Image recognition and processing method, device, equipment and storage medium

Abstract

The disclosed embodiments disclose an image recognition and processing method, device, equipment and storage medium, wherein the image recognition and processing method comprises: acquiring an image of a text page captured by a camera device, recognizing a first character data set and a second character data set in the image based on a recognition model; extracting a feature vector of the first character data set based on a feature vector extraction model, matching a third character data set corresponding to the feature vector from a preset text library, and acquiring a fourth character data set corresponding to the third character data set; inputting the second character data set and the fourth character data set into a pre-trained text matching model, and outputting matching result information. The disclosed embodiments provide an image recognition and processing method, device, terminal and readable medium, which are used to achieve comprehensive correction of exercises, help students strengthen their knowledge mastery, and improve user experience.

Description

Technical Field

[0001] The embodiments of the present disclosure relate to image processing technology, and more particularly to an image recognition and processing method, apparatus, device and storage medium. Background Art

[0002] Students in the school education stage do a lot of exercises every day. After the students finish the exercises, teachers or parents need to spend a lot of energy to check and correct the exercises. At present, there are some teaching products on the market that can help teachers or parents complete the correction of exercises. The shortcomings of the existing technology at least include: only correcting the results of the exercises, not helping students to strengthen their knowledge, and poor user experience. Summary of the invention

[0003] In view of this, the embodiments of the present disclosure provide an image recognition and processing method, apparatus, device and storage medium for realizing comprehensive correction of exercises, helping students to strengthen their knowledge mastery and improving user experience.

[0004] In a first aspect, the present disclosure provides an image recognition and processing method, including:

[0005] Acquire an image of a text page captured by a camera device, and recognize a first character data set and a second character data set in the image based on a recognition model;

[0006] Extracting a feature vector of the first character data set based on a feature vector extraction model, matching a third character data set corresponding to the feature vector from a preset text library, and acquiring a fourth character data set corresponding to the third character data set;

[0007] The second character data set and the fourth character data set are input into a pre-trained text matching model, and matching result information is output.

[0008] In a second aspect, an embodiment of the present disclosure provides an image recognition and processing device, including:

[0009] A recognition module, used for acquiring an image of a text page captured by a camera device, and recognizing a first character data set and a second character data set in the image based on a recognition model;

[0010] a feature extraction module, configured to extract a feature vector of the first character data set based on a feature vector extraction model, match a third character data set corresponding to the feature vector from a preset text library, and obtain a fourth character data set corresponding to the third character data set;

[0011] The matching module is used to input the second character data set and the fourth character data set into a pre-trained text matching model, and output matching result information.

[0012] In a third aspect, an embodiment of the present disclosure provides a terminal device, including:

[0013] one or more processors;

[0014] A memory for storing one or more programs;

[0015] When the one or more programs are executed by the one or more processors, the one or more processors implement the image recognition and processing method as described in any embodiment of the present disclosure.

[0016] In a fourth aspect, an embodiment of the present disclosure provides a computer storage medium on which a computer program is stored, and when the program is executed by a processor, the image recognition and processing method as described in any embodiment of the present disclosure is implemented.

[0017] The disclosed embodiments provide an image recognition and processing method, apparatus, device and storage medium. The terminal can obtain an image of a text page (e.g., an exercise page completed by a student) through a camera device, and recognize a first character data set (e.g., a question stem in an exercise page) and a second character data set (e.g., a student's handwritten answer steps in an exercise page) in the image based on a recognition model; the terminal can extract a feature vector of the first character set based on a feature vector extraction model, match a third character data set corresponding to the feature vector from a preset text library (e.g., a question stem in a preset exercise library), and obtain a fourth character data set corresponding to the third character data set (e.g., a standard step for a reference answer in a preset exercise library); the second character data set and the fourth character data set are input into a pre-trained text matching model, and matching result information is output.

[0018] Through the recognition model, the terminal can recognize the image of the exercise page or test paper page to obtain the question stem and answer steps of the image; through the feature vector extraction model, the terminal can extract the feature vector of the question stem, match the feature vector of the question stem with the feature vector in the preset text library, obtain the corresponding question stem in the preset text library, and obtain the standard steps of the reference answer corresponding to the question stem in the text library; through the pre-trained text matching model, the terminal can match the answer steps with the standard steps and output the matching results, thereby realizing comprehensive correction of exercises, helping students to strengthen their knowledge mastery and improve user experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the embodiments of the present disclosure or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present disclosure. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0020] Figure 1 A flow chart of an image recognition and processing method provided by Embodiment 1 of the present disclosure is shown;

[0021] Figure 2 A flow chart of an image recognition and processing method provided by Embodiment 2 of the present disclosure is shown;

[0022] Figure 3 A schematic diagram of the structure of an image recognition and processing device provided in Embodiment 3 of the present disclosure is shown;

[0023] Figure 4 A schematic diagram of the hardware structure of a terminal device provided in Embodiment 4 of the present disclosure is shown. DETAILED DESCRIPTION

[0024] Embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although certain embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be construed as being limited to the embodiments described herein, which are instead provided for a more thorough and complete understanding of the present disclosure. It should be understood that the drawings and embodiments of the present disclosure are only for exemplary purposes and are not intended to limit the scope of protection of the present disclosure.

[0025] It should be understood that the various steps described in the method embodiments of the present disclosure may be performed in different orders and / or in parallel. In addition, the method embodiments may include additional steps and / or omit the steps shown. The scope of the present disclosure is not limited in this respect.

[0026] The term "including" and its variations used herein are open inclusions, i.e., "including but not limited to". The term "based on" means "based at least in part on". The term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one additional embodiment"; the term "some embodiments" means "at least some embodiments". The relevant definitions of other terms will be given in the following description.

[0027] It should be noted that the concepts such as "first" and "second" mentioned in the present disclosure are only used to distinguish different devices, modules or units, and are not used to limit the order or interdependence of the functions performed by these devices, modules or units.

[0028] It should be noted that the modifications of "one" and "plurality" mentioned in the present disclosure are illustrative rather than restrictive, and those skilled in the art should understand that unless otherwise clearly indicated in the context, it should be understood as "one or more".

[0029] The names of the messages or information exchanged between multiple devices in the embodiments of the present disclosure are only used for illustrative purposes and are not used to limit the scope of these messages or information.

[0030] Embodiment 1

[0031] Figure 1 A flow chart of an image recognition and processing method provided by the first embodiment of the present disclosure is shown. The present embodiment of the present disclosure can be applied to situations of image recognition and processing. For example, after students complete exercise training, the image of the exercise page completed by the students is recognized by the image recognition and processing method provided by the present embodiment of the present disclosure to obtain the student's handwritten answer steps, and the answer steps are further matched with the standard steps of the reference answer to obtain matching result information. The method can be implemented by a terminal, such as a tutoring terminal, and can be implemented specifically by software and / or hardware in the terminal.

[0032] like Figure 1 As shown, the image recognition and processing method provided in the embodiment of the present disclosure includes the following steps:

[0033] S110: Acquire an image of a text page captured by a camera device, and recognize a first character data set and a second character data set in the image based on a recognition model.

[0034] After receiving the image acquisition instruction, the terminal can control the camera (such as a scanner or camera) to acquire the image of the text page according to the image acquisition instruction. The image acquisition instruction can be triggered when the terminal detects the pressing operation of a preset physical key, or after detecting a preset touch gesture on the touch screen, or when detecting a preset voice acquired by the voice acquisition module, or when receiving a communication message sent by other terminals that are connected to the terminal for communication, which is not specifically limited here.

[0035] Among them, the acquisition range of the image of the captured text page can be pre-set, for example, it can be pre-set to a rectangular range of 150mm*100mm; or, the image of the maximum range can be pre-captured by an image acquisition device first; the page corner points in the image of the maximum range are identified; the page range is determined by the page corner points, and the determined page range is used as the acquisition range of the image of the text page.

[0036] The text page may be an exercise page or a test paper page of an exercise book, etc. The first character data set may be at least one question stem in the exercise page or the test paper page, and the second character data set may be the answer steps of at least one question stem written by the user in the exercise page or the test paper page. The recognition model may include one or more sub-recognition models, and each sub-recognition model may be obtained by training the character samples in the training set based on a deep convolutional neural network (CNN).

[0037] Optionally, recognizing a first character data set and a second character data set in an image based on a recognition model includes: performing region recognition on the image based on a region recognition model to obtain a first region and a second region; and recognizing the first region and the second region based on the character recognition model to generate a first character data set and a second character data set.

[0038] The recognition model may include a region recognition model and a character recognition model. The region recognition model may be a model based on a neural network, which uses a trained region recognition model to extract a two-dimensional feature vector from a text page, generates anchor points of different shapes in each grid of the two-dimensional feature vector, and uses a labeling box to label the detected first region and second region. The first region may be understood as the region of printed characters in the text page, and the second region may be understood as the region of handwritten characters in the text page.

[0039] Wherein, the character recognition model can be a model based on a neural network, for example, an optical character recognition (OCR) model, and the character data sets of each region can be recognized by using the trained character recognition model. Wherein, the character recognition model can be established based on a dilated convolution model and an attention model, and the dilated convolution model can be used to extract features from the first region and the second region, and the extracted features can be decoded into characters by the attention model. Wherein, the character recognition model can include a recognition model for printed characters and a recognition model for handwritten characters. According to the character recognition model for printed characters, the printed characters in the first region can be recognized to obtain a first character data set; according to the character recognition model for handwritten characters, the printed characters in the second region can be recognized to obtain a second character data set. Wherein, the recognition model for printed characters and the recognition model for handwritten characters can be trained separately.

[0040] Optionally, recognizing a first character data set and a second character data set in an image based on a recognition model includes: recognizing the character data set in the image based on a character recognition model; and distinguishing the first character data set and the second character data set according to feature parameters of the character data set in the image.

[0041] The character recognition model can be the same as the above-mentioned character recognition model. After the character recognition model completes the recognition of the printed and handwritten character data sets in the image, the printed and handwritten character data sets can be distinguished according to the characteristic parameters of the character data sets in the image, that is, the first character data set and the second character data set can be distinguished.

[0042] Optionally, the feature parameters include at least one of the following: style, size, and spacing.

[0043] Among them, the style of the character data set in the image can be understood as the font of the character, for example, it can include font standards such as Songti, Kaiti and Heiti; the size of the character data set in the image can be understood as the font size of the character, for example, the font size can include font standards such as size 5, size 4 and size 4; the spacing of the character data set in the image can be understood as the distance between the characters. Usually, the style, size and spacing of the first character data set in the original image all meet certain standards, while the style, size and spacing of the second character data set in the original image are more random, so the first character data set and the second character data set can be distinguished by the characteristics of the style, size and spacing of the character data set in the original image.

[0044] S120, extracting a feature vector of the first character data set based on a feature vector extraction model, matching a third character data set corresponding to the feature vector from a preset text library, and acquiring a fourth character data set corresponding to the third character data set.

[0045] The preset text library may be a preset exercise library or a preset test paper library. The third character data set is a question stem in the preset exercise library or the preset test paper library, and the fourth character data set is a standard step of a reference answer corresponding to each question stem in the preset exercise library or the preset test paper library. After matching the question stem in the preset text library corresponding to the question stem of the text page, the standard step of the reference answer corresponding to the question stem can be obtained from the preset text library, that is, the fourth character data set corresponding to the third character data set is obtained.

[0046] The feature vector extraction model may be a model based on a neural network, such as a sent2vec model. The first character data set is input into a pre-trained feature vector extraction model to obtain a feature vector of the first character data set.

[0047] The preset text library may be, for example, a text library stored in a local preset storage location of the terminal, or a text library stored in a preset storage location on the server. The terminal may access the local preset storage location, or access the preset storage location of the server through the network to query the text library. The preset text library may include question stems and standard steps for reference answers corresponding to each question stem.

[0048] After identifying the question stem in the text page, the feature vector of the question stem can be extracted and matched with the feature vectors of each question stem in the preset text library to obtain the same question stem in the text library, and further obtain the standard steps of the reference answer corresponding to the question stem in the text library, laying the foundation for matching the user's answer steps with the standard steps.

[0049] S130, inputting the second character data set and the fourth character data set into a pre-trained text matching model, and outputting matching result information.

[0050] The text matching model may be a model based on a neural network. Based on the text matching model, the second character data set may be identified as a matching type, and different matching result information may be output according to the matching type. Different matching results may be output for different matching types, thereby achieving a comprehensive match for the second character data set, that is, achieving a comprehensive correction of the answer steps completed by the student.

[0051] Optionally, the matching result information includes at least one of the following: step optimization prompts, step error prompts, calculation error prompts, and idea guidance prompts.

[0052] Among them, when students are solving exercises, their thoughts are blocked during the solution process and they cannot complete the solution; or the intermediate process of the solution steps is correct, but the final result of the solution steps is wrong; or the solution steps are incorrect due to the lack of solid grasp of knowledge points, etc.; or the solution process is correct and the answer result is also correct, but the steps in the solution process are redundant and no convenient algorithm is used, etc. For the above situations, the text matching model can identify the matching types as incomplete solution steps, correct intermediate steps of the solution steps but wrong results, wrong intermediate steps of the solution steps, and correct but redundant solution steps.

[0053] When the matching type is incomplete solution steps, it can be considered that the student's thinking is stuck. At this time, the matching result information can generate a thinking guidance prompt based on the character data set in the fourth character data set except the second character data set, that is, generate a thinking guidance prompt based on the other steps in the standard steps of the reference answer except the handwritten solution steps. Therefore, when the student's thinking is stuck, guiding guidance can be provided through the thinking guidance prompt, which helps the student to establish a complete answering idea.

[0054] When the matching type is that the middle step of the solution is correct but the result is wrong, it can be considered that the student is careless in the calculation. In this case, the matching result information can be a prompt of the calculation error to urge students to develop the habit of doing the questions carefully and attentively.

[0055] When the matching type is an error in the middle step of the solution (for example, when the formula is wrong), it can be considered that the student has not mastered the knowledge point firmly. At this time, the matching result information can be a step error prompt generated based on the unmastered knowledge point to help students check for missing knowledge points.

[0056] When the matching type is that the solution steps are correct but redundant (for example, a simple calculation formula is used in the standard problem-solving steps, but the multi-step derivation steps of the simple calculation formula are redundant in the solution steps, then the solution steps are redundant), the matching result information can be optimization prompts for the steps generated according to the simple calculation formula to improve students' problem-solving efficiency.

[0057] In addition, when the matching type is that the answer steps are correct and concise, the matching result information can also be an encouraging reminder to improve students' learning enthusiasm. Different matching result information can be output for different answer situations such as correct, wrong, and incomplete exercises, which can achieve comprehensive and individualized matching for exercises, which can not only guide students to think, but also enhance their grasp of knowledge points.

[0058] Optionally, the text matching model training method includes: training the text matching model based on a fourth character data set, a sample set of second character data sets corresponding to the fourth character data set, and preset matching result information for each second character data set in the sample set.

[0059] The preset matching result can be the matching type and matching result information input after manually correcting each second character data set. Exemplarily, the training method of the text matching model is: the reference answer of Exercise 1 is standard step 1, the sample set of 500 students' solution steps for Exercise 1, and the matching type and matching result after the teacher corrects the 500 solution steps are input into the text matching model for training. After the trained text matching model is input with standard step 1 and any solution step, it can output matching result information.

[0060] Before performing the text matching model, the second character data set may be preprocessed. The preprocessing may include extracting and shuffling the formulas in the second character data set, such as equivalently exchanging or transforming the contents on both sides of the equal sign of the formula (e.g., transforming ab+ac into a(b+c), etc.). Through the preprocessing, the matching of the second character data set with the fourth character data set may be made more accurate.

[0061] In addition, the output matching information can be rendered into an image of a text page, and the rendered text page image can be displayed through a display device to guide students' learning more intuitively.

[0062] An image recognition and processing method provided by an embodiment of the present disclosure is that a terminal can obtain an image of a text page (such as an exercise page completed by a student) through a camera device, and recognize a first character data set (such as a question stem in an exercise page) and a second character data set (such as a student's handwritten answer steps in the exercise page) in the image based on a recognition model; the terminal can extract a feature vector of the first character set based on a feature vector extraction model, match a third character data set corresponding to the feature vector from a preset text library (such as a question stem in a preset exercise library), and obtain a fourth character data set corresponding to the third character data set (such as the standard steps of a reference answer in a preset exercise library); the second character data set and the fourth character data set are input into a pre-trained text matching model, and matching result information is output.

[0063] Through the recognition model, the terminal can recognize the image of the exercise page or test paper page to obtain the question stem and answer steps of the image; through the feature vector extraction model, the terminal can extract the feature vector of the question stem, match the feature vector of the question stem with the feature vector in the preset text library, obtain the corresponding question stem in the preset text library, and obtain the standard steps of the reference answer corresponding to the question stem in the text library; through the pre-trained text matching model, the terminal can match the answer steps with the standard steps and output the matching results, thereby realizing comprehensive correction of exercises, helping students to strengthen their knowledge mastery and improve user experience.

[0064] Embodiment 2

[0065] Based on the above embodiment, this embodiment adds a step of screening the fourth character data set by the terminal, which belongs to the same inventive concept as the image recognition and processing method proposed in the above embodiment. For technical details not fully described in this embodiment, please refer to the above embodiment.

[0066] Optionally, the number of fourth character data sets is at least two; accordingly, after obtaining the fourth character data set corresponding to the third character data set, it also includes: extracting the feature vector of the second character data set and the feature vectors of at least two fourth character data sets based on the feature vector extraction model; calculating the similarity between the feature vector of the second character data set and the feature vectors of at least two fourth character data sets; based on the similarity, determining the target fourth character data set from the feature vectors of at least two fourth character data sets; accordingly, inputting the second character data set and the fourth character data set into a pre-trained text matching model, including: inputting the second character data set and the target fourth character data set into the pre-trained text matching model.

[0067] Figure 2 The flowchart of an image recognition and processing method provided in the second embodiment of the present disclosure is shown. The image recognition and processing method provided in the embodiment of the present disclosure includes the following steps:

[0068] S210, acquiring an image of a text page captured by a camera device, and recognizing a first character data set and a second character data set in the image based on a recognition model;

[0069] S220, extracting a feature vector of the first character data set based on a feature vector extraction model, matching a third character data set corresponding to the feature vector from a preset text library, and acquiring at least two fourth character data sets corresponding to the third character data set;

[0070] S230, extracting a feature vector of the second character data set and at least two feature vectors of the fourth character data set based on the feature vector extraction model;

[0071] S240, calculating the similarity between the feature vector of the second character data set and the feature vectors of at least two fourth character data sets;

[0072] S250, determining a target fourth character data set from feature vectors of at least two fourth character data sets according to the similarity;

[0073] S260: Input the second character data set and the target fourth character data set into a pre-trained text matching model, and output matching result information.

[0074] When the standard steps of the reference answer corresponding to the question stem are standard steps of different problem-solving ideas, there are at least two fourth character data sets. At this time, the fourth character data set with the same problem-solving idea as the second character data set needs to be screened out to improve the accuracy of the matching result information.

[0075] The feature vector extraction model may be the same as the feature vector model provided in the above embodiment. Calculating the similarity between the feature vector of the second character data set and the feature vectors of at least two fourth character data sets may be calculating the Euclidean distance, Manhattan distance or cosine angle between the feature vector of the second character data set and the feature vectors of at least two fourth character data sets. The fourth character data set with the highest similarity may be used as the target fourth character data set to complete the screening of the fourth character data set.

[0076] Based on the above embodiment, the embodiment of the present disclosure can screen the fourth character data set to ensure the accuracy of the matching result information. In addition, the image recognition and processing methods proposed in the embodiment of the present disclosure and the above embodiment belong to the same inventive concept, and the technical details not described in detail in the embodiment can be referred to the above embodiment, and the embodiment of the present disclosure has the same beneficial effects as the above embodiment.

[0077] Embodiment 3

[0078] Figure 3The third embodiment of the present disclosure provides a schematic diagram of the structure of an image recognition and processing device. The present embodiment is applicable to the situation of image recognition and processing. The image recognition and processing device provided by the present disclosure can realize the image recognition and processing method provided by the above embodiment.

[0079] like Figure 3 As shown, the image recognition and processing device in the embodiment of the present disclosure includes:

[0080] A recognition module 310 is used to obtain an image of a text page captured by a camera device, and recognize a first character data set and a second character data set in the image based on a recognition model;

[0081] A feature extraction module 320 is used to extract a feature vector of the first character data set based on a feature vector extraction model, match a third character data set corresponding to the feature vector from a preset text library, and obtain a fourth character data set corresponding to the third character data set;

[0082] The matching module 330 is used to input the second character data set and the fourth character data set into a pre-trained text matching model, and output matching result information.

[0083] Optional, identification module, including:

[0084] A region recognition submodule, used to perform region recognition on the image based on a region recognition model to obtain a first region and a second region;

[0085] The character recognition submodule is used to recognize the first area and the second area based on the character recognition model to generate a first character data set and a second character data set.

[0086] Optional, identification module, including:

[0087] A character data set recognition submodule, used for recognizing a character data set in an image based on a character recognition model;

[0088] The character data set distinguishing submodule is used to distinguish the first character data set and the second character data set according to the characteristic parameters of the character data set in the image.

[0089] Optionally, the feature parameters include at least one of the following: style, size, and spacing.

[0090] Optionally, the number of the fourth character data sets is at least two;

[0091] Correspondingly, the feature extraction module is further used to extract feature vectors of the second character data set and feature vectors of at least two fourth character data sets based on the feature vector extraction model;

[0092] Accordingly, the image recognition and processing device further includes:

[0093] A similarity calculation module, used to calculate the similarity between the feature vector of the second character data set and the feature vectors of at least two fourth character data sets;

[0094] a target fourth character data set determining module, used to determine a target fourth character data set from feature vectors of at least two fourth character data sets according to similarity;

[0095] Correspondingly, the matching module is used to input the second character data set and the target fourth character data set into a pre-trained text matching model.

[0096] Optionally, the text matching model training method includes: training the text matching model based on a fourth character data set, a sample set of second character data sets corresponding to the fourth character data set, and preset matching result information for each second character data set in the sample set.

[0097] Optionally, the matching result information includes at least one of the following: step optimization prompts, step error prompts, calculation error prompts, and idea guidance prompts.

[0098] The image recognition and processing device provided in the embodiment of the present disclosure belongs to the same inventive concept as the image recognition and processing method provided in the above embodiment. The technical details not fully described in the embodiment of the present disclosure can be referred to the above embodiment, and the embodiment of the present disclosure has the same beneficial effects as the above embodiment.

[0099] Embodiment 4

[0100] Figure 4 The hardware structure diagram of a terminal device provided by the fourth embodiment of the present disclosure is shown. The terminal device in the embodiment of the present disclosure may include but is not limited to mobile terminals such as mobile phones, laptop computers, digital broadcast receivers, PDAs (personal digital assistants), PADs (tablet computers), PMPs (portable multimedia players), vehicle-mounted terminals (such as vehicle-mounted navigation terminals), etc., and fixed terminals such as digital TVs, desktop computers, etc. Figure 4 The electronic device shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present disclosure.

[0101] like Figure 4As shown, the electronic device 400 may include a processing device (e.g., a central processing unit, a graphics processing unit, etc.) 401, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 402 or a program loaded from a storage device 408 to a random access memory (RAM) 403. In the RAM 403, various programs and data required for the operation of the electronic device 400 are also stored. The processing device 401, the ROM 402, and the RAM 403 are connected to each other via a bus 404. An input / output (I / O) interface 405 is also connected to the bus 404.

[0102] Typically, the following devices may be connected to the I / O interface 405: an input device 406 including, for example, a touch screen, a touch pad, a keyboard, a mouse, a camera, a microphone, an accelerometer, a gyroscope, etc.; an output device 407 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; a storage device 408 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 409. The communication device 409 may allow the electronic device 400 to communicate with other devices wirelessly or by wire to exchange data. Although Figure 4 The electronic device 400 is shown with various devices, but it should be understood that it is not required to implement or possess all the devices shown. More or fewer devices may be implemented or possessed instead.

[0103] In particular, according to an embodiment of the present disclosure, the process described above with reference to the flowchart can be implemented as a computer software program. For example, an embodiment of the present disclosure includes a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program contains program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from the network through the communication device 409, or installed from the storage device 408, or installed from the ROM 402. When the computer program is executed by the processing device 401, the above-mentioned functions defined in the method of the embodiment of the present disclosure are executed.

[0104] The terminal provided in the embodiment of the present disclosure and the image recognition and processing method provided in the above embodiment belong to the same inventive concept. The technical details not fully described in the embodiment of the present disclosure can be referred to the above embodiment, and the embodiment of the present disclosure has the same beneficial effects as the above embodiment.

[0105] Embodiment 5

[0106] The embodiments of the present disclosure provide a computer storage medium on which a computer program is stored. When the program is executed by a processor, the image recognition and processing method provided in the above embodiments is implemented.

[0107] It should be noted that the computer storage medium disclosed above may be a computer-readable signal medium or a computer-readable storage medium or any combination of the above two. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with 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 a flash memory (FLASH), 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 disclosure, a computer-readable storage medium may be any tangible medium containing or storing a program, which may be used by or in combination with an instruction execution system, device or device. In the present disclosure, a computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, which carries a computer-readable program code. Such propagated data signals may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. Computer readable signal media may also be any computer readable medium other than computer readable storage media, which may send, propagate, or transmit programs for use by or in conjunction with an instruction execution system, apparatus, or device. The program code contained on the computer readable medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (radio frequency), etc., or any suitable combination of the above.

[0108] In some embodiments, the client and the server may communicate using any currently known or future developed network protocol such as HTTP (HyperText Transfer Protocol), and may be interconnected with any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include a local area network ("LAN"), a wide area network ("WAN"), an internet (e.g., the Internet), and a peer-to-peer network (e.g., an ad hoc peer-to-peer network), as well as any currently known or future developed network.

[0109] The computer storage may be included in the electronic device; or it may exist independently without being installed in the electronic device.

[0110] The computer storage carries one or more programs. When the one or more programs are executed by the electronic device, the electronic device: obtains an image of a text page captured by a camera device, and recognizes a first character data set and a second character data set in the image based on a recognition model; extracts a feature vector of the first character data set based on a feature vector extraction model, matches a third character data set corresponding to the feature vector from a preset text library, and obtains a fourth character data set corresponding to the third character data set; inputs the second character data set and the fourth character data set into a pre-trained text matching model, and outputs matching result information.

[0111] Alternatively, the computer-readable medium carries one or more programs, and when the one or more programs are executed by the electronic device, the electronic device: obtains an image of a text page captured by a camera device, and recognizes a first character data set and a second character data set in the image based on a recognition model; extracts a feature vector of the first character data set based on a feature vector extraction model, matches a third character data set corresponding to the feature vector from a preset text library, and obtains a fourth character data set corresponding to the third character data set; inputs the second character data set and the fourth character data set into a pre-trained text matching model, and outputs matching result information.

[0112] Computer program code for performing the operations of the present disclosure may be written in one or more programming languages, or a combination thereof, including object-oriented programming languages, such as Java, Smalltalk, C++, and conventional procedural programming languages, such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., through the Internet using an Internet service provider).

[0113] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present disclosure. In this regard, each square box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and the module, the program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some implementations as replacements, the functions marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two square boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0114] The units involved in the embodiments described in the present disclosure may be implemented by software or hardware, wherein the name of a unit does not, in some cases, limit the unit itself.

[0115] The functions described above herein may be performed at least in part by one or more hardware logic components. For example, without limitation, exemplary types of hardware logic components that may be used include: field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), complex programmable logic devices (CPLDs), and the like.

[0116] In the context of the present disclosure, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or equipment. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or equipment, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, 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 foregoing.

[0117] According to one or more embodiments of the present disclosure, [Example 1] provides an image recognition and processing method, including:

[0118] Acquire an image of a text page captured by a camera device, and recognize a first character data set and a second character data set in the image based on a recognition model;

[0119] Extracting a feature vector of the first character data set based on a feature vector extraction model, matching a third character data set corresponding to the feature vector from a preset text library, and acquiring a fourth character data set corresponding to the third character data set;

[0120] The second character data set and the fourth character data set are input into a pre-trained text matching model, and matching result information is output.

[0121] According to one or more embodiments of the present disclosure, [Example 2] provides an image recognition and processing method, further comprising:

[0122] Optionally, the identifying the first character data set and the second character data set in the image based on the recognition model includes:

[0123] Performing region recognition on the image based on a region recognition model to obtain a first region and a second region;

[0124] The first region and the second region are recognized based on a character recognition model to generate a first character data set and a second character data set.

[0125] According to one or more embodiments of the present disclosure, [Example 3] provides an image recognition and processing method, further comprising:

[0126] Optionally, the identifying the first character data set and the second character data set in the image based on the recognition model includes:

[0127] Recognize a character dataset in the image based on a character recognition model;

[0128] A first character data set and a second character data set are distinguished according to feature parameters of the character data set in the image.

[0129] According to one or more embodiments of the present disclosure, [Example 4] provides an image recognition and processing method, further comprising:

[0130] Optionally, the characteristic parameter includes at least one of the following: style, size and spacing.

[0131] According to one or more embodiments of the present disclosure, [Example 5] provides an image recognition and processing method, further comprising:

[0132] Optionally, the number of the fourth character data sets is at least two;

[0133] Correspondingly, after obtaining the fourth character data set corresponding to the third character data set, the method further includes:

[0134] Extracting a feature vector of the second character data set and feature vectors of at least two fourth character data sets based on the feature vector extraction model;

[0135] Calculating the similarity between the feature vector of the second character data set and the feature vectors of the at least two fourth character data sets;

[0136] determining a target fourth character data set from feature vectors of at least two fourth character data sets according to the similarity;

[0137] Accordingly, the step of inputting the second character data set and the fourth character data set into a pre-trained text matching model includes:

[0138] The second character data set and the target fourth character data set are input into a pre-trained text matching model.

[0139] According to one or more embodiments of the present disclosure, [Example 6] provides an image recognition and processing method, further comprising:

[0140] Optionally, the text matching model is trained by:

[0141] The text matching model is trained according to a fourth character data set, a sample set of second character data sets corresponding to the fourth character data set, and preset matching result information for each second character data set in the sample set.

[0142] According to one or more embodiments of the present disclosure, [Example 7] provides an image recognition and processing method, further comprising:

[0143] Optionally, the matching result information includes at least one of the following: step optimization prompts, step error prompts, calculation error prompts and idea guidance prompts.

[0144] According to one or more embodiments of the present disclosure, [Example 8] provides an image recognition and processing device, including:

[0145] A recognition module, used for acquiring an image of a text page captured by a camera device, and recognizing a first character data set and a second character data set in the image based on a recognition model;

[0146] a feature extraction module, configured to extract a feature vector of the first character data set based on a feature vector extraction model, match a third character data set corresponding to the feature vector from a preset text library, and obtain a fourth character data set corresponding to the third character data set;

[0147] The matching module is used to input the second character data set and the fourth character data set into a pre-trained text matching model, and output matching result information.

[0148] The above description is only a preferred embodiment of the present disclosure and an explanation of the technical principles used. Those skilled in the art should understand that the scope of disclosure involved in the present disclosure is not limited to the technical solutions formed by a specific combination of the above technical features, but should also cover other technical solutions formed by any combination of the above technical features or their equivalent features without departing from the above disclosed concept. For example, the above features are replaced with the technical features with similar functions disclosed in the present disclosure (but not limited to) by each other to form a technical solution.

[0149] In addition, although each operation is described in a specific order, this should not be understood as requiring these operations to be performed in the specific order shown or in a sequential order. Under certain circumstances, multitasking and parallel processing may be advantageous. Similarly, although some specific implementation details are included in the above discussion, these should not be interpreted as limiting the scope of the present disclosure. Some features described in the context of a separate embodiment can also be implemented in a single embodiment in combination. On the contrary, the various features described in the context of a single embodiment can also be implemented in multiple embodiments individually or in any suitable sub-combination mode.

[0150] Although the subject matter has been described in language specific to structural features and / or methodological logical actions, it should be understood that the subject matter defined in the appended claims is not necessarily limited to the specific features or actions described above. On the contrary, the specific features and actions described above are merely example forms of implementing the claims.

Claims

1. An image recognition and processing method, characterized in that: include: Acquire an image of a text page captured by a camera device, and recognize a first character data set and a second character data set in the image based on a recognition model, wherein the first character data set is a question stem in an exercise page or an examination paper page; and the second character data set is a user's handwritten answering steps for the question stem in the exercise page or the examination paper page; Extracting a feature vector of the first character data set based on a feature vector extraction model, matching a third character data set corresponding to the feature vector from a preset text library, and acquiring a fourth character data set corresponding to the third character data set; Inputting the second character data set and the fourth character data set into a pre-trained text matching model, and outputting matching result information; The third character data set is a question stem in a preset question bank or a preset test paper bank, and the fourth character data set is a standard step of a reference answer corresponding to each question stem in the preset question bank or the preset test paper bank; the matching result information includes at least one of the following: a step optimization prompt, a step error prompt, a calculation error prompt, a thought guidance prompt, and an encouraging prompt; Among them, the matching types identified by the text matching model are incomplete solution steps, correct intermediate steps of the solution steps but wrong results, wrong intermediate steps of the solution steps, correct solution steps but redundant, and correct and concise solution steps.

2. The method according to claim 1, characterized in that The step of recognizing the first character data set and the second character data set in the image based on the recognition model includes: Performing region recognition on the image based on a region recognition model to obtain a first region and a second region; The first region and the second region are recognized based on a character recognition model to generate a first character data set and a second character data set.

3. The method according to claim 1, characterized in that The step of recognizing the first character data set and the second character data set in the image based on the recognition model includes: Recognize a character dataset in the image based on a character recognition model; A first character data set and a second character data set are distinguished according to feature parameters of the character data set in the image.

4. The method according to claim 3, characterized in that The characteristic parameters include at least one of the following: style, size and spacing.

5. The method according to claim 1, characterized in that The number of the fourth character data sets is at least two; Correspondingly, after obtaining the fourth character data set corresponding to the third character data set, the method further includes: Extracting a feature vector of the second character data set and feature vectors of at least two fourth character data sets based on the feature vector extraction model; Calculating the similarity between the feature vector of the second character data set and the feature vectors of the at least two fourth character data sets; determining a target fourth character data set from feature vectors of at least two fourth character data sets according to the similarity; Accordingly, the step of inputting the second character data set and the fourth character data set into a pre-trained text matching model includes: The second character data set and the target fourth character data set are input into a pre-trained text matching model.

6. The method according to claim 1, characterized in that The training method of the text matching model includes: The text matching model is trained according to a fourth character data set, a sample set of second character data sets corresponding to the fourth character data set, and preset matching result information for each second character data set in the sample set.

7. An image recognition and processing device, characterized in that: include: A recognition module is used to obtain an image of a text page captured by a camera device, and recognize a first character data set and a second character data set in the image based on a recognition model, wherein the first character data set is a question stem in an exercise page or an examination paper page; and the second character data set is a user's handwritten answering steps for the question stem in the exercise page or the examination paper page; a feature extraction module, configured to extract a feature vector of the first character data set based on a feature vector extraction model, match a third character data set corresponding to the feature vector from a preset text library, and obtain a fourth character data set corresponding to the third character data set; A matching module, used for inputting the second character data set and the fourth character data set into a pre-trained text matching model, and outputting matching result information; The third character data set is a question stem in a preset question bank or a preset test paper bank, and the fourth character data set is a standard step of a reference answer corresponding to each question stem in the preset question bank or the preset test paper bank; the matching result information includes at least one of the following: a step optimization prompt, a step error prompt, a calculation error prompt, a thought guidance prompt, and an encouraging prompt; Among them, the matching types identified by the text matching model are incomplete solution steps, correct intermediate steps of the solution steps but wrong results, wrong intermediate steps of the solution steps, correct solution steps but redundant, and correct and concise solution steps.

8. A terminal device, characterized in that: The terminal comprises: one or more processors; A memory for storing one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement the image recognition and processing method as described in any one of claims 1-6.

9. A computer storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the image recognition and processing method as described in any one of claims 1 to 6 is implemented.

Images