CN116189219A

CN116189219A - Policy identification method, device, equipment and computer readable storage medium

Info

Publication number: CN116189219A
Application number: CN202310088705.2A
Authority: CN
Inventors: 胡冰
Original assignee: Guangzhou Yingshang Information Technology Co ltd
Current assignee: Guangzhou Yingshang Information Technology Co ltd
Priority date: 2023-01-19
Filing date: 2023-01-19
Publication date: 2023-05-30

Abstract

Embodiments of the present disclosure provide a policy identification method, apparatus, device, and computer-readable storage medium. The method comprises the following steps: uploading a policy picture to be identified to a first OCR recognition model of the client; identifying the policy picture through the first OCR identification model to obtain a first policy identification text; extracting first key information according to the first policy identification text and the policy picture; judging whether the first key information meets the preset text requirement or not; if the images do not accord with the first OCR recognition model, uploading the images to a second OCR recognition model of a server so as to recognize the images through the second OCR recognition model. In this way, the policy key information can be prevented from being manually determined by the insurance user or the insurance business personnel, the determination efficiency and accuracy of the policy key information are improved, and the subsequent insurance business can be rapidly executed.

Description

Policy identification method, device, equipment and computer readable storage medium

Technical Field

The present disclosure relates to the field of insurance, and in particular, to the technical field of policy identification.

Background

At present, in order to determine whether the purchased insurance product is suitable, an insurance user needs to take out the insurance policy of the purchased insurance, and then inform insurance service personnel of key information in the insurance policy, but the insurance user needs to highly cooperate to ensure that the uploaded key information is accurate, but in reality, many insurance users cannot determine which key information needs to be uploaded, even cannot find the key information, which is not beneficial to continuing to execute subsequent insurance services (such as insurance consultation service, insurance assessment service and the like), and the analysis mode of the key information of the insurance policy obviously needs a large amount of manual operation, and has extremely low efficiency.

Disclosure of Invention

The present disclosure provides a policy identification method, apparatus, device, and computer-readable storage medium.

According to a first aspect of the present disclosure, a policy identification method is provided. The method comprises the following steps:

uploading a policy picture to be identified to a first OCR recognition model of the client;

identifying the policy picture through the first OCR identification model to obtain a first policy identification text;

extracting first key information according to the first policy identification text and the policy picture;

judging whether the first key information meets the preset text requirement or not;

if the images do not accord with the first OCR recognition model, uploading the images to a second OCR recognition model of a server so as to recognize the images through the second OCR recognition model.

In the aspect and any possible implementation manner described above, there is further provided an implementation manner, where the extracting the first key information according to the first policy identification text and the policy picture includes:

extracting feature information in the policy picture, wherein the feature information comprises: texture information and/or color information;

and carrying out information classification on the first policy identification text according to the characteristic information so as to extract the first key information.

In the foregoing aspect and any possible implementation manner, there is further provided an implementation manner, where the determining whether the first key information meets a preset text requirement includes:

performing preset judgment on the first key information to obtain a first judgment result, wherein the preset judgment comprises at least one of the following steps: information type judgment, information quantity judgment and information length judgment;

and if the first judging result meets the preset text requirement, judging that the first key information meets the preset text requirement, otherwise, judging that the first key information does not meet the preset text requirement.

Aspects and any one of the possible implementations as described above, further providing an implementation, the first OCR recognition model is a compressed model of the second OCR recognition model.

In aspects and any one of the possible implementations described above, there is further provided an implementation, the first OCR recognition model is acquired by:

obtaining a compression rate of a preset model;

acquiring a weight coefficient threshold value and a weight coefficient type corresponding to the preset model compression rate;

deleting the nodes of which the node weight coefficients are lower than the weight coefficient threshold value in the second OCR recognition model, and performing type conversion on the weight coefficients of the nodes in the second OCR recognition model according to the weight coefficient types so as to obtain the first OCR recognition model.

In accordance with aspects and any one of the possible implementations described above, there is further provided an implementation in which the predetermined model compression rate is dependent on a recognition rate and an accuracy of the first OCR recognition model.

Aspects and any one of the possible implementations as described above, further providing an implementation of uploading the policy picture to a second OCR recognition model of a server to recognize the policy picture through the second OCR recognition model, including:

uploading the encrypted policy picture to the second OCR recognition model to recognize the policy picture through the second OCR recognition model to obtain a second policy recognition text, wherein the server is used for extracting second key information according to the second policy recognition text and the policy picture; judging whether the second key information meets the preset text requirement or not, encrypting a second judging result and then transmitting the second judging result to the client;

and receiving the encrypted second judgment result issued by the server.

According to a second aspect of the present disclosure, a policy identification device is provided. The device comprises:

the uploading module is used for uploading the policy picture to be identified to a first OCR identification model of the client;

the identification module is used for identifying the policy picture through the first OCR identification model to obtain a first policy identification text;

the extraction module is used for extracting first key information according to the first policy identification text and the policy picture;

the judging module is used for judging whether the first key information accords with a preset text requirement or not;

and the processing module is used for uploading the policy picture to a second OCR recognition model of the server if the policy picture does not accord with the policy picture, so as to recognize the policy picture through the second OCR recognition model.

According to a third aspect of the present disclosure, an electronic device is provided. The electronic device includes: a memory and a processor, the memory having stored thereon a computer program, the processor implementing the method as described above when executing the program.

According to a fourth aspect of the present disclosure, there is provided a computer readable storage medium having stored thereon a computer program which when executed by a processor implements a method according to the first and/or second aspects of the present disclosure.

In the method, the first policy identification text is obtained by uploading the policy picture to be identified to the first OCR identification model of the client, the policy picture can be automatically identified through the first OCR identification model, then the first key information in the policy is extracted according to the first policy identification text and the policy picture, and then whether the first key information meets the preset text requirement is judged, if yes, the first key information is directly output, and if not, the policy picture is uploaded to the second OCR identification model of the server, so that the policy picture is identified through the second OCR identification model, and therefore, the policy key information can be prevented from being determined manually by an insurance user or insurance service personnel, the determination efficiency and accuracy of the policy key information are obviously improved, and the method is beneficial to rapidly executing subsequent insurance services.

It should be understood that what is described in this summary is not intended to limit the critical or essential features of the embodiments of the disclosure nor to limit the scope of the disclosure. Other features of the present disclosure will become apparent from the following description.

Drawings

The above and other features, advantages and aspects of embodiments of the present disclosure will become more apparent by reference to the following detailed description when taken in conjunction with the accompanying drawings. For a better understanding of the present disclosure, and without limiting the disclosure thereto, the same or similar reference numerals denote the same or similar elements, wherein:

FIG. 1 illustrates a flow chart of a policy identification method according to an embodiment of the present disclosure;

FIG. 2 illustrates a flow chart of another policy identification method according to an embodiment of the disclosure;

FIG. 3 illustrates a block diagram of a policy identification device according to an embodiment of the disclosure;

fig. 4 illustrates a block diagram of an exemplary electronic device capable of implementing embodiments of the present disclosure.

Detailed Description

For the purposes of making the objects, technical solutions and advantages of the embodiments of the present disclosure more apparent, the technical solutions of the embodiments of the present disclosure will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present disclosure, and it is apparent that the described embodiments are some embodiments of the present disclosure, but not all embodiments. All other embodiments, which can be made by one of ordinary skill in the art based on the embodiments in this disclosure without inventive faculty, are intended to be within the scope of this disclosure.

In addition, the term "and/or" herein is merely an association relationship describing an association object, and means that three relationships may exist, for example, a and/or B may mean: a exists alone, A and B exist together, and B exists alone. In addition, the character "/" herein generally indicates that the front and rear associated objects are an "or" relationship.

Fig. 1 shows a flowchart of a policy identification method 100 according to an embodiment of the present disclosure. The method 100 may include:

step 110, uploading a policy picture to be identified to a first OCR recognition model of the client;

OCR English is called Optical Character Recognition (OCR for short) and is characterized by that it utilizes optical technology and computer technology to read out the characters printed or written on the paper, and converts them into a format which can be accepted by computer and can be understood by human body.

The policy pictures can be one or more pictures formed by shooting the policy of the insurance purchased by the insurance user by using the image acquisition device such as a camera, a camera and the like, and the policy of the insurance purchased by the insurance user can be one or more, can be a part of the policy of the insurance or can be all the policies of the insurance.

Insurance of the present disclosure includes, but is not limited to, social insurance and business insurance, wherein:

social insurance includes senior insurance, medical insurance, uneconomic insurance, industrial injury insurance, and fertility insurance;

commercial insurance is classified into property insurance and life insurance, wherein property insurance is classified into three kinds of insurance types, namely property loss insurance, liability insurance and credit guarantee insurance.

The first OCR recognition model and the second OCR recognition model can be realized through a neural network, namely, the first OCR recognition model and the second OCR recognition model are obtained through training the neural network, and the training can be performed according to an AI (Artificial Intelligence ) technology and a large number of sample policy pictures.

Step 120, identifying the policy picture through the first OCR identification model to obtain a first policy identification text;

before the first OCR recognition model recognizes the policy picture, the first OCR recognition model may be used to correct an input direction of the policy picture, for example, a direction of inputting the policy picture into the preset OCR recognition model is unified into a horizontal input or a vertical input.

In addition, the policy picture can be preprocessed before being input into the first OCR recognition model; wherein the pretreatment includes, but is not limited to: one or more of noise reduction processing, binarization processing, sharpening processing, inclination correction processing, gray scale processing and the like, which is beneficial to improving the image quality of the policy picture and extracting the first key information more accurately.

Finally, before the policy picture is input into the first OCR recognition model, the policy picture may be compressed by using a target picture compression algorithm to reduce the recognition load of the client, but the compression methods of the policy picture are various, and different compression methods may affect the recognition accuracy of the first OCR recognition model to different extents, so when the target picture compression algorithm is selected, the target picture compression algorithm may be selected in the following manner to reduce adverse effects on the recognition accuracy of the first OCR recognition model as much as possible:

acquiring a plurality of sample pictures;

respectively selecting different picture compression algorithms for compressing a plurality of sample pictures to obtain compressed pictures respectively corresponding to the plurality of sample pictures;

inputting each compressed picture into a first OCR recognition model respectively to obtain key information of each compressed picture;

counting the identification accuracy of key information of each compressed picture;

and taking a picture compression algorithm for identifying the compressed picture with the highest accuracy as the target picture compression algorithm.

The recognition accuracy of the key information of each compressed picture can be calculated by comparing the key information of each compressed picture extracted by the first OCR recognition model with the key information extracted from the sample picture corresponding to each compressed picture by using the first OCR recognition model.

The image compression algorithm is used for compressing the image by reducing the space and resolution occupied by the image, and the format of the image may be changed before and after compression.

Step 130, extracting first key information according to the first policy identification text and the policy picture;

the first key information may be: name, sex, age, policy type, identification number, date of birth, insurance payment mode, period, amount, name, address, and policy restriction terms.

In addition, the first OCR recognition model and the second OCR recognition model can be determined according to insurance companies and insurance policy types, namely, OCR recognition models corresponding to different insurance policy types under different insurance companies can be different, so that different OCR recognition models can be selected to extract key information according to the current insurance companies and the current insurance policy types in the insurance policy pictures, and the accuracy of the key information can be improved. The insurance policy type is used for representing dangerous seeds, for example, the insurance policy type of the pension insurance is the pension insurance dangerous seed, and the insurance policy type of the serious disease insurance is the serious disease dangerous seed.

Step 140, judging whether the first key information meets the preset text requirement;

the preset text requirement may be an information length requirement, an information quantity requirement, or the like.

And step 150, if the images do not accord with the images, uploading the images to a second OCR recognition model of a server so as to recognize the images through the second OCR recognition model.

The method comprises the steps of uploading a policy picture to be identified to a first OCR identification model of a client, automatically identifying the policy picture through the first OCR identification model to obtain a first policy identification text, extracting first key information in the policy according to the first policy identification text and the policy picture, judging whether the first key information meets preset text requirements or not, directly outputting the first key information if the first key information meets the preset text requirements, and uploading the policy picture to a second OCR identification model of a server if the first key information does not meet the preset text requirements, so that the policy picture is identified through the second OCR identification model, thereby avoiding an insurance user or an insurance business personnel from manually determining the policy key information, obviously improving the determination efficiency and accuracy of the policy key information and being beneficial to rapidly executing subsequent insurance business.

In addition, the policy picture is uploaded to a first OCR recognition model of the client, and is uploaded to a second OCR recognition model on the basis that the first OCR recognition model cannot accurately extract first key information, but is directly uploaded to the second OCR recognition model, so that the OCR service can be centralized, the need of deploying the OCR service on a server is avoided, the OCR service can be deployed to the client in a lightweight manner, the deployment cost of the centralized server can be reduced, the calculated amount of the server is reduced, the server is prevented from being down due to the receipt of a large number of policy recognition requests of insurance users, most of policy recognition is not centralized in the calculation of the client, the cost of deploying the centralized server is greatly reduced, the original 100% centralized OCR service of the OCR service is reduced to 5%, privacy leakage caused by uploading the policy picture to the server is also reduced, and personal privacy data safety is ensured.

In some embodiments, the extracting the first key information according to the first policy identification text and the policy picture includes:

the feature information includes but is not limited to texture information, color information, and may also include SIFT features of the picture, where the SIFT features are Scale Invariant Feature Transform, also called scale invariant features, and finally represent the input image as a 128-dimensional feature vector set, where the SIFT features have characteristics of rotation, scaling, translation, illumination invariant, and the like.

Because the characteristic information of the same key information in different types of insurance policies of the same insurance company often has a certain rule, as the texture and color characteristics of names in different types of insurance policies of an insurance company may be the same, or even the characteristic information of the same key information under different types of insurance policies of different insurance companies often has a certain rule, as the texture and color characteristics of names of insurance companies under different types of insurance policies of different insurance companies may be the same, a preset relationship exists between the characteristic information and the key information and/or the types of the key information, so that the information classification is performed on the first insurance policy identification text according to the preset relationship and the characteristic information in the extracted insurance policy picture, and whether the different information in the first insurance policy identification text is the name, the sex, the insurance restriction term or other types of information is determined, so that the first key information can be accurately extracted.

In some embodiments, the determining whether the first key information meets a preset text requirement includes:

The preset text requirement is one or a combination of information type requirement, information quantity requirement and information length requirement.

By carrying out preset judgment on the first key information, whether the first key information meets the preset text requirement or not can be accurately determined, if yes, the fact that the first key information extracted by the first OCR recognition model is correct is indicated, the fact that the policy picture is uploaded to the second OCR recognition model on the server is not needed, if not, the fact that the first key information extracted by the first OCR recognition model is wrong is indicated, and the fact that the policy picture is uploaded to the second OCR recognition model on the server is needed.

In some embodiments, the first OCR recognition model is a compressed model of the second OCR recognition model.

The first OCR recognition model is a compression model of the second OCR recognition model, namely the first OCR recognition model is a light-weight model of the second OCR recognition model, so that the OCR recognition model running on the client is different from the OCR recognition model running on a common server, the OCR recognition model can be ensured to normally run on the client to correctly recognize the policy picture, and the comprehensive cost of OCR service can be reduced by more than 10 times.

In some embodiments, the first OCR recognition model is obtained by:

obtaining a compression rate of a preset model;

the preset model compression rate refers to a model compression rate of the first OCR recognition model with respect to the second OCR recognition model and/or a model compression rate of the first OCR recognition model with respect to the original OCR recognition model.

The preset model compression rate is used for indicating how much the second OCR recognition model needs to be compressed to obtain the proper light-weight first OCR recognition model.

The first OCR recognition model and the second OCR recognition model can be realized through a neural network, the neural network is provided with a plurality of convolution layers, and the number of the convolution layers is unchanged when the neural network is compressed, but the number of nodes and the weight coefficient type are converted. While the nodes in the second OCR recognition model, namely neurons, and the node weight coefficients are used for representing the connection strength between the neurons, the smaller the node weight coefficients of two neurons are, the lower the connection strength between the two neurons is.

The weight coefficient types can be floating point type, integer type (such as int type), decimal type and the like, and the weight coefficient type corresponding to the preset model compression rate is preferably a data type with small occupied space for compressing the OCR model.

the corresponding weight coefficient threshold value and weight coefficient type can be accurately determined by setting the association relation among the model compression rate, the weight coefficient threshold value and the weight coefficient type in advance and combining the association relation with the preset model compression rate.

The compression rate of the preset model is obtained, a weight coefficient threshold value and a weight coefficient type corresponding to the compression rate of the preset model can be determined, then nodes with node weight coefficients lower than the weight coefficient threshold value in the second OCR recognition model are deleted, namely nodes with smaller connection strength in the second OCR recognition model are deleted, and the weight coefficients of the nodes in the second OCR recognition model are automatically type-converted according to the weight coefficient type, so that the occupied space of the OCR recognition model is further reduced, a first light-weight OCR recognition model is obtained, the fact that the OCR recognition model can normally operate on a client side to correctly recognize policy pictures as much as possible is ensured, and the fact that the policy pictures need to be recognized by using the OCR recognition model on a server is avoided as much as possible is avoided, so that the comprehensive cost of OCR service can be reduced by more than 10 times.

In some embodiments, the predetermined model compression rate is a function of the recognition rate and accuracy of the first OCR recognition model.

The predetermined model compression rate may be determined according to the recognition rate and accuracy of the first OCR recognition model, in particular,

the second OCR recognition model can be compressed according to different model compression rates to obtain a plurality of different first OCR recognition models;

respectively inputting the same sample policy pictures into each first OCR recognition model to obtain policy key information output by each first OCR recognition model and recognition duration corresponding to each first OCR recognition model (namely, duration required by outputting the policy key information); wherein the identification time length corresponds to the identification rate, the shorter the identification time length is, the higher the identification rate is, the longer the identification time length is, and the lower the identification rate is

Inputting the same sample policy to a second OCR recognition model to obtain policy key information output by the second OCR recognition model;

matching the policy key information output by each first OCR recognition model with the policy key information output by the second OCR recognition model to obtain the matching degree corresponding to the policy key information output by each first OCR recognition model;

determining the matching degree corresponding to the policy key information output by each first OCR recognition model as the accuracy of each first OCR recognition model;

selecting a final first OCR recognition model according to the accuracy rate of each first OCR recognition model and the recognition duration corresponding to each first OCR recognition model;

and taking the model compression rate corresponding to the final first OCR recognition model as a preset model compression rate.

In one embodiment, selecting the final first OCR recognition model according to the accuracy of each first OCR recognition model and the recognition duration corresponding to each first OCR recognition model includes:

determining an accuracy threshold and an identification duration threshold;

comparing the accuracy rate of each first OCR recognition model with an accuracy rate threshold value, and comparing the recognition duration corresponding to each first OCR recognition model with a recognition duration threshold value;

selecting an OCR recognition model with the accuracy higher than an accuracy threshold and the recognition time longer than a recognition time length threshold from the first OCR recognition models as a candidate OCR recognition model;

if the candidate OCR recognition model includes a plurality of OCR recognition models, one OCR recognition model is randomly selected from the candidate OCR recognition models as a final first OCR recognition model.

Of course, if there is only one OCR recognition model candidate, the OCR recognition model candidate is determined as the final first OCR recognition model.

In some embodiments, uploading the policy picture to a second OCR recognition model of a server to recognize the policy picture through the second OCR recognition model includes:

uploading the encrypted policy picture to the second OCR recognition model to recognize the policy picture through the second OCR recognition model to obtain a second policy recognition text, wherein the server is used for extracting second key information according to the second policy recognition text and the policy picture; judging whether the second key information meets the preset text requirement or not, encrypting a second judging result and then transmitting the second judging result to the client; the algorithm used in encryption may be an asymmetric encryption algorithm.

And receiving the encrypted second judgment result issued by the server. In addition, if the second key information meets the preset text requirement, the server can directly encrypt the second key information and then send the encrypted second key information to the client, and the client receives the encrypted second key information sent by the server; if the second key information does not meet the preset text requirement, the server can also directly send the prompt of re-uploading the policy picture to the client, and the client receives the prompt of re-uploading the policy picture sent by the server.

The second key information can be extracted by encrypting the policy picture and uploading the policy picture to the second OCR recognition model, so that the policy picture is prevented from being leaked in the process of uploading the policy picture to the server, the security of the policy picture is ensured, the second key information can be extracted by utilizing the second OCR recognition model to recognize the policy picture, and then whether the second key information meets the preset text requirement is encrypted and then transmitted to the client, so that the client can clearly and definitely recognize the policy picture finally and correctly.

Specifically, if the second judging result is that the second key information meets the preset text requirement, displaying the second key information through the client according to a preset display condition, wherein the preset display condition includes but is not limited to: and immediately displaying when the second key information is received, displaying when the preset time is reached, displaying when the noise of the current environment of the user is lower than a preset noise threshold value, displaying when the brightness of the current environment of the user is higher than the preset noise threshold value, and the like.

If the second key information does not meet the preset text requirement as a second judging result, prompting the insurance user to upload the policy picture again on the client.

The technical scheme of the present disclosure is described in further detail below with reference to fig. 2:

the client off-line OCR service is a lightweight AI algorithm specially developed for ARM architecture chips, the accuracy comparable to a large AI model and the execution speed are achieved by using less computing resources, a model file (namely an installation package of a first OCR recognition model) of the algorithm is integrated to a mobile client after encryption processing, the first OCR recognition model is operated to be completely off-line, the comprehensive use success rate can reach 95%, and a user does not need to upload any data to the cloud.

As shown in fig. 2, a required OCR picture file is uploaded to a client offline OCR service (i.e., a first OCR recognition model), the first OCR recognition model adjusts the picture direction, determines a region to which key information belongs, performs text recognition on the region to which the key information belongs, obtains a first policy recognition text, and then extracts the first key information in the text, if the first key information meets a preset text requirement, which is a text content to be uploaded by a client, the first key information is displayed on the client, if the first key information does not meet the preset text requirement, a centralized server is requested to perform OCR service on the policy picture, at this time, the policy picture to be uploaded by a user is locally encrypted offline through a self-grinding asymmetric encryption algorithm and then uploaded to the server for OCR recognition, so that data security can be ensured, and after the centralized server recognizes the extracted content (i.e., a second key information), different contents are returned according to whether the second key information meets the requirement.

Specifically, if the user uploads the policy picture which does not meet the requirement, the client is informed to open the guide page to prompt the user to upload the correct policy picture again, if the user uploads the policy picture which meets the requirement, the extracted content (namely the second key information) is encrypted through an asymmetric encryption algorithm and then returned to the client to be automatically filled into the client for display.

The architecture can reduce the original 100% of centralization service to 5%, and the OCR model running on the centralization server (namely, the second OCR recognition model is also a light OCR recognition model, namely, the second OCR recognition model is also a compression model relative to the original OCR recognition model, but the second OCR recognition model is installed on the server, the processing capacity of the server is higher than that of the client, therefore, the model compression rate of the second OCR recognition model relative to the original OCR recognition model is higher than that of the first OCR recognition model relative to the second OCR recognition model, if the preset model compression rate is 80%, the model compression rate of the second OCR recognition model relative to the original OCR recognition model is possibly 90% or a value higher than 80%) is also designed to be light, the second OCR recognition model can not run on a CPU server with high cost like other servers, but can run on a common CPU server, and the comprehensive cost can be reduced by more than 10 times. The framework can be enough for users to ensure personal privacy data security, has good support degree for each mainstream platform, and is win-win for enterprises and users in combination.

It should be noted that, for simplicity of description, the foregoing method embodiments are all described as a series of acts, but it should be understood by those skilled in the art that the present disclosure is not limited by the order of acts described, as some steps may be performed in other orders or concurrently in accordance with the present disclosure. Further, those skilled in the art will also appreciate that the embodiments described in the specification are all alternative embodiments, and that the acts and modules referred to are not necessarily required by the present disclosure.

The foregoing is a description of embodiments of the method, and the following further describes embodiments of the present disclosure through examples of apparatus.

Fig. 3 shows a block diagram of a policy identification device 300 according to an embodiment of the disclosure. As shown in fig. 3, the apparatus 300 includes:

an uploading module 310, configured to upload a policy picture to be identified to a first OCR recognition model of the client;

the recognition module 320 is configured to recognize the policy picture through the first OCR recognition model to obtain a first policy recognition text;

an extracting module 330, configured to extract first key information according to the first policy identification text and the policy picture;

a judging module 340, configured to judge whether the first key information meets a preset text requirement;

and the processing module 350 is configured to upload the policy picture to a second OCR recognition model of a server if the policy picture does not match, so as to recognize the policy picture through the second OCR recognition model.

It will be clear to those skilled in the art that, for convenience and brevity of description, specific working procedures of the described modules may refer to corresponding procedures in the foregoing method embodiments, which are not described herein again.

The present disclosure also provides, in accordance with embodiments of the present disclosure, an electronic device and a non-transitory computer-readable storage medium storing computer instructions.

Fig. 4 shows a schematic block diagram of an electronic device 400 that may be used to implement embodiments of the present disclosure. Electronic devices are intended to represent various forms of digital computers, such as laptops, desktops, workstations, personal digital assistants, servers, blade servers, mainframes, and other appropriate computers. The electronic device may also represent various forms of mobile devices, such as personal digital processing, cellular telephones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions, are meant to be exemplary only, and are not meant to limit implementations of the disclosure described and/or claimed herein.

The device 400 comprises a computing unit 401 that may perform various suitable actions and processes in accordance with a computer program stored in a Read Only Memory (ROM) 402 or loaded from a storage unit 408 into a Random Access Memory (RAM) 403. In RAM 403, various programs and data required for the operation of device 400 may also be stored. The computing unit 401, ROM 402, and RAM 403 are connected to each other by a bus 404. An input/output (I/O) interface 405 is also connected to bus 404.

Various components in device 400 are connected to I/O interface 405, including: an input unit 406 such as a keyboard, a mouse, etc.; an output unit 407 such as various types of displays, speakers, and the like; a storage unit 408, such as a magnetic disk, optical disk, etc.; and a communication unit 409 such as a network card, modem, wireless communication transceiver, etc. The communication unit 409 allows the device 400 to exchange information/data with other devices via a computer network, such as the internet, and/or various telecommunication networks.

The computing unit 401 may be a variety of general purpose and/or special purpose processing components having processing and computing capabilities. Some examples of computing unit 401 include, but are not limited to, a Central Processing Unit (CPU), a Graphics Processing Unit (GPU), various specialized Artificial Intelligence (AI) computing chips, various computing units running machine learning model algorithms, a Digital Signal Processor (DSP), and any suitable processor, controller, microcontroller, etc. The computing unit 401 performs the various methods and processes described above, such as method 100. For example, in some embodiments, the method 100 may be implemented as a computer software program tangibly embodied on a machine-readable medium, such as the storage unit 408. In some embodiments, part or all of the computer program may be loaded and/or installed onto the device 400 via the ROM 402 and/or the communication unit 409. One or more of the steps of the method 100 described above may be performed when a computer program is loaded into RAM 403 and executed by the computing unit 401. Alternatively, in other embodiments, the computing unit 401 may be configured to perform the method 100 by any other suitable means (e.g., by means of firmware).

Various implementations of the systems and techniques described here above may be implemented in digital electronic circuitry, integrated circuit systems, field Programmable Gate Arrays (FPGAs), application Specific Integrated Circuits (ASICs), application Specific Standard Products (ASSPs), systems On Chip (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and/or combinations thereof. These various embodiments may include: implemented in one or more computer programs, the one or more computer programs may be executed and/or interpreted on a programmable system including at least one programmable processor, which may be a special purpose or general-purpose programmable processor, that may receive data and instructions from, and transmit data and instructions to, a storage system, at least one input device, and at least one output device.

Program code for carrying out methods of the present disclosure may be written in any combination of one or more programming languages. These program code may be provided to a processor or controller of a general purpose computer, special purpose computer, or other programmable data processing apparatus such that the program code, when executed by the processor or controller, causes the functions/operations specified in the flowchart and/or block diagram to be implemented. The program code may execute entirely on the machine, partly on the machine, as a stand-alone software package, partly on the machine and partly on a remote machine or entirely on the remote machine or server.

In the context of this disclosure, a machine-readable medium may be a tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device. The machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. The machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. More specific examples of a machine-readable storage medium would include an electrical connection based on one or more wires, a portable computer diskette, 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 disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

To provide for interaction with a user, the systems and techniques described here can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to a user; and a keyboard and pointing device (e.g., a mouse or trackball) by which a user can provide input to the computer. Other kinds of devices may also be used to provide for interaction with a user; for example, feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form, including acoustic input, speech input, or tactile input.

The systems and techniques described here can be implemented in a computing system that includes a background component (e.g., as a data server), or that includes a middleware component (e.g., an application server), or that includes a front-end component (e.g., a user computer having a graphical user interface or a web browser through which a user can interact with an implementation of the systems and techniques described here), or any combination of such background, middleware, or front-end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: local Area Networks (LANs), wide Area Networks (WANs), and the internet.

The computing system may include clients and servers. The client and server are typically remote from each other and typically interact through a communication network. The relationship of client and server arises by virtue of computer programs running on the respective computers and having a client-server relationship to each other. The server may be a cloud server, a server of a distributed system, or a server incorporating a blockchain.

It should be appreciated that various forms of the flows shown above may be used to reorder, add, or delete steps. For example, the steps recited in the present disclosure may be performed in parallel or sequentially or in a different order, provided that the desired results of the technical solutions of the present disclosure are achieved, and are not limited herein.

The above detailed description should not be taken as limiting the scope of the present disclosure. It will be apparent to those skilled in the art that various modifications, combinations, sub-combinations and alternatives are possible, depending on design requirements and other factors. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present disclosure are intended to be included within the scope of the present disclosure.

Claims

1. The policy identification method is suitable for the client and is characterized by comprising the following steps:

2. The method of claim 1, wherein the extracting first key information from the first policy identification text and the policy picture comprises:

3. The method of claim 1, wherein determining whether the first key information meets a preset text requirement comprises:

4. The method of claim 1, wherein the step of determining the position of the substrate comprises,

the first OCR recognition model is a compressed model of the second OCR recognition model.

5. The method of claim 4, wherein the step of determining the position of the first electrode is performed,

the first OCR recognition model is obtained by the following steps:

obtaining a compression rate of a preset model;

6. The method of claim 5, wherein the step of determining the position of the probe is performed,

the preset model compression rate is determined according to the recognition rate and the accuracy of the first OCR recognition model.

7. The method of claim 1, wherein the step of determining the position of the substrate comprises,

uploading the policy picture to a second OCR recognition model of a server to recognize the policy picture through the second OCR recognition model, including:

and receiving the encrypted second judgment result issued by the server.

8. A policy identification device adapted for use with a client, comprising:

9. An electronic device, comprising:

at least one processor; and

a memory communicatively coupled to the at least one processor; wherein,,

the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the method of any one of claims 1-7.

10. A non-transitory computer readable storage medium storing computer instructions for causing the computer to perform the method of any one of claims 1-7.