction from the computer-readable storage medium, and the processor executes the computer instruction, so that the computer device executes Figure 1 The method shown.

[0087] This embodiment also provides a storage medium storing instructions or programs that can execute a method for extracting visual rich document information based on layout relevance provided by an embodiment of the method of the present invention. When the instructions or programs are run, any combination of implementation steps of the method embodiment can be executed, and the corresponding functions and beneficial effects of the method can be obtained.

[0088] In some selectable embodiments, the function / operation mentioned in the block diagram may not occur in the order mentioned in the operation diagram. For example, depending on the function / operation involved, the two boxes shown in succession can actually be executed substantially simultaneously or the boxes can sometimes be executed in reverse order. In addition, the embodiment presented and described in the flow chart of the present invention is provided by way of example, for the purpose of providing a more comprehensive understanding of technology. The disclosed method is not limited to the operation and logic flow presented herein. Selectable embodiments are expected, wherein the order of various operations is changed and the sub-operation of a part for which is described as a larger operation is performed independently.

[0089] In addition, although the present invention is described in the context of functional modules, it should be understood that, unless otherwise specified, one or more of the functions and / or features described may be integrated into a single physical device and / or software module, or one or more functions and / or features may be implemented in separate physical devices or software modules. It is also understood that a detailed discussion of the actual implementation of each module is unnecessary for understanding the present invention. More specifically, in view of the properties, functions, and internal relationships of the various functional modules in the device disclosed herein, the actual implementation of the module will be understood within the conventional skills of the engineer. Therefore, those skilled in the art can implement the present invention set forth in the claims without excessive experimentation using ordinary techniques. It is also understood that the specific concepts disclosed are merely illustrative and are not intended to limit the scope of the present invention, which is determined by the full scope of the appended claims and their equivalents.

[0090] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc., which can store program codes.

[0091] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in conjunction with such instruction execution systems, devices or apparatuses. For the purposes of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in conjunction with such instruction execution systems, devices or apparatuses.

[0092] More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or more wires (electronic device), a portable computer disk case (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be a paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering or, if necessary, processing in another suitable manner, and then stored in a computer memory.

[0093] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, a plurality of steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0094] In the above description of this specification, the description with reference to the terms "one embodiment / example", "another embodiment / example" or "certain embodiments / examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner.

[0095] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the claims and their equivalents.

[0096] The above is a specific description of the preferred implementation of the present invention, but the present invention is not limited to the above embodiments. Those skilled in the art may make various equivalent modifications or substitutions without violating the spirit of the present invention. These equivalent modifications or substitutions are all included in the scope defined by the claims of this application.

Claims

1. A method for extracting visually rich document information based on layout relevance, characterized in that: The following steps are involved: Acquire visually rich document images and annotate the visually rich document images to obtain a data set; the visually rich document images include ticket images, invoice images, certificate images, and document images; Construct a document information extraction model for layout relevance, and use the data set to train the document information extraction model; obtain visually rich document images, input the visually rich document images into the trained document information extraction model, and obtain information extraction results; Among them, layout relevance refers to the positional relationship between static fields and dynamic fields. Static fields are fields with fixed text in the same template, and dynamic fields are fields in the same template that change according to actual content. The training process of the document information extraction model includes: The different semantics in the field are represented as a 1024-dimensional embedding vector, the positional features of the field are quantized into numbers, and the processed semantic and positional features are input as input features; For each category in the preset model, calculate the mean of all input features in this category as the category center of each category; Classification is done by measuring the distance between the sample and the class center; The document information extraction model processes the input visually rich document image as follows: Get each individual field in the visual rich document image as a node in the graph; Get the edge connection relationship between nodes; where the edge connection relationship is {|X i-j |,|Y i-j |, W i / W j , H i / H j },|X i-j | represents the distance between two field nodes on the horizontal axis, |Y i-j | represents the distance between two field nodes on the ordinate, W i / W j Indicates the ratio of the widths of the two rectangular boxes corresponding to the two field nodes, H i / H j Indicates the ratio of the heights of the two rectangular boxes corresponding to the two field nodes; Get the connection relationship between all static fields and dynamic fields, and get the matching relationship between static fields and dynamic fields based on the connection relationship; The matching relationship is obtained in the following way: Get the matching probability value between the preset field and all fields; Selecting a matching relationship with a matching probability value greater than a threshold value indicates a matching relationship between two fields; The matching relationship includes a one-to-one matching relationship, a one-to-many matching relationship, and a many-to-one matching relationship; When a one-to-many matching relationship or a many-to-one matching relationship occurs, the optimal probability matching is obtained according to the probability value.

2. According to the method for extracting visually rich document information based on layout relevance according to claim 1, it is characterized in that: The obtaining of the optimal probability match according to the probability value includes: Get the probability value R of the classification result of each dynamic field; Sort the probability value R of each dynamic field about the classification result from large to small; Traverse element i in the set of probability values ​​R; Add the top three probability values ​​of elements i to the set Q; Traverse all elements j except element i in the set of probability values ​​R; If the cumulative probability sum of the current probability value is greater than the cumulative probability sum of the existing probabilities in the set, then the current probability value is added to the set Q and the old value is removed; Update the set Q.

3. A visual rich document information extraction system based on layout relevance, applied to a visual rich document information extraction method based on layout relevance as described in any one of claims 1-2, characterized in that: include: A data acquisition module is used to acquire visually rich document images and annotate the visually rich document images to obtain a data set; Model training module, used to build a document information extraction model for layout relevance, and use the data set to train the document information extraction model; An information extraction module is used to obtain visually rich document images, input the visually rich document images into the trained document information extraction model, and obtain information extraction results; Among them, layout relevance refers to the positional relationship between static fields and dynamic fields. Static fields are fields with fixed text in the same template, and dynamic fields are fields in the same template that change according to actual content.

4. A visually rich document information extraction system based on layout relevance, characterized in that: include: at least one processor; at least one memory for storing at least one program; When the at least one program is executed by the at least one processor, the at least one processor implements the method according to any one of claims 1 to 2.

5. A computer-readable storage medium storing a program executable by a processor, characterized in that: The processor-executable program is used to perform the method according to any one of claims 1 to 2 when executed by the processor.

Images