CN113705468A - Digital image identification method based on artificial intelligence and related equipment - Google Patents
Digital image identification method based on artificial intelligence and related equipment Download PDFInfo
- Publication number
- CN113705468A CN113705468A CN202111005704.4A CN202111005704A CN113705468A CN 113705468 A CN113705468 A CN 113705468A CN 202111005704 A CN202111005704 A CN 202111005704A CN 113705468 A CN113705468 A CN 113705468A
- Authority
- CN
- China
- Prior art keywords
- image
- text
- digital
- recognition
- information
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
Abstract
The invention relates to artificial intelligence, and provides a digital image identification method based on artificial intelligence and related equipment. The method includes the steps of obtaining an image to be recognized according to a digital image recognition request, recognizing an image table in the image to be recognized, performing text recognition on the recognized image based on the image table to obtain an image text, extracting a first digital text according to the image table and a key text in the image text, cutting the image to be recognized according to the first digital text to obtain a digital image, inputting the digital image into a digital recognition model to obtain a second digital text, and determining the first digital text or the second digital text as a target digital text if the first digital text is the same as the second digital text. The invention can accurately identify the handwritten numbers in the image. In addition, the invention also relates to a block chain technology, and the target digital text can be stored in the block chain.
Description
Technical Field
The invention relates to the technical field of artificial intelligence, in particular to a digital image identification method based on artificial intelligence and related equipment.
Background
Digital image recognition refers to recognizing digital text contained in an image. With the development of artificial intelligence, digital texts in images are mainly recognized by an ocr (optical character recognition) algorithm at present.
However, in some images, (e.g., house deal assessment images), the digital text contained in these images is often handwritten by the user, which results in current OCR algorithms being unable to accurately recognize handwritten digits in the images because the text handwritten for the same digit by different users differs.
Disclosure of Invention
In view of the foregoing, it is desirable to provide a digital image recognition method and related apparatus based on artificial intelligence, which can accurately recognize handwritten digits in an image.
In one aspect, the present invention provides an artificial intelligence-based digital image recognition method, including:
when a digital image identification request is received, acquiring an image to be identified according to the digital image identification request;
identifying an image table in the image to be identified;
performing text recognition on the image to be recognized based on the image table to obtain an image text;
extracting a first digital text from the image text according to the image table and a key text in the image text;
cutting the image to be recognized according to the first digital text to obtain a digital image;
inputting the digital image into a pre-trained digital recognition model to obtain a second digital text;
and if the first digital text is the same as the second digital text, determining the first digital text or the second digital text as a target digital text.
According to a preferred embodiment of the present invention, before inputting the digital image into a pre-trained digital recognition model to obtain a second digital text, the method further comprises:
constructing a learner, wherein the learner comprises a convolution layer, a pooling layer and a full-connection layer;
acquiring a plurality of handwritten image samples, wherein each handwritten image sample comprises a sample image and a handwritten text in the sample image;
generating an image vector of the sample image from image pixels of the sample image;
performing convolution processing on the image vector based on the convolution layer to obtain convolution characteristics;
inputting the convolution features into the pooling layer to obtain text features;
acquiring a weight matrix in the full-connection layer, and calculating the product of the text features and the weight matrix to obtain a text vector;
mapping the text vector to obtain a predicted text of the learner on the sample image;
comparing each character in the predicted text with each character in the handwritten text, and counting the number of the characters in the predicted text and the number of the characters in the handwritten text to obtain the editing distance between the predicted text and the handwritten text;
counting the total number of characters of the predicted text to obtain a first number, and counting the total number of characters of the handwritten text to obtain a second number;
determining the numerical value with the largest value in the first quantity and the second quantity as a target quantity;
calculating the ratio of the editing distance to the target number to obtain the recognition error rate of the learner on the sample image;
and adjusting learning parameters in the learner according to the recognition error rate until the recognition error rate is not reduced any more, so as to obtain the digital recognition model.
According to a preferred embodiment of the present invention, the acquiring an image to be recognized according to the digital image recognition request includes:
analyzing the message of the digital image identification request to obtain data information carried by the message;
acquiring a storage path and a preset label from the data information;
and acquiring an image corresponding to the preset label from the storage path as the image to be identified.
According to a preferred embodiment of the present invention, the identifying the image table in the image to be identified includes:
carrying out binarization processing on the image to be identified to obtain a binary image;
acquiring an image line segment in the binary image based on a preset function;
and carrying out cross processing on the image line segments according to the line segment positions of the image line segments in the binary image to obtain the image table.
According to a preferred embodiment of the present invention, the image table includes a plurality of cells, the image text includes text information and a cell in which the text information is located, and extracting a first digital text from the image text according to the image table and a key text in the image text includes:
performing semantic analysis on the text information to obtain semantic features of the text information;
determining a text field where the text information is located according to the semantic features;
acquiring a plurality of preset vocabularies from a domain vocabulary according to the text field, and performing word segmentation processing on the text information to obtain a plurality of text segmented words;
extracting a participle which is the same as any preset vocabulary from the plurality of text participles to serve as the key text;
obtaining a cell where the key text is located from the image text as a key cell;
acquiring a cell associated with the key cell from the image table as a target cell;
and extracting the first digital text from the image text according to the target cell.
According to a preferred embodiment of the present invention, the cutting the image to be recognized according to the first digital text to obtain a digital image includes:
obtaining a cell where the first digital text is located from the image text as the target cell;
acquiring coordinate information of the target cell from the image to be identified;
and cutting the image to be recognized according to the coordinate information to obtain a digital image.
According to a preferred embodiment of the invention, the method further comprises:
if the first digital text is different from the second digital text, generating feedback information according to the first digital text and the second digital text;
extracting, when a response result based on the feedback information is received, confirmation information from the response result;
counting the response quantity of the confirmation information which is not the second digital text;
and when the response quantity is larger than the preset quantity, adjusting the model parameters of the digital recognition model according to the confirmation information.
In another aspect, the present invention further provides an artificial intelligence based digital image recognition apparatus, including:
the device comprises an acquisition unit, a recognition unit and a recognition unit, wherein the acquisition unit is used for acquiring an image to be recognized according to a digital image recognition request when the digital image recognition request is received;
the identification unit is used for identifying an image table in the image to be identified;
the identification unit is further used for performing text identification on the image to be identified based on the image table to obtain an image text;
the extraction unit is used for extracting a first digital text from the image text according to the image table and the key text in the image text;
the cutting unit is used for cutting the image to be recognized according to the first digital text to obtain a digital image;
the input unit is used for inputting the digital image into a pre-trained digital recognition model to obtain a second digital text;
the determining unit is further configured to determine the first digital text or the second digital text as a target digital text if the first digital text is the same as the second digital text.
In another aspect, the present invention further provides an electronic device, including:
a memory storing computer readable instructions; and
a processor executing computer readable instructions stored in the memory to implement the artificial intelligence based digital image recognition method.
In another aspect, the present invention also provides a computer-readable storage medium having computer-readable instructions stored therein, which are executed by a processor in an electronic device to implement the artificial intelligence based digital image recognition method.
According to the technical scheme, the first digital text can be accurately extracted from the image text through the image table and the key text, the problem that the second digital text cannot be identified by the digital identification model due to inaccurate extraction of the first digital text is avoided, certain convenience is provided for the digital identification model to identify the second digital text, and the identification accuracy of the target digital text can be improved by further combining the comparison between the first digital text and the second digital text.
Drawings
FIG. 1 is a flow chart of the digital image recognition method based on artificial intelligence of the present invention.
FIG. 2 is a functional block diagram of a digital image recognition device based on artificial intelligence according to a preferred embodiment of the present invention.
FIG. 3 is a schematic structural diagram of an electronic device implementing a digital image recognition method based on artificial intelligence according to a preferred embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention will be described in detail with reference to the accompanying drawings and specific embodiments.
FIG. 1 is a flow chart of the digital image recognition method based on artificial intelligence according to the preferred embodiment of the present invention. The order of the steps in the flow chart may be changed and some steps may be omitted according to different needs.
The digital image recognition method based on artificial intelligence can acquire and process related data based on artificial intelligence technology. Among them, Artificial Intelligence (AI) is a theory, method, technique and application system that simulates, extends and expands human Intelligence using a digital computer or a machine controlled by a digital computer, senses the environment, acquires knowledge and uses the knowledge to obtain the best result.
The artificial intelligence infrastructure generally includes technologies such as sensors, dedicated artificial intelligence chips, cloud computing, distributed storage, big data processing technologies, operation/interaction systems, mechatronics, and the like. The artificial intelligence software technology mainly comprises a computer vision technology, a robot technology, a biological recognition technology, a voice processing technology, a natural language processing technology, machine learning/deep learning and the like.
The Digital image recognition method based on artificial intelligence is applied to one or more electronic devices, which are devices capable of automatically performing numerical calculation and/or information processing according to computer readable instructions set or stored in advance, and the hardware thereof includes, but is not limited to, a microprocessor, an Application Specific Integrated Circuit (ASIC), a Programmable Gate Array (FPGA), a Digital Signal Processor (DSP), an embedded device, and the like.
The electronic device may be any electronic product capable of performing human-computer interaction with a user, for example, a Personal computer, a tablet computer, a smart phone, a Personal Digital Assistant (PDA), a game machine, an interactive Internet Protocol Television (IPTV), a smart wearable device, and the like.
The electronic device may include a network device and/or a user device. Wherein the network device includes, but is not limited to, a single network electronic device, an electronic device group consisting of a plurality of network electronic devices, or a Cloud Computing (Cloud Computing) based Cloud consisting of a large number of hosts or network electronic devices.
The network in which the electronic device is located includes, but is not limited to: the internet, a wide area Network, a metropolitan area Network, a local area Network, a Virtual Private Network (VPN), and the like.
S11, when receiving the digital image identification request, acquiring the image to be identified according to the digital image identification request.
In at least one embodiment of the invention, the digital image recognition request may be triggered to be generated by a user having a text recognition requirement.
The image to be recognized is an image which needs to be subjected to handwritten number recognition. The image to be identified can record domain information in a plurality of domains.
In at least one embodiment of the present invention, the electronic device obtaining the image to be recognized according to the digital image recognition request includes:
analyzing the message of the digital image identification request to obtain data information carried by the message;
acquiring a storage path and a preset label from the data information;
and acquiring an image corresponding to the preset label from the storage path as the image to be identified.
Wherein, the preset label is used for indicating that the image is not subjected to digital identification processing.
By analyzing the message, the storage path and the preset label can be quickly acquired, so that the acquisition efficiency of the image to be identified is improved, and the image to be digitally identified can be accurately acquired through the preset label.
And S12, identifying an image table in the image to be identified.
In at least one embodiment of the present invention, the image table refers to a table in the image to be recognized. The image table includes a plurality of cells.
In at least one embodiment of the invention, the method further comprises:
and preprocessing the image to be recognized.
By preprocessing the image to be recognized, noise information in the image to be recognized can be eliminated, and therefore recognition of the image table and the image text is improved.
In at least one embodiment of the present invention, the electronic device identifying an image table in the image to be identified includes:
carrying out binarization processing on the image to be identified to obtain a binary image;
acquiring an image line segment in the binary image based on a preset function;
and carrying out cross processing on the image line segments according to the line segment positions of the image line segments in the binary image to obtain the image table.
Wherein the preset function may be a HoughLinesP () function.
The line segment position refers to a position of the image line segment in the binary image.
By carrying out binarization processing on the image to be identified, the median value of the binary image comprises binary information, so that the identification accuracy of the image line segment is improved, and the identification accuracy of the image table is improved.
Specifically, the electronic device analyzes the image to be recognized based on a cv2.cvtColor () function to obtain the binary image.
And S13, performing text recognition on the image to be recognized based on the image table to obtain an image text.
In at least one embodiment of the present invention, the image text refers to text information in the image table.
In at least one embodiment of the present invention, the performing, by the electronic device, text recognition on the image to be recognized based on the image table to obtain an image text includes:
positioning the position of the image table in the image to be identified to obtain an image layer;
carrying out corrosion treatment on the image layer according to a corrosion algorithm to obtain a text characteristic layer;
performing expansion processing on the text feature layer by adopting a nearest neighbor search algorithm to obtain a first area;
extracting a plurality of character forming characteristics from the first region, and integrating the character forming characteristics to obtain at least one line of characters;
and cutting all the characters in the at least one line of characters by adopting different cutting intervals based on different characters to obtain the image text.
By positioning the image table, analysis on all pixels in the image to be recognized can be avoided, the recognition efficiency of the image text is improved, and by integrating the plurality of character forming features, the at least one line of characters is cut again, so that the recognition accuracy of the image text can be improved.
S14, extracting a first digital text from the image text according to the image table and the key text in the image text.
In at least one embodiment of the present invention, the key text refers to a label text corresponding to a handwritten numeral in the image text. For example, the key text may be: and (6) summing.
The first digital text refers to handwritten numbers extracted from the image to be recognized based on an OCR algorithm.
In at least one embodiment of the present invention, the image table includes a plurality of cells, the image text includes text information and a cell in which the text information is located, and the extracting, by the electronic device, the first digital text from the image text according to the image table and a key text in the image text includes:
performing semantic analysis on the text information to obtain semantic features of the text information;
determining a text field where the text information is located according to the semantic features;
acquiring a plurality of preset vocabularies from a domain vocabulary according to the text field, and performing word segmentation processing on the text information to obtain a plurality of text segmented words;
extracting a participle which is the same as any preset vocabulary from the plurality of text participles to serve as the key text;
obtaining a cell where the key text is located from the image text as a key cell;
acquiring a cell associated with the key cell from the image table as a target cell;
and extracting the first digital text from the image text according to the target cell.
Wherein the text field may be a house purchase contract field or the like. It is understood that the text fields differ and the corresponding predetermined vocabulary also differs. For example, if the text field is a house purchase contract field, the preset vocabulary may be: total, etc., and the text field is an outpatient service expense list, the preset vocabulary may be: total amount, etc.
By performing semantic analysis on the text information, the determination accuracy of the text field can be improved, further, the key texts in the image text can be comprehensively extracted by comparing the plurality of text participles with the plurality of preset vocabularies, and further, the key cells where the key texts are located can be accurately determined according to the mapping relation between the text information stored in the image text and the cells, so that the extraction accuracy of the first digital text is improved.
Specifically, the semantic analysis of the text information by the electronic device to obtain the semantic features of the text information includes:
obtaining an MLM (masked Language model) network layer and an NSP (Next sequence prefix) network layer from a network library;
splicing the MLM network layer and the NSP network layer to obtain a semantic vector network layer;
and processing the text information by utilizing the semantic vector network layer to obtain the semantic features.
The text information is processed through the MLM network layer and the NSP network layer, so that semantic features with context semantic information can be obtained, and the representation capability of the semantic features on the text information is improved.
Specifically, the determining, by the electronic device, the text field where the text information is located according to the semantic features includes:
acquiring domain characteristics of all domains from a domain library;
calculating the feature similarity of the semantic features and the domain features;
and determining the field corresponding to the field feature with the minimum feature similarity as the text field.
Through the analysis of the feature similarity, the situation that the semantic features cannot be matched with the text field from the field library due to certain deviation is avoided, and the accuracy of the text field is improved.
And S15, cutting the image to be recognized according to the first digital text to obtain a digital image.
In at least one embodiment of the present invention, the digital image refers to an image slice containing the first digital text. It is understood that the digital image belongs to a part of the image to be recognized.
In at least one embodiment of the present invention, the electronic device, according to the first digital text, cutting the image to be recognized to obtain a digital image, includes:
obtaining a cell where the first digital text is located from the image text as the target cell;
acquiring coordinate information of the target cell from the image to be identified;
and cutting the image to be recognized according to the coordinate information to obtain a digital image.
The coordinate information is used for indicating the position of the target cell in the image to be recognized.
Through the coordinate information corresponding to the target unit grid, not only can the picture information contained in the number needing to be identified be completely cut out, but also redundant picture information can be prevented from being cut out from the image to be identified, and therefore the influence on the identification of the second digital text is avoided.
Specifically, the electronic device cuts the image to be recognized according to the coordinate information, and obtaining a digital image includes:
positioning a detection frame in the image to be identified according to the coordinate information;
based on the detection frame, calling a cut () function to cut the image to be identified to obtain the digital image.
And S16, inputting the digital image into a pre-trained digital recognition model to obtain a second digital text.
In at least one embodiment of the present invention, the number recognition model refers to a model for recognizing handwritten numbers in an image.
In at least one embodiment of the present invention, the second digital text refers to a text obtained after the number recognition model recognizes handwritten numbers in the digital image.
In at least one embodiment of the present invention, before inputting the digital image into the pre-trained digital recognition model to obtain the second digital text, the method further comprises:
constructing a learner, wherein the learner comprises a convolution layer, a pooling layer and a full-connection layer;
acquiring a plurality of handwritten image samples, wherein each handwritten image sample comprises a sample image and a handwritten text in the sample image;
generating an image vector of the sample image from image pixels of the sample image;
performing convolution processing on the image vector based on the convolution layer to obtain convolution characteristics;
inputting the convolution features into the pooling layer to obtain text features;
acquiring a weight matrix in the full-connection layer, and calculating the product of the text features and the weight matrix to obtain a text vector;
mapping the text vector to obtain a predicted text of the learner on the sample image;
comparing each character in the predicted text with each character in the handwritten text, and counting the number of the characters in the predicted text and the number of the characters in the handwritten text to obtain the editing distance between the predicted text and the handwritten text;
counting the total number of characters of the predicted text to obtain a first number, and counting the total number of characters of the handwritten text to obtain a second number;
determining the numerical value with the largest value in the first quantity and the second quantity as a target quantity;
calculating the ratio of the editing distance to the target number to obtain the recognition error rate of the learner on the sample image;
and adjusting learning parameters in the learner according to the recognition error rate until the recognition error rate is not reduced any more, so as to obtain the digital recognition model.
Each handwritten image sample comprises a sample image and handwritten text in the sample image. The handwritten text in the sample image includes multiple scripts of any one of the numbers 0-9.
The image pixel refers to a pixel value of each pixel point in the sample image.
The edit distance refers to the number of each character in the predicted text different from the number of characters at the corresponding position in the handwritten text. For example, if the predicted text is 234 and the handwritten text is 236, the edit distance is 1.
The image vectors representing the sample images can be accurately generated through the image pixels, the prediction capability of the learner on the sample images can be accurately determined based on the image vectors, the learning parameters are adjusted through the recognition error rate, and the recognition accuracy of the digital recognition model on the handwritten text is improved. According to the invention, the digital recognition model is generated by the plurality of handwritten images, so that the application scene of the digital recognition model to the handwritten versions of different numbers can be improved, and the recognition accuracy of the digital recognition model to the handwritten numbers is improved. Specifically, the electronic device performs mapping processing on the text vector to obtain the predicted text of the learner on the sample image includes:
acquiring a mapping vector of each digital text in a text mapping table;
calculating the similarity of the text vector and the mapping vector;
and determining the digital text corresponding to the mapping vector with the minimum similarity as the predicted text.
And the text mapping table stores the corresponding relation between a plurality of handwritten numbers and mapping vectors.
Through the implementation mode, the predicted text corresponding to the text vector can be rapidly determined based on the text mapping table.
In at least one embodiment of the present invention, the electronic device inputs the digital image into the digital recognition model, and a manner of obtaining the second digital text is similar to a manner of generating the prediction text of the sample image by the electronic device, which is not described in detail herein.
S17, if the first digital text is the same as the second digital text, determining the first digital text or the second digital text as a target digital text.
In at least one embodiment of the present invention, the target digital text refers to a handwritten number to be recognized from the image to be recognized.
It is emphasized that, to further ensure the privacy and security of the target digital text, the target digital text may also be stored in a node of a blockchain.
In at least one embodiment of the invention, the method further comprises:
if the first digital text is different from the second digital text, generating feedback information according to the first digital text and the second digital text;
extracting, when a response result based on the feedback information is received, confirmation information from the response result;
counting the response quantity of the confirmation information which is not the second digital text;
and when the response quantity is larger than the preset quantity, adjusting the model parameters of the digital recognition model according to the confirmation information.
Wherein the response result may be a confirmation result for the feedback information.
The response quantity refers to the total quantity of the second digital texts generated by the digital recognition model and the real results in the image to be recognized.
The preset number may be set according to recognition accuracy of the digital recognition model.
By the implementation method, when the total number of the second digital text generated by the digital recognition model and the real result in the image to be recognized are different is larger than the preset number, the digital recognition model can be further adjusted to improve the accuracy of the digital recognition model in recognizing the handwritten numbers.
Specifically, the generating, by the electronic device, feedback information according to the first digital text and the second digital text includes:
acquiring a preset message, wherein the preset message comprises a first preset label and a second preset label;
and writing the first digital text into a position corresponding to the first preset label, and writing the second digital text into a position corresponding to the second preset label to obtain the feedback information.
The first preset label is a label corresponding to the identification mode of the first digital text, and the second preset label is a label corresponding to the identification mode of the second digital text.
Through the preset message, the feedback information can be generated quickly, and a user can know the specific identification modes of the first digital text and the second digital text conveniently.
According to the technical scheme, the first digital text can be accurately extracted from the image text through the image table and the key text, the problem that the second digital text cannot be identified by the digital identification model due to inaccurate extraction of the first digital text is avoided, certain convenience is provided for the digital identification model to identify the second digital text, and the identification accuracy of the target digital text can be improved by further combining the comparison between the first digital text and the second digital text.
FIG. 2 is a functional block diagram of a digital image recognition device based on artificial intelligence according to a preferred embodiment of the present invention. The artificial intelligence based digital image recognition device 11 comprises a generation unit 110, an acquisition unit 111, a recognition unit 112, an extraction unit 113, a cutting unit 114, an input unit 115, a determination unit 116, a preprocessing unit 117, a statistics unit 118, an adjustment unit 119, a construction unit 120, a convolution unit 121 and a mapping unit 122. The module/unit referred to herein is a series of computer readable instruction segments that can be accessed by the processor 13 and perform a fixed function and that are stored in the memory 12. In the present embodiment, the functions of the modules/units will be described in detail in the following embodiments.
When receiving a digital image recognition request, the acquisition unit 111 acquires an image to be recognized according to the digital image recognition request.
In at least one embodiment of the invention, the digital image recognition request may be triggered to be generated by a user having a text recognition requirement.
The image to be recognized is an image which needs to be subjected to handwritten number recognition. The image to be identified can record domain information in a plurality of domains.
In at least one embodiment of the present invention, the acquiring unit 111 acquires the image to be recognized according to the digital image recognition request includes:
analyzing the message of the digital image identification request to obtain data information carried by the message;
acquiring a storage path and a preset label from the data information;
and acquiring an image corresponding to the preset label from the storage path as the image to be identified.
Wherein, the preset label is used for indicating that the image is not subjected to digital identification processing.
By analyzing the message, the storage path and the preset label can be quickly acquired, so that the acquisition efficiency of the image to be identified is improved, and the image to be digitally identified can be accurately acquired through the preset label.
The identifying unit 112 identifies an image table in the image to be identified.
In at least one embodiment of the present invention, the image table refers to a table in the image to be recognized. The image table includes a plurality of cells.
In at least one embodiment of the present invention, the preprocessing unit 117 preprocesses the image to be recognized.
By preprocessing the image to be recognized, noise information in the image to be recognized can be eliminated, and therefore recognition of the image table and the image text is improved.
In at least one embodiment of the present invention, the identifying unit 112 identifies an image table in the image to be identified, including:
carrying out binarization processing on the image to be identified to obtain a binary image;
acquiring an image line segment in the binary image based on a preset function;
and carrying out cross processing on the image line segments according to the line segment positions of the image line segments in the binary image to obtain the image table.
Wherein the preset function may be a HoughLinesP () function.
The line segment position refers to a position of the image line segment in the binary image.
By carrying out binarization processing on the image to be identified, the median value of the binary image comprises binary information, so that the identification accuracy of the image line segment is improved, and the identification accuracy of the image table is improved.
Specifically, the identification unit 112 analyzes the image to be identified based on a cv2.cvtcolor () function to obtain the binary image.
The recognition unit 112 performs text recognition on the image to be recognized based on the image table to obtain an image text.
In at least one embodiment of the present invention, the image text refers to text information in the image table.
In at least one embodiment of the present invention, the recognizing unit 112 performs text recognition on the image to be recognized based on the image table, and obtaining an image text includes:
positioning the position of the image table in the image to be identified to obtain an image layer;
carrying out corrosion treatment on the image layer according to a corrosion algorithm to obtain a text characteristic layer;
performing expansion processing on the text feature layer by adopting a nearest neighbor search algorithm to obtain a first area;
extracting a plurality of character forming characteristics from the first region, and integrating the character forming characteristics to obtain at least one line of characters;
and cutting all the characters in the at least one line of characters by adopting different cutting intervals based on different characters to obtain the image text.
By positioning the image table, analysis on all pixels in the image to be recognized can be avoided, the recognition efficiency of the image text is improved, and by integrating the plurality of character forming features, the at least one line of characters is cut again, so that the recognition accuracy of the image text can be improved.
The extracting unit 113 extracts a first digital text from the image text according to the image table and a key text in the image text.
In at least one embodiment of the present invention, the key text refers to a label text corresponding to a handwritten numeral in the image text. For example, the key text may be: and (6) summing.
The first digital text refers to handwritten numbers extracted from the image to be recognized based on an OCR algorithm.
In at least one embodiment of the present invention, the image table includes a plurality of cells, the image text includes text information and a cell in which the text information is located, and the extracting unit 113 extracts a first digital text from the image text according to the image table and a key text in the image text includes:
performing semantic analysis on the text information to obtain semantic features of the text information;
determining a text field where the text information is located according to the semantic features;
acquiring a plurality of preset vocabularies from a domain vocabulary according to the text field, and performing word segmentation processing on the text information to obtain a plurality of text segmented words;
extracting a participle which is the same as any preset vocabulary from the plurality of text participles to serve as the key text;
obtaining a cell where the key text is located from the image text as a key cell;
acquiring a cell associated with the key cell from the image table as a target cell;
and extracting the first digital text from the image text according to the target cell.
Wherein the text field may be a house purchase contract field or the like. It is understood that the text fields differ and the corresponding predetermined vocabulary also differs. For example, if the text field is a house purchase contract field, the preset vocabulary may be: total, etc., and the text field is an outpatient service expense list, the preset vocabulary may be: total amount, etc.
By performing semantic analysis on the text information, the determination accuracy of the text field can be improved, further, the key texts in the image text can be comprehensively extracted by comparing the plurality of text participles with the plurality of preset vocabularies, and further, the key cells where the key texts are located can be accurately determined according to the mapping relation between the text information stored in the image text and the cells, so that the extraction accuracy of the first digital text is improved.
Specifically, the extracting unit 113 performs semantic analysis on the text information, and obtaining the semantic features of the text information includes:
obtaining an MLM (masked Language model) network layer and an NSP (Next sequence prefix) network layer from a network library;
splicing the MLM network layer and the NSP network layer to obtain a semantic vector network layer;
and processing the text information by utilizing the semantic vector network layer to obtain the semantic features.
The text information is processed through the MLM network layer and the NSP network layer, so that semantic features with context semantic information can be obtained, and the representation capability of the semantic features on the text information is improved.
Specifically, the determining, by the extracting unit 113, the text field where the text information is located according to the semantic features includes:
acquiring domain characteristics of all domains from a domain library;
calculating the feature similarity of the semantic features and the domain features;
and determining the field corresponding to the field feature with the minimum feature similarity as the text field.
Through the analysis of the feature similarity, the situation that the semantic features cannot be matched with the text field from the field library due to certain deviation is avoided, and the accuracy of the text field is improved.
The cutting unit 114 cuts the image to be recognized according to the first digital text to obtain a digital image.
In at least one embodiment of the present invention, the digital image refers to an image slice containing the first digital text. It is understood that the digital image belongs to a part of the image to be recognized.
In at least one embodiment of the present invention, the cutting unit 114 cuts the image to be recognized according to the first digital text, and obtaining a digital image includes:
obtaining a cell where the first digital text is located from the image text as the target cell;
acquiring coordinate information of the target cell from the image to be identified;
and cutting the image to be recognized according to the coordinate information to obtain a digital image.
The coordinate information is used for indicating the position of the target cell in the image to be recognized.
Through the coordinate information corresponding to the target unit grid, not only can the picture information contained in the number needing to be identified be completely cut out, but also redundant picture information can be prevented from being cut out from the image to be identified, and therefore the influence on the identification of the second digital text is avoided.
Specifically, the cutting unit 114 cuts the image to be recognized according to the coordinate information, and obtaining a digital image includes:
positioning a detection frame in the image to be identified according to the coordinate information;
based on the detection frame, calling a cut () function to cut the image to be identified to obtain the digital image.
The input unit 115 inputs the digital image into a pre-trained digital recognition model to obtain a second digital text.
In at least one embodiment of the present invention, the number recognition model refers to a model for recognizing handwritten numbers in an image.
In at least one embodiment of the present invention, the second digital text refers to a text obtained after the number recognition model recognizes handwritten numbers in the digital image.
In at least one embodiment of the present invention, before the digital image is input into the pre-trained digital recognition model to obtain the second digital text, the construction unit 120 constructs a learner, which includes a convolutional layer, a pooling layer, and a full-link layer;
the acquiring unit 111 acquires a plurality of handwritten image samples, each handwritten image sample including a sample image and a handwritten text in the sample image;
the generation unit 110 generates an image vector of the sample image from image pixels of the sample image;
the convolution unit 121 performs convolution processing on the image vector based on the convolution layer to obtain convolution characteristics;
the input unit 115 inputs the convolution features into the pooling layer to obtain text features;
the obtaining unit 111 obtains a weight matrix in the full connection layer, and calculates a product of the text feature and the weight matrix to obtain a text vector;
the mapping unit 122 performs mapping processing on the text vector to obtain a predicted text of the learner on the sample image;
the counting unit 118 compares each character in the predicted text with each character in the handwritten text, and counts the number of characters in the predicted text and the number of characters in the handwritten text, so as to obtain the edit distance between the predicted text and the handwritten text;
the counting unit 118 counts the total number of characters of the predicted text to obtain a first number, and counts the total number of characters of the handwritten text to obtain a second number;
the determining unit 116 determines a maximum value of the first number and the second number as a target number;
the statistical unit 118 calculates a ratio of the editing distance to the target number to obtain an identification error rate of the learner on the sample image;
the adjusting unit 119 adjusts the learning parameters in the learner according to the recognition error rate until the recognition error rate is no longer reduced, so as to obtain the digital recognition model.
Each handwritten image sample comprises a sample image and handwritten text in the sample image. The handwritten text in the sample image includes multiple scripts of any one of the numbers 0-9.
The image pixel refers to a pixel value of each pixel point in the sample image.
The edit distance refers to the number of each character in the predicted text different from the number of characters at the corresponding position in the handwritten text. For example, if the predicted text is 234 and the handwritten text is 236, the edit distance is 1.
The image vectors representing the sample images can be accurately generated through the image pixels, the prediction capability of the learner on the sample images can be accurately determined based on the image vectors, the learning parameters are adjusted through the recognition error rate, and the recognition accuracy of the digital recognition model on the handwritten text is improved. According to the invention, the digital recognition model is generated by the plurality of handwritten images, so that the application scene of the digital recognition model to the handwritten versions of different numbers can be improved, and the recognition accuracy of the digital recognition model to the handwritten numbers is improved. Specifically, the mapping unit 122 performs mapping processing on the text vector to obtain the predicted text of the learner on the sample image includes:
acquiring a mapping vector of each digital text in a text mapping table;
calculating the similarity of the text vector and the mapping vector;
and determining the digital text corresponding to the mapping vector with the minimum similarity as the predicted text.
And the text mapping table stores the corresponding relation between a plurality of handwritten numbers and mapping vectors.
Through the implementation mode, the predicted text corresponding to the text vector can be rapidly determined based on the text mapping table.
In at least one embodiment of the present invention, the input unit 115 inputs the digital image into the digital recognition model, and a manner of obtaining the second digital text is similar to a manner of generating the predicted text of the sample image, which is not described in detail herein.
If the first digital text is the same as the second digital text, the determining unit 116 determines the first digital text or the second digital text as a target digital text.
In at least one embodiment of the present invention, the target digital text refers to a handwritten number to be recognized from the image to be recognized.
It is emphasized that, to further ensure the privacy and security of the target digital text, the target digital text may also be stored in a node of a blockchain.
In at least one embodiment of the present invention, if the first digital text is different from the second digital text, the generating unit 110 generates feedback information according to the first digital text and the second digital text;
when receiving a response result based on the feedback information, the extracting unit 113 extracts confirmation information from the response result;
the counting unit 118 counts the number of responses of the confirmation information which is not the second digital text;
when the response number is greater than the preset number, the adjusting unit 119 adjusts the model parameters of the digital recognition model according to the confirmation information.
Wherein the response result may be a confirmation result for the feedback information.
The response quantity refers to the total quantity of the second digital texts generated by the digital recognition model and the real results in the image to be recognized.
The preset number may be set according to recognition accuracy of the digital recognition model.
By the implementation method, when the total number of the second digital text generated by the digital recognition model and the real result in the image to be recognized are different is larger than the preset number, the digital recognition model can be further adjusted to improve the accuracy of the digital recognition model in recognizing the handwritten numbers.
Specifically, the generating unit 110 generates feedback information according to the first digital text and the second digital text, including:
acquiring a preset message, wherein the preset message comprises a first preset label and a second preset label;
and writing the first digital text into a position corresponding to the first preset label, and writing the second digital text into a position corresponding to the second preset label to obtain the feedback information.
The first preset label is a label corresponding to the identification mode of the first digital text, and the second preset label is a label corresponding to the identification mode of the second digital text.
Through the preset message, the feedback information can be generated quickly, and a user can know the specific identification modes of the first digital text and the second digital text conveniently.
According to the technical scheme, the digital recognition model is generated through the plurality of handwritten images, the application scene of the digital recognition model to handwritten versions of different numbers can be improved, the recognition accuracy of the digital recognition model to the handwritten numbers is improved, the first digital text can be accurately extracted from the image text through the image table and the key text, the problem that the digital recognition model cannot recognize the second digital text due to inaccurate extraction of the first digital text is avoided, certain convenience is provided for the digital recognition model to recognize the second digital text, and the recognition accuracy of the target digital text can be improved by further combining the comparison between the first digital text and the second digital text.
Fig. 3 is a schematic structural diagram of an electronic device according to a preferred embodiment of the present invention for implementing an artificial intelligence-based digital image recognition method.
In one embodiment of the present invention, the electronic device 1 includes, but is not limited to, a memory 12, a processor 13, and computer readable instructions stored in the memory 12 and executable on the processor 13, such as an artificial intelligence based digital image recognition program.
It will be appreciated by a person skilled in the art that the schematic diagram is only an example of the electronic device 1 and does not constitute a limitation of the electronic device 1, and that it may comprise more or less components than shown, or some components may be combined, or different components, e.g. the electronic device 1 may further comprise an input output device, a network access device, a bus, etc.
The Processor 13 may be a Central Processing Unit (CPU), other general purpose Processor, a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field Programmable Gate Array (FPGA) or other Programmable logic device, discrete Gate or transistor logic device, discrete hardware component, etc. The processor 13 is an operation core and a control center of the electronic device 1, and is connected to each part of the whole electronic device 1 by various interfaces and lines, and executes an operating system of the electronic device 1 and various installed application programs, program codes, and the like.
Illustratively, the computer readable instructions may be partitioned into one or more modules/units that are stored in the memory 12 and executed by the processor 13 to implement the present invention. The one or more modules/units may be a series of computer readable instruction segments capable of performing specific functions, which are used for describing the execution process of the computer readable instructions in the electronic device 1. For example, the computer readable instructions may be divided into a generation unit 110, an acquisition unit 111, a recognition unit 112, an extraction unit 113, a cutting unit 114, an input unit 115, a determination unit 116, a preprocessing unit 117, a statistics unit 118, an adjustment unit 119, a construction unit 120, a convolution unit 121, and a mapping unit 122.
The memory 12 may be used for storing the computer readable instructions and/or modules, and the processor 13 implements various functions of the electronic device 1 by executing or executing the computer readable instructions and/or modules stored in the memory 12 and invoking data stored in the memory 12. The memory 12 may mainly include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application program required by at least one function (such as a sound playing function, an image playing function, etc.), and the like; the storage data area may store data created according to use of the electronic device, and the like. The memory 12 may include non-volatile and volatile memories, such as: a hard disk, a memory, a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) Card, a Flash memory Card (Flash Card), at least one magnetic disk storage device, a Flash memory device, or other storage device.
The memory 12 may be an external memory and/or an internal memory of the electronic device 1. Further, the memory 12 may be a memory having a physical form, such as a memory stick, a TF Card (Trans-flash Card), or the like.
The integrated modules/units of the electronic device 1 may be stored in a computer-readable storage medium if they are implemented in the form of software functional units and sold or used as separate products. Based on such understanding, all or part of the flow of the method according to the above embodiments may be implemented by hardware that is configured to be instructed by computer readable instructions, which may be stored in a computer readable storage medium, and when the computer readable instructions are executed by a processor, the steps of the method embodiments may be implemented.
Wherein the computer readable instructions comprise computer readable instruction code which may be in source code form, object code form, an executable file or some intermediate form, and the like. The computer-readable medium may include: any entity or device capable of carrying said computer readable instruction code, recording medium, U-disk, removable hard disk, magnetic disk, optical disk, computer Memory, Read-Only Memory (ROM), Random Access Memory (RAM).
The block chain is a novel application mode of computer technologies such as distributed data storage, point-to-point transmission, a consensus mechanism, an encryption algorithm and the like. A block chain (Blockchain), which is essentially a decentralized database, is a series of data blocks associated by using a cryptographic method, and each data block contains information of a batch of network transactions, so as to verify the validity (anti-counterfeiting) of the information and generate a next block. The blockchain may include a blockchain underlying platform, a platform product service layer, an application service layer, and the like.
In conjunction with fig. 1, the memory 12 of the electronic device 1 stores computer readable instructions to implement an artificial intelligence based digital image recognition method, and the processor 13 executes the computer readable instructions to implement:
when a digital image identification request is received, acquiring an image to be identified according to the digital image identification request;
identifying an image table in the image to be identified;
performing text recognition on the image to be recognized based on the image table to obtain an image text;
extracting a first digital text from the image text according to the image table and a key text in the image text;
cutting the image to be recognized according to the first digital text to obtain a digital image;
inputting the digital image into a pre-trained digital recognition model to obtain a second digital text;
and if the first digital text is the same as the second digital text, determining the first digital text or the second digital text as a target digital text.
Specifically, the processor 13 may refer to the description of the relevant steps in the embodiment corresponding to fig. 1 for a specific implementation method of the computer readable instructions, which is not described herein again.
In the embodiments provided in the present invention, it should be understood that the disclosed system, apparatus and method may be implemented in other ways. For example, the above-described apparatus embodiments are merely illustrative, and for example, the division of the modules is only one logical functional division, and other divisions may be realized in practice.
The computer readable storage medium has computer readable instructions stored thereon, wherein the computer readable instructions when executed by the processor 13 are configured to implement the steps of:
when a digital image identification request is received, acquiring an image to be identified according to the digital image identification request;
identifying an image table in the image to be identified;
performing text recognition on the image to be recognized based on the image table to obtain an image text;
extracting a first digital text from the image text according to the image table and a key text in the image text;
cutting the image to be recognized according to the first digital text to obtain a digital image;
inputting the digital image into a pre-trained digital recognition model to obtain a second digital text;
and if the first digital text is the same as the second digital text, determining the first digital text or the second digital text as a target digital text.
The modules described as separate parts may or may not be physically separate, and parts displayed as modules may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the present embodiment.
In addition, functional modules in the embodiments of the present invention may be integrated into one processing unit, or each unit may exist alone physically, or two or more units are integrated into one unit. The integrated unit can be realized in a form of hardware, or in a form of hardware plus a software functional module.
The present embodiments are therefore to be considered in all respects as illustrative and not restrictive, the scope of the invention being indicated by the appended claims rather than by the foregoing description, and all changes which come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein. Any reference signs in the claims shall not be construed as limiting the claim concerned.
Furthermore, it is obvious that the word "comprising" does not exclude other elements or steps, and the singular does not exclude the plural. The plurality of units or devices may also be implemented by one unit or device through software or hardware. The terms first, second, etc. are used to denote names, but not any particular order.
Finally, it should be noted that the above embodiments are only for illustrating the technical solutions of the present invention and not for limiting, and although the present invention is described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that modifications or equivalent substitutions may be made on the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.
Claims (10)
1. A digital image recognition method based on artificial intelligence is characterized by comprising the following steps:
when a digital image identification request is received, acquiring an image to be identified according to the digital image identification request;
identifying an image table in the image to be identified;
performing text recognition on the image to be recognized based on the image table to obtain an image text;
extracting a first digital text from the image text according to the image table and a key text in the image text;
cutting the image to be recognized according to the first digital text to obtain a digital image;
inputting the digital image into a pre-trained digital recognition model to obtain a second digital text;
and if the first digital text is the same as the second digital text, determining the first digital text or the second digital text as a target digital text.
2. The artificial intelligence based digital image recognition method of claim 1, wherein prior to inputting the digital image into a pre-trained digital recognition model resulting in a second digital text, the method further comprises:
constructing a learner, wherein the learner comprises a convolution layer, a pooling layer and a full-connection layer;
acquiring a plurality of handwritten image samples, wherein each handwritten image sample comprises a sample image and a handwritten text in the sample image;
generating an image vector of the sample image from image pixels of the sample image;
performing convolution processing on the image vector based on the convolution layer to obtain convolution characteristics;
inputting the convolution features into the pooling layer to obtain text features;
acquiring a weight matrix in the full-connection layer, and calculating the product of the text features and the weight matrix to obtain a text vector;
mapping the text vector to obtain a predicted text of the learner on the sample image;
comparing each character in the predicted text with each character in the handwritten text, and counting the number of the characters in the predicted text and the number of the characters in the handwritten text to obtain the editing distance between the predicted text and the handwritten text;
counting the total number of characters of the predicted text to obtain a first number, and counting the total number of characters of the handwritten text to obtain a second number;
determining the numerical value with the largest value in the first quantity and the second quantity as a target quantity;
calculating the ratio of the editing distance to the target number to obtain the recognition error rate of the learner on the sample image;
and adjusting learning parameters in the learner according to the recognition error rate until the recognition error rate is not reduced any more, so as to obtain the digital recognition model.
3. The artificial intelligence based digital image recognition method of claim 1, wherein the obtaining the image to be recognized according to the digital image recognition request comprises:
analyzing the message of the digital image identification request to obtain data information carried by the message;
acquiring a storage path and a preset label from the data information;
and acquiring an image corresponding to the preset label from the storage path as the image to be identified.
4. The artificial intelligence based digital image recognition method of claim 1 wherein the recognizing an image form in the image to be recognized comprises:
carrying out binarization processing on the image to be identified to obtain a binary image;
acquiring an image line segment in the binary image based on a preset function;
and carrying out cross processing on the image line segments according to the line segment positions of the image line segments in the binary image to obtain the image table.
5. The artificial intelligence based digital image recognition method of claim 1, wherein the image table comprises a plurality of cells, the image text comprises text information and the cells in which the text information is located, and the extracting a first digital text from the image text according to the image table and the key text in the image text comprises:
performing semantic analysis on the text information to obtain semantic features of the text information;
determining a text field where the text information is located according to the semantic features;
acquiring a plurality of preset vocabularies from a domain vocabulary according to the text field, and performing word segmentation processing on the text information to obtain a plurality of text segmented words;
extracting a participle which is the same as any preset vocabulary from the plurality of text participles to serve as the key text;
obtaining a cell where the key text is located from the image text as a key cell;
acquiring a cell associated with the key cell from the image table as a target cell;
and extracting the first digital text from the image text according to the target cell.
6. The artificial intelligence based digital image recognition method of claim 5, wherein the cutting the image to be recognized according to the first digital text to obtain a digital image comprises:
obtaining a cell where the first digital text is located from the image text as the target cell;
acquiring coordinate information of the target cell from the image to be identified;
and cutting the image to be recognized according to the coordinate information to obtain a digital image.
7. The artificial intelligence based digital image recognition method of claim 1, wherein the method further comprises:
if the first digital text is different from the second digital text, generating feedback information according to the first digital text and the second digital text;
extracting, when a response result based on the feedback information is received, confirmation information from the response result;
counting the response quantity of the confirmation information which is not the second digital text;
and when the response quantity is larger than the preset quantity, adjusting the model parameters of the digital recognition model according to the confirmation information.
8. An artificial intelligence based digital image recognition apparatus, comprising:
the device comprises an acquisition unit, a recognition unit and a recognition unit, wherein the acquisition unit is used for acquiring an image to be recognized according to a digital image recognition request when the digital image recognition request is received;
the identification unit is used for identifying an image table in the image to be identified;
the identification unit is further used for performing text identification on the image to be identified based on the image table to obtain an image text;
the extraction unit is used for extracting a first digital text from the image text according to the image table and the key text in the image text;
the cutting unit is used for cutting the image to be recognized according to the first digital text to obtain a digital image;
the input unit is used for inputting the digital image into a pre-trained digital recognition model to obtain a second digital text;
the determining unit is further configured to determine the first digital text or the second digital text as a target digital text if the first digital text is the same as the second digital text.
9. An electronic device, characterized in that the electronic device comprises:
a memory storing computer readable instructions; and
a processor executing computer readable instructions stored in the memory to implement the artificial intelligence based digital image recognition method of any one of claims 1 to 7.
10. A computer-readable storage medium characterized by: the computer readable storage medium has stored therein computer readable instructions which are executed by a processor in an electronic device to implement the artificial intelligence based digital image recognition method according to any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111005704.4A CN113705468A (en) | 2021-08-30 | 2021-08-30 | Digital image identification method based on artificial intelligence and related equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111005704.4A CN113705468A (en) | 2021-08-30 | 2021-08-30 | Digital image identification method based on artificial intelligence and related equipment |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113705468A true CN113705468A (en) | 2021-11-26 |
Family
ID=78656911
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111005704.4A Pending CN113705468A (en) | 2021-08-30 | 2021-08-30 | Digital image identification method based on artificial intelligence and related equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113705468A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114639107A (en) * | 2022-04-21 | 2022-06-17 | 北京百度网讯科技有限公司 | Table image processing method, apparatus and storage medium |
| CN117558019A (en) * | 2024-01-11 | 2024-02-13 | 武汉理工大学 | Method for automatically extracting symbol map parameters from PDF format component manual |
-
2021
- 2021-08-30 CN CN202111005704.4A patent/CN113705468A/en active Pending
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114639107A (en) * | 2022-04-21 | 2022-06-17 | 北京百度网讯科技有限公司 | Table image processing method, apparatus and storage medium |
| CN114639107B (en) * | 2022-04-21 | 2023-03-24 | 北京百度网讯科技有限公司 | Table image processing method, apparatus and storage medium |
| CN117558019A (en) * | 2024-01-11 | 2024-02-13 | 武汉理工大学 | Method for automatically extracting symbol map parameters from PDF format component manual |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111694826B (en) | Data enhancement method and device based on artificial intelligence, electronic equipment and medium | |
| CN113032528B (en) | Case analysis method, case analysis device, case analysis equipment and storage medium | |
| CN112989826B (en) | Test question score determining method, device, equipment and medium based on artificial intelligence | |
| CN113094478B (en) | Expression reply method, device, equipment and storage medium | |
| CN111639487A (en) | Classification model-based field extraction method and device, electronic equipment and medium | |
| CN114090794A (en) | Event map construction method based on artificial intelligence and related equipment | |
| CN113656547B (en) | Text matching method, device, equipment and storage medium | |
| CN113705468A (en) | Digital image identification method based on artificial intelligence and related equipment | |
| CN113408278A (en) | Intention recognition method, device, equipment and storage medium | |
| CN115222443A (en) | Client group division method, device, equipment and storage medium | |
| CN113536770B (en) | Text analysis method, device and equipment based on artificial intelligence and storage medium | |
| CN114037545A (en) | Client recommendation method, device, equipment and storage medium | |
| CN112232203A (en) | Pedestrian recognition method and device, electronic equipment and storage medium | |
| CN112949305B (en) | Negative feedback information acquisition method, device, equipment and storage medium | |
| CN113627186B (en) | Entity relation detection method based on artificial intelligence and related equipment | |
| CN113627576B (en) | Code scanning information detection method, device, equipment and storage medium | |
| CN113486848B (en) | Document table identification method, device, equipment and storage medium | |
| CN113420545B (en) | Abstract generation method, device, equipment and storage medium | |
| CN113408265B (en) | Semantic analysis method, device and equipment based on human-computer interaction and storage medium | |
| CN113342977B (en) | Invoice image classification method, device, equipment and storage medium | |
| CN112989820B (en) | Legal document positioning method, device, equipment and storage medium | |
| CN114581177A (en) | Product recommendation method, device, equipment and storage medium | |
| CN113516205A (en) | Data classification method, device, equipment and storage medium based on artificial intelligence | |
| CN113434895B (en) | Text decryption method, device, equipment and storage medium | |
| CN113283421B (en) | Information identification method, device, equipment and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20221010 Address after: Room 2601 (Unit 07), Qianhai Free Trade Building, No. 3048, Xinghai Avenue, Nanshan Street, Qianhai Shenzhen-Hong Kong Cooperation Zone, Shenzhen, Guangdong 518000 Applicant after: Shenzhen Ping An Smart Healthcare Technology Co.,Ltd. Address before: 1-34 / F, Qianhai free trade building, 3048 Xinghai Avenue, Mawan, Qianhai Shenzhen Hong Kong cooperation zone, Shenzhen, Guangdong 518000 Applicant before: Ping An International Smart City Technology Co.,Ltd. |
|
| TA01 | Transfer of patent application right |