Chart preparation information identification method and device, storage medium and electronic equipment

Abstract

The invention provides a chart preparation information identification method and device, a storage medium and electronic equipment. The identification method comprises the steps that a to-be-identified chart is acquired; and on the basis of a pre-trained recognition model, chart preparation information corresponding to the to-be-recognized chart is determined, and the chart preparation information at least comprises a preparation tool and drawing parameters. According to the method, the chart preparation information corresponding to the to-be-identified chart can be accurately identified through the pre-trained identification model, namely, the preparation tool and the drawing parameters corresponding to the to-be-identified chart can be identified without depending on the literature to which the to-be-identified chart belongs, so that the flexibility is relatively high; moreover, there is no need to traverse a large number of charts, and the recognition efficiency is effectively improved.

Description

Technical Field

[0001] This application relates to the technical field of chart recognition, and particularly to a method, device, storage medium and electronic device for identifying chart preparation information. Background Art

[0002] In existing literature or public materials, professional tables and charts are usually presented in various forms. Some are generated by specific software, while others may be created through programming languages. However, although these charts may be very effective in conveying information, their aesthetics and visualization effects vary. Generally, elegant and intuitive charts are crucial for conveying data and ideas. However, when seeing an elegant and intuitive chart and wanting to imitate and create with one's own data, there is a lack of information on what software or programming language was used to generate it and the relevant charting parameters.

[0003] Currently, there are mainly two ways to know how a chart is prepared: The first is to directly search for relevant information on chart preparation in the literature to which the chart belongs, but this method depends on the literature itself, and most literature does not mention relevant information on chart preparation; the second is to use the chart similarity search function of a search engine to find charts similar to the chart in the literature, and then view the corresponding preparation information of the chart. However, the probability of finding a chart similar to the chart in the literature by this method is relatively low, and a large number of charts need to be traversed, resulting in low efficiency. Even if a similar chart can be found, the preparation information of the chart similar to the chart in the literature may not be obtained.

[0004] Therefore, there is an urgent need for a method to obtain chart preparation information. Summary of the Invention

[0005] In view of this, the purpose of the embodiments of this application is to provide a method, device, storage medium and electronic device for identifying chart preparation information, which can accurately and efficiently identify the chart preparation information corresponding to the chart to be identified.

[0006] In a first aspect, the embodiments of this application provide a method for identifying chart preparation information, which includes:

[0007] Obtain the chart to be identified;

[0008] Based on a pre-trained recognition model, determine the chart preparation information corresponding to the chart to be identified, where the chart preparation information at least includes a preparation tool and charting parameters.

[0009] In a possible implementation manner, the determining the chart preparation information corresponding to the chart to be identified based on a pre-trained recognition model includes:

[0010] Input the chart to be recognized into the recognition model so that the recognition model converts the chart to be recognized into a first vector;

[0011] Calculate the first vector through the recognition model to obtain a second vector;

[0012] Convert the second vector into chart preparation information corresponding to the chart to be recognized.

[0013] In a possible implementation manner, the calculating the second vector by the recognition model for the first vector includes:

[0014] Obtain supplementary information determined by the user, where the supplementary information can describe the type and / or attributes of the chart to be recognized;

[0015] Input the supplementary information into the recognition model so that the recognition model converts the supplementary information into a third vector;

[0016] Calculate the second vector by the recognition model for the first vector and the third vector.

[0017] In a possible implementation manner, before determining the chart preparation information corresponding to the chart to be recognized based on a pre-trained recognition model, it further includes:

[0018] Obtain a plurality of chart preparation information;

[0019] Convert each piece of chart preparation information into a candidate vector that can characterize it through the recognition model and store it;

[0020] The calculating the second vector by the recognition model for the first vector includes:

[0021] Screen, through the recognition model, a candidate vector that matches the first vector as the second vector.

[0022] In a possible implementation manner, before determining the chart preparation information corresponding to the chart to be recognized based on a pre-trained recognition model, it further includes:

[0023] Obtain a plurality of preparation tools and a plurality of mapping parameters;

[0024] Convert each preparation tool and each set of mapping parameters into sub-vectors through the recognition model;

[0025] The calculating the second vector by the recognition model for the first vector includes:

[0026] Perform prediction on the first vector through the recognition model to obtain a second vector formed by a plurality of sub-vectors.

[0027] In a possible implementation, it further includes the step of generating an identification model:

[0028] Obtain a training sample set, where the training sample set includes a plurality of chart samples and a plurality of text samples;

[0029] Input the chart samples and the text samples into the identification model to be trained to obtain a theoretical pairing result, where the theoretical pairing result includes the pairing scores of each chart vector and each text vector;

[0030] Determine the trained identification model based on the theoretical pairing result and the actual pairing result.

[0031] In a possible implementation, the determining the trained identification model based on the theoretical pairing result and the actual pairing result includes:

[0032] Use a loss function to calculate the theoretical pairing result and the actual pairing result to obtain a calculation result;

[0033] When the calculation result meets a preset condition, determine that the trained identification model is obtained;

[0034] When the calculation result does not meet the preset condition, adjust the parameters of the identification model until the calculation result meets the preset condition.

[0035] In a second aspect, an embodiment of the present application further provides an identification device for chart preparation information, which includes:

[0036] An acquisition module configured to acquire a chart to be identified;

[0037] A determination module configured to determine the chart preparation information corresponding to the chart to be identified based on a pre-trained identification model, where the chart preparation information at least includes a preparation tool and drawing parameters.

[0038] In a third aspect, an embodiment of the present application further provides a storage medium, where a computer program is stored on the computer-readable storage medium, and when the computer program is run by a processor, it executes the steps of the identification method of chart preparation information as described in any one of the above.

[0039] In a fourth aspect, an embodiment of the present application further provides an electronic device, which includes: a processor and a memory, the memory stores machine-readable instructions executable by the processor, when the electronic device runs, the processor communicates with the memory through a bus, and when the machine-readable instructions are executed by the processor, it executes the steps of the identification method of chart preparation information as described in any one of the above.

[0040] Through the pre-trained recognition model, this application can accurately identify the chart preparation information corresponding to the chart to be recognized, that is, identify the preparation tools and drawing parameters corresponding to the chart to be recognized, without relying on the literature to which the chart to be recognized belongs, and has high flexibility; moreover, there is no need to traverse a large number of charts, effectively improving the recognition efficiency.

[0041] To make the above objects, features, and advantages of this application more obvious and understandable, the following specifically gives preferred embodiments and, in conjunction with the accompanying drawings, the detailed description is as follows. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] To more clearly illustrate the technical solutions in this application or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments recorded in this application. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0043] Figure 1 Shows the flowchart of the method for identifying chart preparation information provided by this application;

[0044] Figure 2 Shows the schematic diagram of the recognition model provided by this application for determining the chart preparation information corresponding to the chart to be recognized;

[0045] Figure 3 Shows the flowchart of obtaining the second vector in the method for identifying chart preparation information provided by this application;

[0046] Figure 4 Shows the flowchart of generating the recognition model in the method for identifying chart preparation information provided by this application;

[0047] Figure 5 Shows the schematic diagram of training the recognition model provided by this application;

[0048] Figure 6 Shows the structural schematic diagram of the device for identifying chart preparation information provided by this application;

[0049] Figure 7 Shows the structural schematic diagram of the electronic device provided by this application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0050] Reference is made herein to the accompanying drawings to describe the various aspects and features of this application.

[0051] It should be understood that various modifications can be made to the embodiments applied for herein. Therefore, the above description should not be construed as limiting, but merely as an example of the embodiments. Those skilled in the art will envision other modifications within the scope and spirit of the present application.

[0052] The accompanying drawings, which are included in and constitute a part of this specification, illustrate embodiments of the present application and, together with the general description of the present application given above and the detailed description of the embodiments given below, serve to explain the principles of the present application.

[0053] These and other features of the present application will become apparent from the following description of the preferred forms of the embodiments given by way of non-limiting example with reference to the accompanying drawings.

[0054] It should also be understood that although the present application has been described with reference to some specific examples, those skilled in the art can surely implement many other equivalent forms of the present application, which have the features as described in the claims and thus are all within the scope of protection defined thereby.

[0055] When combined with the accompanying drawings, the above and other aspects, features, and advantages of the present application will become more apparent in view of the following detailed description.

[0056] Specific embodiments of the present application will be described hereinafter with reference to the accompanying drawings; however, it should be understood that the embodiments applied for are merely examples of the present application and can be implemented in various ways. Well-known and / or repetitive functions and structures are not described in detail to avoid obscuring the present application with unnecessary or redundant details. Therefore, the specific structural and functional details applied for herein are not intended to be limiting, but merely as a basis and representative basis for the claims to teach those skilled in the art to use the present application in substantially any suitable detailed structure in a variety of ways.

[0057] This specification may use the phrases "in one embodiment", "in another embodiment", "in yet another embodiment", or "in other embodiments", which may each refer to one or more of the same or different embodiments according to the present application.

[0058] First, for the convenience of understanding the present application, a method for identifying chart preparation information provided by the present application will be introduced in detail. As Figure 1 shown, the method for identifying chart preparation information provided by the embodiments of the present application specifically includes S101 and S102.

[0059] S101, obtain the chart to be identified.

[0060] In specific implementation, the chart to be identified includes pictures and tables in various fields. For example, molecular diagrams in the chemical field, device diagrams in the mechanical field, etc.

[0061] Furthermore, the chart to be recognized can be a part of an article, such as a chart in a published paper; the chart to be recognized can also be a part of a presentation document, such as a chart in an internal company meeting; the chart to be recognized can also be an independent chart, such as a chart for promotional content, etc.

[0062] Optionally, when the user needs to determine the chart preparation information corresponding to the chart in a PDF file, obtain the file determined by the user and perform chart recognition on the file to extract the chart to be recognized. Of course, when the user needs to determine the chart preparation information corresponding to a chart, the chart is determined as the chart to be recognized.

[0063] S102, based on a pre-trained recognition model, determine the chart preparation information corresponding to the chart to be recognized, where the chart preparation information at least includes a preparation tool and drawing parameters.

[0064] After obtaining the chart to be recognized, based on a pre-trained recognition model, determine the chart preparation information corresponding to the chart to be recognized. Among them, the recognition model is used to recognize the chart preparation information corresponding to the chart.

[0065] When determining the chart preparation information corresponding to the chart to be recognized based on a pre-trained recognition model, first input the chart to be recognized into the recognition model so that the recognition model converts the chart to be recognized into a first vector. Optionally, the recognition model includes an image encoder, and the image encoder is used to convert the chart to be recognized into a first vector. Among them, the image encoder can be various neural networks or networks pre-trained in various types of pictures.

[0066] After that, continue to calculate the first vector through the recognition model to obtain a second vector, and then convert the second vector into the chart preparation information corresponding to the chart to be recognized. Among them, the chart preparation information at least includes a preparation tool and drawing parameters. The preparation tool includes an application program and / or a programming language, and the drawing parameters at least include one or more of line type, thickness, color, and contrast. For example, the chart preparation information can include application program A and parameter 1, the chart preparation information can also include application program C and parameter 2, and the chart preparation information can also include programming language S and parameter 3, etc.

[0067] In one example, the recognition model prepares parameters corresponding to multiple chart preparation information stored by itself as candidate vectors corresponding to the chart preparation information, searches for a second vector corresponding to the first vector, that is, filters candidate vectors matching the first vector as the second vector, so as to obtain the chart preparation information corresponding to the chart to be recognized. Based on this, before determining the chart preparation information corresponding to the chart to be recognized based on a pre-trained recognition model, it is necessary to obtain multiple chart preparation information. Then, based on the characteristics of the model input and output, that is, the model can recognize vectors, each chart preparation information is converted into a candidate vector that can represent it and stored through the recognition model, so that when the recognition model receives the chart to be recognized, it can filter out candidate vectors matching the first vector, that is, find the chart preparation information corresponding to the chart to be recognized from multiple chart preparation information. Among them, the recognition model also includes a text encoder, and then the text encoder is used to convert each chart preparation information into a candidate vector that can represent it. Moreover, the text encoder in the embodiments of the present application can be various neural network architectures, or a pre-trained text encoding or generation model, such as Bert, GPT, etc.

[0068] As an example, Figure 2 shows a schematic diagram of the recognition model determining the chart preparation information corresponding to the chart to be recognized. Referring to Figure 2 , the recognition model includes an image encoder and a text encoder. The image encoder converts Chart 1 into a first vector, and the text encoder converts Text 1, Text 2, and Text... into candidate vectors respectively, and calculates the matching degree between each candidate vector and the first vector. 1 represents a high matching degree, and 0 represents a low matching degree. Then, candidate vectors matching the first vector are filtered out as the second vector, and the second vector is converted into chart preparation information.

[0069] Furthermore, as the number of new preparation tools or new mapping parameters increases, the candidate vectors stored in the recognition model that can represent chart preparation information will also increase. Therefore, in practical applications, new preparation tools and / or new mapping parameters can be obtained in real time or periodically, so as to update the recognition model based on the new preparation tools and / or new mapping parameters, that is, update the candidate vectors stored in the recognition model, thereby ensuring that the chart preparation information recognized by the recognition model is more accurate.

[0070] Among them, updating the recognition model includes generating multiple new chart preparation information based on the new preparation tools and / or new mapping parameters, and converting the new chart preparation information into a second vector and storing it in the recognition model.

[0071] In another example, when the recognition model calculates the second vector from the first vector, it can also make predictions on the first vector to obtain multiple sub-vectors, and form the second vector based on the multiple sub-vectors, so as to obtain the chart preparation information corresponding to the chart to be recognized. Based on this, before determining the chart preparation information corresponding to the chart to be recognized based on the pre-trained recognition model, obtain multiple preparation tools and multiple mapping parameters, and use the text encoder in the recognition model to convert each preparation tool and each set of mapping parameters into sub-vectors respectively. Here, each preparation tool corresponds to a sub-vector, and each set of mapping parameters also corresponds to a sub-vector. Of course, the parameter types and quantities included in each set of mapping parameters can be different.

[0072] After that, the recognition model makes predictions on the first vector to obtain the second vector formed by multiple sub-vectors. Optionally, when the recognition model makes predictions on the first vector, a pre-set string or a randomly generated string can be used as the starting string, and the text encoder or language predictor included in the recognition model is used to make predictions according to the language rules to screen out multiple sub-vectors, and then form the second vector, that is, obtain the corresponding chart preparation information.

[0073] It should be noted that when different methods are used to obtain the second vector, the chart preparation information determined for the same chart to be recognized may be the same or different. For example, for the chart to be recognized a, if the second vector is determined by the method of screening candidate vectors that match the first vector as the second vector through the recognition model, the determined chart preparation information is that application C generates the chart to be recognized according to parameter 2; if the second vector is determined by the method of making predictions on the first vector through the recognition model to obtain the second vector formed by multiple sub-vectors, the determined chart preparation information is that the chart to be recognized is generated according to parameter 3 through programming language S.

[0074] As one example, Figure 3 The method flow chart shows taking the first vector as the input of the recognition model to enable the recognition model to calculate the second vector from the first vector, where the specific steps include S301 - S303.

[0075] S301, obtain the supplementary information determined by the user, and the supplementary information can describe the type and / or attributes of the chart to be recognized.

[0076] S302, input the supplementary information into the recognition model to enable the recognition model to convert the supplementary information into a third vector.

[0077] S303, calculate the first vector and the third vector through the recognition model to obtain the second vector.

[0078] In another example, when the recognition model calculates the first vector, supplementary information determined by the user can also be obtained, and this supplementary information can describe the type and / or attributes of the chart to be recognized. For example, the user intuitively obtains the type of the chart to be recognized, such as a histogram, a waveform chart, a three-row two-column table, etc.

[0079] Further, after obtaining the supplementary information, the supplementary information is input into the recognition model, so that the text encoder included in the recognition model converts the supplementary information into a third vector, and then the recognition model calculates the first vector and the third vector to obtain a second vector. By combining the supplementary information determined by the user, the accuracy of the obtained second vector, that is, the chart preparation information, can be improved.

[0080] It should be noted that when calculating the second vector by the recognition model for the first vector and the third vector, the method of screening candidate vectors matching the first vector and the third vector as the second vector by the recognition model can also be adopted, or the method of predicting the first vector and the third vector by the recognition model to obtain a second vector formed by multiple sub-vectors can be adopted. Details are not elaborated here.

[0081] As an example, Figure 4 The flowchart of the method for generating the recognition model is shown, and the specific steps include S401 - S403.

[0082] S401, obtain a training sample set, where the training sample set includes multiple chart samples and multiple text samples.

[0083] S402, input the chart samples and text samples into the recognition model to be trained to obtain a theoretical pairing result, and the theoretical pairing result includes the pairing scores of each chart vector and each text vector.

[0084] S403, determine the trained recognition model based on the theoretical pairing result and the actual pairing result.

[0085] The embodiments of the present application also provide the steps for generating the recognition model. Specifically, first obtain a training sample set, where the training sample set includes multiple chart samples and multiple text samples. Of course, one or more text samples corresponding to the chart samples may exist in the multiple text samples, or there may be no text samples corresponding to the chart samples.

[0086] After that, the chart samples and text samples are input into the recognition model to be trained. The image encoder included in the recognition model is used to convert each chart sample into a chart vector, and the text encoder included in the recognition model is used to convert each text sample into a text vector. Then, a matching calculation is performed on the chart vector and the text vector to obtain a theoretical pairing result, which includes the pairing scores of each chart vector and each text vector. Here, the recognition model to be trained is an initially constructed model without parameter adjustment.

[0087] Optionally, it can be set that the recognition model includes a binary classifier, which is used to perform a matching calculation on the chart vector and the text vector to obtain a theoretical pairing result. As an example, the theoretical pairing result includes 0 and 1. For example, when the chart vector and the text vector match, the corresponding pairing score of the two is 1; when the chart vector and the text vector do not match, the corresponding pairing score of the two is 0. Of course, the embodiments of the present application are not limited thereto. At this time, the schematic diagram of training the recognition model can be referred to Figure 5 , Figure 4 which shows that the recognition model includes an image encoder and a text encoder, and the part in the dashed box is the theoretical pairing result.

[0088] Correspondingly, it can be set that the recognition model includes a language predictor, which is used to perform text prediction according to language rules to form a complete and clear text description. When training the recognition model, the language predictor predicts the chart vector according to language rules to obtain the text vector corresponding to the chart vector, that is, to obtain the theoretical pairing result.

[0089] After obtaining the theoretical pairing result based on the recognition model to be trained, a trained recognition model is determined based on the theoretical pairing result and the actual pairing result. Here, the actual pairing result includes sample pairs formed by each chart sample and its matching text sample, and the actual pairing result can be used as the label of each chart sample.

[0090] As an example, when determining the trained recognition model based on the theoretical pairing result and the actual pairing result, a loss function can be used to calculate the theoretical pairing result and the actual pairing result to obtain a calculation result. Here, the loss function can be a cross-entropy loss function to determine the difference between the theoretical pairing result and the actual pairing result based on the cross-entropy loss function.

[0091] When the calculation result meets the preset conditions, a trained recognition model is determined; when the calculation result does not meet the preset conditions, the parameters of the recognition model are adjusted until the calculation result meets the preset conditions. Here, the preset condition is that the difference between the theoretical pairing result and the actual pairing result is less than a threshold. For example, the cross-entropy being lower than the preset threshold means meeting the preset conditions, etc.

[0092] Through the pre-trained recognition model, the embodiments of the present application can accurately identify the chart preparation information corresponding to the chart to be recognized, that is, identify the preparation tools and drawing parameters corresponding to the chart to be recognized, without relying on the literature to which the chart to be recognized belongs, and have high flexibility; moreover, there is no need to traverse a large number of charts, effectively improving the recognition efficiency.

[0093] Based on the same inventive concept, the second aspect of the present application further provides an apparatus for recognizing chart preparation information. Since the principle of solving problems by the apparatus in the present application is similar to the above-mentioned method for recognizing chart preparation information in the present application, the implementation of the apparatus can refer to the implementation of the method, and the repeated parts will not be elaborated.

[0094] See Figure 6 As shown, the apparatus for recognizing chart preparation information includes:

[0095] An acquisition module 601 configured to acquire the chart to be recognized;

[0096] A determination module 602 configured to determine the chart preparation information corresponding to the chart to be recognized based on a pre-trained recognition model, where the chart preparation information at least includes preparation tools and drawing parameters.

[0097] In another embodiment, the determination module 602 is specifically configured to:

[0098] Input the chart to be recognized into the recognition model, so that the recognition model converts the chart to be recognized into a first vector;

[0099] Calculate the first vector through the recognition model to obtain a second vector;

[0100] Convert the second vector into the chart preparation information corresponding to the chart to be recognized.

[0101] In another embodiment, the determination module 602 is further configured to:

[0102] Acquire supplementary information determined by the user, where the supplementary information can describe the type and / or attributes of the chart to be recognized;

[0103] Input the supplementary information into the recognition model, so that the recognition model converts the supplementary information into a third vector;

[0104] Calculate the first vector and the third vector through the recognition model to obtain a second vector.

[0105] In another embodiment, the recognition apparatus further includes:

[0106] A storage module 604, configured to obtain a plurality of chart preparation information; convert each of the chart preparation information into a candidate vector capable of characterizing it through the recognition model, and store it;

[0107] The determination module 602 is further configured to screen candidate vectors that match the first vector through the recognition model as the second vector.

[0108] In another embodiment, the recognition device further includes:

[0109] A conversion module 605, configured to obtain a plurality of preparation tools and a plurality of mapping parameters;

[0110] Convert each preparation tool and each set of mapping parameters into sub-vectors respectively through the recognition model;

[0111] The determination module 602 is further configured to perform prediction on the first vector through the recognition model to obtain a second vector formed by a plurality of sub-vectors.

[0112] In another embodiment, the recognition device further includes a generation module 603, configured to:

[0113] Obtain a training sample set, where the training sample set includes a plurality of chart samples and a plurality of text samples;

[0114] Input the chart samples and the text samples into the recognition model to be trained to obtain a theoretical pairing result, where the theoretical pairing result includes the pairing scores of each chart vector and each text vector;

[0115] Determine a trained recognition model based on the theoretical pairing result and the actual pairing result.

[0116] In another embodiment, the generation module 603 is further configured to:

[0117] Calculate the theoretical pairing result and the actual pairing result using a loss function to obtain a calculation result;

[0118] When the calculation result meets a preset condition, determine that a trained recognition model is obtained;

[0119] When the calculation result does not meet the preset condition, adjust the parameters of the recognition model until the calculation result meets the preset condition.

[0120] Through the pre-trained recognition model of the present application, the chart preparation information corresponding to the chart to be recognized can be accurately recognized, that is, the preparation tools and mapping parameters corresponding to the chart to be recognized can be recognized, without relying on the literature to which the chart to be recognized belongs, and the flexibility is relatively high; moreover, there is no need to traverse a large number of charts, effectively improving the recognition efficiency.

[0121] The third aspect of the present application also provides a storage medium, which is a computer-readable medium and stores a computer program. When the computer program is executed by a processor, it implements the method provided in any embodiment of the present application, including the following steps:

[0122] S11, obtaining a chart to be recognized;

[0123] S12, based on a pre-trained recognition model, determining chart preparation information corresponding to the chart to be recognized, where the chart preparation information at least includes a preparation tool and mapping parameters.

[0124] Through the pre-trained recognition model, the present application can accurately recognize the chart preparation information corresponding to the chart to be recognized, that is, recognize the preparation tool and mapping parameters corresponding to the chart to be recognized, without relying on the literature to which the chart to be recognized belongs, and has high flexibility; moreover, there is no need to traverse a large number of charts, effectively improving the recognition efficiency.

[0125] It should be noted that the above storage medium of the present application may be a computer-readable signal medium, a computer-readable storage medium, or any combination of the two. A computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the above. More specific examples of a computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present application, a computer-readable storage medium may be any tangible medium that contains or stores a program, and the program can be used by or in combination with an instruction execution system, apparatus, or device. In the present application, a computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, which carries computer-readable program code. Such a propagated data signal may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. A computer-readable signal medium may also be any storage medium other than a computer-readable storage medium, and the computer-readable signal medium can send, propagate, or transmit a program for use by or in combination with an instruction execution system, apparatus, or device. The program code contained on the storage medium can be transmitted by any suitable medium, including but not limited to: wires, optical cables, RF (radio frequency), etc., or any suitable combination of the above.

[0126] The fourth aspect of the present application also provides an electronic device, such asFigure 7 As shown in the figure, the electronic device at least includes a memory 701 and a processor 702. A computer program is stored on the memory 701. When the processor 702 executes the computer program on the memory 701, the method provided by any embodiment of the present application is implemented. Exemplarily, the method executed by the computer program of the electronic device is as follows:

[0127] S21, obtain the chart to be recognized;

[0128] S22, based on a pre-trained recognition model, determine the chart preparation information corresponding to the chart to be recognized, where the chart preparation information at least includes a preparation tool and drawing parameters.

[0129] Through the pre-trained recognition model of the present application, the chart preparation information corresponding to the chart to be recognized can be accurately recognized, that is, the preparation tool and drawing parameters corresponding to the chart to be recognized are recognized, without relying on the literature to which the chart to be recognized belongs, and the flexibility is relatively high; moreover, there is no need to traverse a large number of charts, effectively improving the recognition efficiency.

[0130] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architectures, functions, and operations of systems, methods, and computer program products according to various embodiments of the present application. In this regard, each block in the flowchart or block diagram may represent a module, a program segment, or a part of code, and this module, program segment, or part of code contains one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order than marked in the accompanying drawings. For example, two consecutive blocks shown may actually be executed substantially in parallel, and they may sometimes be executed in the reverse order, depending on the functions involved. It should also be noted that each block in the block diagram and / or flowchart, and the combination of blocks in the block diagram and / or flowchart, can be implemented by a dedicated hardware-based system for performing the specified functions or operations, or can be implemented by a combination of dedicated hardware and computer instructions.

[0131] The above description is only a preferred embodiment of the present application and an explanation of the technical principles applied. Those skilled in the art should understand that the scope of disclosure involved in the present application is not limited to the technical solutions formed by the specific combination of the above technical features, and should also cover other technical solutions formed by any combination of the above technical features or their equivalent features without departing from the above disclosure concept. For example, the technical solutions formed by mutually replacing the above features with technical features having similar functions (but not limited to) disclosed in the present application.

[0132] In addition, although the operations are depicted in a particular order, this should not be construed as requiring that the operations be performed in the particular order shown or in a sequential order. In certain circumstances, multitasking and parallel processing may be advantageous. Similarly, although several specific implementation details are included in the foregoing description, these should not be construed as limiting the scope of the present application. Certain features that are described in the context of separate embodiments may also be implemented in combination in a single embodiment. Conversely, the various features that are described in the context of a single embodiment may also be implemented separately or in any suitable sub-combination in multiple embodiments.

[0133] Although the subject matter has been described in language specific to structural features and / or methodological acts, it is to be understood that the subject matter defined in the appended claims is not necessarily limited to the specific features or acts described above. Rather, the specific features and acts described above are merely example forms of implementing the claims.

[0134] The foregoing has described in detail multiple embodiments of the present application, but the present application is not limited to these specific embodiments. Those skilled in the art can make various variations and modifications to the embodiments on the basis of the concept of the present application, and these variations and modifications should all fall within the scope claimed by the present application.

Claims

1. A method for identifying chart preparation information, characterized in that: include: Get the chart to be identified; Based on the pre-trained recognition model, the chart preparation information corresponding to the chart to be recognized is determined, wherein the chart preparation information at least includes a preparation tool and a drawing parameter.

2. The identification method according to claim 1, characterized in that: The determining, based on the pre-trained recognition model, the chart preparation information corresponding to the chart to be recognized includes: Inputting the to-be-recognized graph into the recognition model, so that the recognition model converts the to-be-recognized graph into a first vector; Calculate the first vector using the recognition model to obtain a second vector; The second vector is converted into chart preparation information corresponding to the chart to be identified.

3. The identification method according to claim 2, characterized in that: The step of calculating the first vector by using the recognition model to obtain the second vector includes: Acquire supplementary information determined by a user, where the supplementary information can describe the type and / or attribute of the chart to be identified; Inputting the supplementary information into the recognition model so that the recognition model converts the supplementary information into a third vector; The first vector and the third vector are calculated by using the recognition model to obtain a second vector.

4. The identification method according to claim 2, characterized in that: Before determining the chart preparation information corresponding to the chart to be identified based on the pre-trained recognition model, the method further includes: Obtain multiple chart preparation information; Convert each of the chart preparation information into a candidate vector capable of representing the same through the recognition model, and store the candidate vector; The calculating the first vector by using the recognition model to obtain the second vector includes: A candidate vector matching the first vector is selected through the recognition model as the second vector.

5. The identification method according to claim 2, characterized in that: Before determining the chart preparation information corresponding to the chart to be identified based on the pre-trained recognition model, the method further includes: Obtain multiple preparation tools and multiple mapping parameters; Each preparation tool and each set of mapping parameters are converted into sub-vectors respectively by the recognition model; The calculating the first vector by using the recognition model to obtain the second vector includes: The first vector is predicted by the recognition model to obtain a second vector formed by a plurality of sub-vectors.

6. The identification method according to claim 1, characterized in that: It also includes the steps to generate a recognition model: Acquire a training sample set, wherein the training sample set includes a plurality of chart samples and a plurality of text samples; Inputting the chart sample and the text sample into a recognition model to be trained to obtain a theoretical pairing result, wherein the theoretical pairing result includes a pairing score between each chart vector and each text vector; Based on the theoretical pairing results and the actual pairing results, a trained recognition model is determined.

7. The identification method according to claim 6, characterized in that: The step of determining a trained recognition model based on the theoretical pairing result and the actual pairing result includes: Calculating the theoretical pairing result and the actual pairing result using a loss function to obtain a calculation result; When the calculation result meets the preset conditions, determining to obtain a trained recognition model; When the calculation result does not meet the preset condition, the parameters of the recognition model are adjusted until the calculation result meets the preset condition.

8. A device for identifying chart preparation information, characterized in that: include: An acquisition module configured to acquire a chart to be identified; The determination module is configured to determine the chart preparation information corresponding to the chart to be identified based on a pre-trained recognition model, wherein the chart preparation information at least includes a preparation tool and drawing parameters.

9. A storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the method for identifying chart preparation information as claimed in any one of claims 1 to 7 are executed.

10. An electronic device, characterized in that: include: A processor and a memory, wherein the memory stores machine-readable instructions executable by the processor, and when the electronic device is running, the processor and the memory communicate via a bus, and when the machine-readable instructions are executed by the processor, the steps of the method for identifying chart preparation information as described in any one of claims 1 to 7 are performed.

Images