CN112115921A - True and false identification method and device and electronic equipment - Google Patents
True and false identification method and device and electronic equipment Download PDFInfo
- Publication number
- CN112115921A CN112115921A CN202011065166.3A CN202011065166A CN112115921A CN 112115921 A CN112115921 A CN 112115921A CN 202011065166 A CN202011065166 A CN 202011065166A CN 112115921 A CN112115921 A CN 112115921A
- Authority
- CN
- China
- Prior art keywords
- target
- main body
- feature
- real
- subject
- 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
- 238000000034 method Methods 0.000 title claims abstract description 69
- 238000013528 artificial neural network Methods 0.000 claims description 60
- 230000015654 memory Effects 0.000 claims description 19
- 238000000605 extraction Methods 0.000 claims description 17
- 238000012545 processing Methods 0.000 claims description 17
- 239000013598 vector Substances 0.000 claims description 16
- 238000012937 correction Methods 0.000 claims description 8
- 239000000126 substance Substances 0.000 claims description 2
- 238000013473 artificial intelligence Methods 0.000 abstract description 3
- 238000013135 deep learning Methods 0.000 abstract description 3
- 238000012795 verification Methods 0.000 description 15
- 238000001514 detection method Methods 0.000 description 12
- 230000008569 process Effects 0.000 description 11
- 239000010410 layer Substances 0.000 description 10
- 238000003384 imaging method Methods 0.000 description 9
- 238000013527 convolutional neural network Methods 0.000 description 6
- 238000010586 diagram Methods 0.000 description 6
- 239000000284 extract Substances 0.000 description 6
- 230000009467 reduction Effects 0.000 description 6
- 238000004891 communication Methods 0.000 description 4
- 230000006870 function Effects 0.000 description 4
- 230000002452 interceptive effect Effects 0.000 description 4
- 238000012549 training Methods 0.000 description 4
- 238000004590 computer program Methods 0.000 description 3
- 238000012986 modification Methods 0.000 description 3
- 230000004048 modification Effects 0.000 description 3
- 230000011218 segmentation Effects 0.000 description 3
- 230000000694 effects Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 230000003993 interaction Effects 0.000 description 2
- 239000004973 liquid crystal related substance Substances 0.000 description 2
- 238000012015 optical character recognition Methods 0.000 description 2
- 238000007781 pre-processing Methods 0.000 description 2
- 238000000513 principal component analysis Methods 0.000 description 2
- 230000003321 amplification Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000001413 cellular effect Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000013434 data augmentation Methods 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000010191 image analysis Methods 0.000 description 1
- 238000003709 image segmentation Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000010295 mobile communication Methods 0.000 description 1
- 238000003199 nucleic acid amplification method Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000011176 pooling Methods 0.000 description 1
- 230000000306 recurrent effect Effects 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 230000001953 sensory effect Effects 0.000 description 1
- 239000002356 single layer Substances 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
- G06V30/40—Document-oriented image-based pattern recognition
- G06V30/41—Analysis of document content
- G06V30/413—Classification of content, e.g. text, photographs or tables
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2415—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on parametric or probabilistic models, e.g. based on likelihood ratio or false acceptance rate versus a false rejection rate
-
- 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
-
- 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
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/22—Image preprocessing by selection of a specific region containing or referencing a pattern; Locating or processing of specific regions to guide the detection or recognition
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
- G06V30/10—Character recognition
- G06V30/14—Image acquisition
- G06V30/148—Segmentation of character regions
- G06V30/153—Segmentation of character regions using recognition of characters or words
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/40—Spoof detection, e.g. liveness detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
- G06V30/10—Character recognition
- G06V30/28—Character recognition specially adapted to the type of the alphabet, e.g. Latin alphabet
- G06V30/287—Character recognition specially adapted to the type of the alphabet, e.g. Latin alphabet of Kanji, Hiragana or Katakana characters
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Data Mining & Analysis (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Computing Systems (AREA)
- Molecular Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Probability & Statistics with Applications (AREA)
- Human Computer Interaction (AREA)
- Collating Specific Patterns (AREA)
Abstract
The application discloses a method and a device for identifying authenticity and electronic equipment, and relates to the technical field of artificial intelligence such as computer vision, deep learning and character recognition. The specific implementation scheme is as follows: acquiring a first image; under the condition that the first image is detected to comprise a target main body, identifying text content in the target main body and/or extracting characteristic information of the target main body, wherein the target main body comprises at least one of a stamp and a signature, and the characteristic information comprises at least one of a shape characteristic, a topological characteristic and a convolution characteristic; and determining the target authenticity identification result of the target main body based on the character content of the target main body and/or the characteristic information of the target main body. That is, in this embodiment, the authenticity of the target subject can be determined by using at least one of the text content of the target subject and/or the shape feature, the topology feature, and the convolution feature of the target subject, and the authenticity can be identified without manual checking, so that the efficiency of authenticity identification can be improved.
Description
Technical Field
The application relates to the technical field of artificial intelligence, in particular to the technical field of computer vision, deep learning and character recognition, and particularly relates to a method and a device for identifying authenticity and electronic equipment.
Background
In daily life or work, a large number of documents or contracts are generated, and in order to ensure authenticity of many important documents or contracts, authenticity verification of the above seals, signatures and the like is required.
At present, the authenticity of the main bodies such as the seal, the signature and the like in the document or the contract is usually verified by manual verification.
Disclosure of Invention
The application provides a method and a device for identifying authenticity and electronic equipment.
In a first aspect, an embodiment of the present application provides a method for authenticating authenticity, the method including:
acquiring a first image;
under the condition that the first image is detected to comprise a target main body, identifying text content in the target main body and/or extracting feature information of the target main body, wherein the target main body comprises at least one of a seal and a signature, and the feature information comprises at least one of a shape feature, a topological feature and a convolution feature;
and determining the target authenticity identification result of the target main body based on the text content of the target main body and/or the characteristic information of the target main body.
In the authenticity identification method of the embodiment, the first image is acquired, and when the target main body is detected in the first image, the text content in the target main body is identified and/or the characteristic information of the target main body is extracted, and the target authenticity identification result of the target main body is determined based on the text content of the target main body and/or the characteristic information of the target main body. The target main body comprises at least one of a seal and a signature, the characteristic information can comprise at least one of a shape characteristic, a topology characteristic and a convolution characteristic, namely in the embodiment, the authenticity of the target main body can be determined by utilizing the text content of the target main body and/or at least one of the shape characteristic, the topology characteristic and the convolution characteristic of the target main body, the authenticity does not need to be identified through manual checking, and the authenticity identification efficiency can be improved. Meanwhile, the authenticity identification method of the embodiment can be used for checking the situation that errors are easy to generate when authenticity is identified through manual checking, and can improve identification accuracy.
In a second aspect, an embodiment of the present application provides an apparatus for authenticating a genuine person, the apparatus including:
the acquisition module is used for acquiring a first image;
the processing module is used for identifying text content in a target main body and/or extracting feature information of the target main body under the condition that the first image is detected to comprise the target main body, wherein the target main body comprises at least one of a seal and a signature, and the feature information comprises at least one of a shape feature, a topological feature and a convolution feature;
and the determining module is used for determining the target authenticity identification result of the target main body based on the text content of the target main body and/or the characteristic information of the target main body.
In the authenticity identification process of the authenticity identification device of the embodiment, the first image is obtained, and under the condition that the target main body is detected in the first image, the character content in the target main body is identified and/or the characteristic information of the target main body is extracted, and the target authenticity identification result of the target main body is determined based on the character content of the target main body and/or the characteristic information of the target main body. The target main body comprises at least one of a seal and a signature, the characteristic information can comprise at least one of a shape characteristic, a topology characteristic and a convolution characteristic, namely in the embodiment, the authenticity of the target main body can be determined by utilizing the text content of the target main body and/or at least one of the shape characteristic, the topology characteristic and the convolution characteristic of the target main body, the authenticity does not need to be identified through manual checking, and the authenticity identification efficiency can be improved. Meanwhile, the authenticity identification method of the embodiment can be used for checking the situation that errors are easy to generate when authenticity is identified through manual checking, and can improve identification accuracy.
In a third aspect, an embodiment of the present application further provides an electronic device, including:
at least one processor; and
a memory communicatively coupled to the at least one processor; wherein the content of the first and second substances,
the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the methods provided by the embodiments of the present application.
In a fourth aspect, an embodiment of the present application further provides a non-transitory computer-readable storage medium storing computer instructions for causing a computer to perform the method provided by the embodiments of the present application.
Drawings
The drawings are included to provide a better understanding of the present solution and are not intended to limit the present application. Wherein:
fig. 1 is a schematic flow chart of an authentication method according to an embodiment provided in the present application;
fig. 2 is a second schematic flowchart of an authentication method according to an embodiment of the present application;
fig. 3 is a third schematic flowchart of an authentication method according to an embodiment of the present application;
fig. 4 is a schematic diagram of an authentication system implementing an authentication method according to an embodiment provided in the present application;
fig. 5 is a fourth schematic flowchart of an authentication method according to an embodiment of the present application;
fig. 6 is one of the structural diagrams of the authentication apparatus according to an embodiment provided in the present application;
fig. 7 is a second structural diagram of an authentication device according to an embodiment of the present application;
fig. 8 is a third structural view of an authentication device according to an embodiment of the present application;
fig. 9 is a block diagram of an electronic device for implementing the authentication method according to the embodiment of the present application.
Detailed Description
The following description of the exemplary embodiments of the present application, taken in conjunction with the accompanying drawings, includes various details of the embodiments of the application for the understanding of the same, which are to be considered exemplary only. Accordingly, those of ordinary skill in the art will recognize that various changes and modifications of the embodiments described herein can be made without departing from the scope and spirit of the present application. Also, descriptions of well-known functions and constructions are omitted in the following description for clarity and conciseness.
As shown in fig. 1, according to an embodiment of the present application, there is provided an authenticity identifying method applicable to an electronic device, the method including:
step S101: a first image is acquired.
The first image may be an image obtained by shooting a subject to be identified (i.e., an object to be identified), an image obtained by scanning, or an image obtained by screenshot. In the authentication process, the first image may be input to an execution subject, such as an electronic device, that performs the authentication method. For example, for a contract with a seal, an image corresponding to the contract can be obtained by imaging the contract with an imaging device. For example, a document having a signature is captured by an imaging device, and an image corresponding to the document is obtained.
Step S102: and under the condition that the first image is detected to comprise the target main body, identifying the text content in the target main body and/or extracting the characteristic information of the target main body.
Wherein the target subject includes at least one of a stamp and a signature, and the feature information includes at least one of a shape feature, a topology feature, and a convolution feature.
The first images obtained by shooting different subjects to be authenticated are different in target subject, for example, the first image obtained by shooting a contract with a seal includes a target subject as a seal, and the first image obtained by shooting a file with a signature includes a target subject as a signature. In this embodiment, in a case that it is detected that the first image includes at least one of a stamp and a signature, the text content in the target subject may be identified and/or the feature information of the target subject may be extracted, as an example, for a stamp, the text content in the stamp may be identified and the shape feature of the stamp may be extracted, and for a signature, the topological feature and/or the convolution feature of the signature may be extracted.
As an example, the first image may be subject to target detection through a pre-trained target detection model, and in a case that the first image is detected to include a target subject, text content in the target subject is identified and/or feature information of the target subject is extracted. As an example, the object detection model may be yolov3 detection model (an object detection network) through which object detection may improve object detection efficiency.
Step S103: and determining the target authenticity identification result of the target main body based on the character content of the target main body and/or the characteristic information of the target main body.
After the text content in the target main body is identified and/or the characteristic information of the target main body is extracted, the text content of the target main body and/or the characteristic information of the target main body can be used for determining the target authenticity identification result of the target main body, the authenticity of the seal and/or the signature in the first image can be determined, and the seal and/or the signature authenticity identification is realized. The target authenticity identification result comprises a true identification result or a false identification result.
In the authenticity identification method of the embodiment, the first image is acquired, and when the target main body is detected in the first image, the text content in the target main body is identified and/or the characteristic information of the target main body is extracted, and the target authenticity identification result of the target main body is determined based on the text content of the target main body and/or the characteristic information of the target main body. The target main body comprises at least one of a seal and a signature, the characteristic information can comprise at least one of a shape characteristic, a topology characteristic and a convolution characteristic, namely in the embodiment, the authenticity of the target main body can be determined by utilizing the text content of the target main body and/or at least one of the shape characteristic, the topology characteristic and the convolution characteristic of the target main body, the authenticity does not need to be identified through manual checking, and the authenticity identification efficiency can be improved. Meanwhile, the authenticity identification method of the embodiment can be used for checking the situation that errors are easy to generate when authenticity is identified through manual checking, and can improve identification accuracy.
Optionally, the target main body includes a stamp, and the characteristic information of the target main body includes a shape characteristic of the stamp. In this embodiment, determining the target authenticity identification result of the target subject based on the text content of the target subject and/or the feature information of the target subject includes: comparing the text content of the seal with the real text content of the seal to obtain a first authenticity identification result of the seal; determining a second authenticity identification result of the seal based on the shape characteristics of the seal; and determining the target authenticity result according to the first authenticity identification result and the second authenticity identification result. That is, in this embodiment, as shown in fig. 2, there is provided a method for authenticating authenticity, including the steps of:
step S201: a first image is acquired.
The first image may be an image obtained by shooting a subject to be identified (i.e., an object to be identified), an image obtained by scanning, or an image obtained by screenshot. In the authentication process, the first image may be input to an execution subject, such as an electronic device, that performs the authentication method. For example, for a contract with a seal, an image corresponding to the contract can be obtained by imaging the contract with an imaging device. For example, a document having a signature is captured by an imaging device, and an image corresponding to the document is obtained.
Step S202: and under the condition that the first image is detected to comprise the target main body, identifying the text content in the target main body and/or extracting the characteristic information of the target main body.
The target body comprises a seal, and the characteristic information comprises shape characteristics.
The first images obtained by shooting different to-be-authenticated main bodies comprise different target main bodies, for example, in the first image obtained by shooting a contract with a seal, the target main body comprises the seal, and in the first image obtained by shooting a file with a signature, the target main body comprises the signature. In this embodiment, in the case that it is detected that the first image includes at least one of a stamp and a signature, the text content in the target body may be identified and/or the feature information of the target body may be extracted, and as an example, for a stamp, the text content in the stamp may be identified and the shape feature of the stamp may be extracted. In this embodiment, the target main body includes a stamp, and the feature information of the target main body includes a shape feature of the stamp, that is, when it is detected that the first image includes the stamp, the text content in the stamp is recognized and the shape feature of the stamp is extracted.
Step S203: and comparing the text content of the seal with the real text content of the seal to obtain a first authenticity identification result of the seal.
Step S204: and determining a second authenticity identification result of the seal based on the shape characteristics of the seal.
Step S205: and determining the target authenticity result according to the first authenticity identification result and the second authenticity identification result.
Under the condition that the first image comprises the seal, recognizing the text content in the seal, extracting the shape characteristic of the seal, and then comparing the text content of the seal with the real text content of the seal to obtain a first authenticity identification result of the seal. And determining a second authenticity identification result of the seal based on the shape characteristics of the seal. That is, the stamp authentication involves two aspects, namely, on one hand, the comparison of the text content, and on the other hand, the authenticity is judged by the shape feature of the stamp image (region) in the first image. Therefore, the accuracy of identifying the authenticity of the seal can be improved.
As an example, extracting the shape feature of the stamp may include performing feature extraction on the stamp through a VGG16 model (a convolutional neural network) to obtain the shape feature of the stamp.
As an example, determining the target authenticity result from the first authenticity verification result and the second authenticity verification result includes: determining the target authentication result as a true authentication result under the condition that the first authenticity authentication result and the second authenticity authentication result are both true authentication results; or under the condition that at least one of the first authenticity identification result and the second authenticity identification result is a false and true identification result, determining that the target identification result is a false identification result.
As an example, recognizing the text content in the stamp may include: detecting characters in the seal to obtain a character main body of the seal, and then performing character recognition on the character main body to obtain character contents of the seal. As an example, characters in the stamp can be detected by an LOMO model (a text detector), and characters on the stamp, including bent characters on a red stamp, can be detected more accurately by the LOMO model. The LOMO model includes a Direct Regressor (DR), an Iterative Refinement Module (IRM), and a Shape Expression Module (SEM) connected in sequence. First, the DR branch generates a quadrangular text propusals (preselected box). The IRM then progressively perceives the entire long text through iterative refinement based on the extracted feature blocks of the text explosals. And then, by considering the geometric properties of the text example, including text regions, text center line and boundary offset and the like, introducing the SEM to obtain a more accurate curved text representation. By utilizing the LOMO model, the bent characters on the red seal can be detected very accurately. As an example, the text body may be subjected to text recognition through a text recognition model, that is, a detected text is recognized, and the text recognition model may adopt a CRNN (Convolutional Neural Network, RNN, Convolutional Recurrent Neural Network) model or the like.
Optionally, the feature information includes at least one of a topological feature and a convolution feature. In this embodiment, determining the target authenticity identification result of the target subject based on the text content of the target subject and/or the feature information of the target subject includes: inputting the topological characteristic of the target subject and the real topological characteristic of the real subject corresponding to the target subject into a first neural network, and outputting a first real probability of the target subject through the first neural network; and/or inputting the convolution characteristic of the target subject and the real convolution characteristic of the real subject corresponding to the target subject into a second neural network, and outputting a second real probability of the target subject through the second neural network; and determining the authenticity identification result of the target based on the first real probability and/or the second real probability. That is, in the present embodiment, as shown in fig. 3, there is provided a method for authenticating authenticity, including the steps of:
step S301: a first image is acquired.
The first image may be an image obtained by shooting a subject to be identified (i.e., an object to be identified), an image obtained by scanning, or an image obtained by screenshot. In the authentication process, the first image may be input to an execution subject, such as an electronic device, that performs the authentication method. For example, for a contract with a seal, an image corresponding to the contract can be obtained by imaging the contract with an imaging device. For example, a document having a signature is captured by an imaging device, and an image corresponding to the document is obtained.
Step S302: and under the condition that the first image is detected to comprise the target main body, identifying the text content in the target main body and/or extracting the characteristic information of the target main body.
Wherein the feature information includes at least one of a topological feature and a convolution feature.
The first images obtained by shooting different to-be-authenticated main bodies comprise different target main bodies, for example, in the first image obtained by shooting a contract with a seal, the target main body comprises the seal, and in the first image obtained by shooting a file with a signature, the target main body comprises the signature. In this embodiment, when it is detected that the first image includes at least one of a stamp and a signature, the text content in the target subject may be identified and/or the feature information of the target subject may be extracted. As one example, for a signature, topological and/or convolutional features of the signature may be extracted. For example, in the present embodiment, the target subject includes a signature, and the feature information of the target subject includes a topological feature and/or a convolution feature of the signature, that is, in a case where it is detected that the first image includes the signature, and the topological feature and/or the convolution feature of the signature is extracted.
Step S303: inputting the topological characteristic of the target subject and the real topological characteristic of the real subject corresponding to the target subject into a first neural network, and outputting a first real probability of the target subject through the first neural network; and/or inputting the convolution characteristic of the target subject and the real convolution characteristic of the real subject corresponding to the target subject into a second neural network, and outputting a second real probability of the target subject through the second neural network.
Step S304: and determining the authenticity identification result of the target based on the first real probability and/or the second real probability.
In this embodiment, at least one of the topological feature and the convolution feature of the target subject is used to identify the authenticity of the target subject, for example, the topological feature of the target subject and the real topological feature of the real subject corresponding to the target subject may be input to the first neural network, and the first real probability of the target subject may be output through the first neural network, and/or the convolution feature of the target subject and the real convolution feature of the real subject corresponding to the target subject may be input to the second neural network, and the second real probability of the target subject may be output through the second neural network. Therefore, the first true probability and/or the second true probability can be obtained, and then the target authenticity identification result can be determined based on the first true probability and/or the second true probability so as to improve the accuracy of the authenticity identification of the target main body.
As one example, the first neural network may be a first BP (back propagation) neural network and the second neural network may be a second convolutional neural network. As one example, the target subject may be a signature.
Optionally, the inputting the topological feature of the target subject and the real topological feature of the real subject corresponding to the target subject into the first neural network, and outputting the first real probability of the target subject through the first neural network, includes: obtaining a first feature vector based on the topological feature and the real topological feature of the target subject; and inputting the first feature vector into a first neural network to obtain a first true probability.
Taking the target subject as the signature as an example, the corresponding real subject is the real signature, and the signature notes of each person have own characteristics and relative stability. In the authentication process, the signature can be regarded as a plane undirected graph, and then topological features of the undirected graph are extracted through an algorithm. The topological features may include: a connecting sheet (handwriting connected with each other), a mesh (a closed blank area surrounded by the handwriting), a one-degree fixed point (an end point of the handwriting) and a multi-degree vertex (a triple point, a quadruple point and the like formed by intersecting handwriting).
To provide topological discrimination, signatures can be skeletonized using OpenCV (which is a BSD (Berkeley Software Distribution) -based licensed (open source) distributed cross-platform computer vision and machine learning Software library that can run on Linux, Windows, Android, and Mac OS operating systems). The skeleton is the shape characteristic of the character symbol in the signature and is composed of a plurality of thin curves or circular arcs. The curves or arcs can better keep the connectivity of the original shape of the character symbol, show the topological property of the character symbol, are important representations of the structural shape, and can simplify the signature character into a skeleton image. And (3) skeleton extraction, namely extracting the central pixel outline of characters in the signature on the image. The extraction of the skeleton of the characters can simplify the features of the image and is also beneficial to feature extraction in subsequent deep learning. The method is realized by continuously corroding from the periphery of the character target to the center of the character target until the character target can not be corroded any more, and leaving a single-layer pixel width, namely the character skeleton.
After the topological structure (namely the topological characteristic) of the signature is obtained, vectorizing the topological characteristic of the signature and the topological characteristic of the real signature (the extraction process is similar to that of the topological characteristic of the signature, but the signature is different, namely the real signature) to obtain a first characteristic vector, inputting the first characteristic vector into a first neural network, performing second classification, judging whether the signature and the compared real signature are from the same person, and giving a corresponding confidence coefficient (namely a first real probability).
The real topological characteristic of the real main body can be obtained in advance, after the topological characteristic of the target main body is obtained, the topological characteristic of the target main body and the real topological characteristic can be vectorized to obtain a first characteristic vector, the first characteristic vector is input into the first neural network, and the first real probability is output through the first neural network. Thus, the accuracy of the first true probability can be improved.
Optionally, inputting the convolution feature of the target subject and the true convolution feature of the true subject corresponding to the target subject into the second neural network, and outputting a second true probability of the target subject through the second neural network, including: carrying out binarization processing on the target main body to obtain a first binarization main body; performing character cutting on the first binarized main body to obtain cut characters of the first binarized main body; extracting convolution characteristics of the cut character of the first binarized main body through a first convolution neural network; and inputting the convolution characteristic of the cut character of the first binarized main body and the real convolution characteristic of the cut character of the real main body into a second neural network to obtain a second real probability.
The real convolution characteristics of the cut characters of the real main body can be extracted in advance, the extraction process of the real convolution characteristics of the cut characters of the real main body is similar to the process of the convolution characteristics of the cut characters of the target main body, namely, the real main body is subjected to binarization processing in advance to obtain a second binarization main body; performing character cutting on the second binarization main body to obtain cut characters of the second binarization main body; and extracting the convolution characteristics of the cut characters of the second binarization main body through the first convolution neural network to realize the extraction of the convolution characteristics of the real main body. Obtaining a first binarization main body by carrying out binarization processing on a target main body; performing character cutting on the first binarized main body to obtain cut characters of the first binarized main body; extracting the convolution characteristics of the cut characters of the first binary main body through the first convolution neural network to realize the extraction of the convolution characteristics of the target main body. And then inputting the convolution characteristic of the cut character of the first binarized main body and the real convolution characteristic of the cut character of the real main body into a second neural network, and outputting a second real probability through the second neural network, so that the accuracy of the second real probability can be improved.
As an example, for character cutting, the character can be cut by the connected component algorithm in OpenCV. The cutting by the connected component algorithm is selected based on the situation that continuous strokes exist in the signature, and for the situation, continuous strokes of characters are not cut by the connected component algorithm, but are input into the first convolutional neural network as important features.
As an example, before inputting the convolution feature of the cut character of the first binarized subject and the real convolution feature of the cut character of the real subject into the second neural network, the method may further include: and carrying out dimension reduction processing on the convolution characteristics of the cut characters of the first binarized main body to obtain first dimension reduction characteristics, and encoding the first dimension reduction characteristics in a Fisher Vectors encoding mode to obtain second characteristics, so that the characteristics with the same dimension and global characteristics can be formed. Therefore, the second characteristic of the cut character of the first binarized main body and the third characteristic of the cut character of the real main body can be input into the second neural network, and the second real probability is output through the second neural network, wherein the third characteristic is obtained by performing dimension reduction processing on the convolution characteristic of the cut character of the real main body and encoding the dimension reduced characteristic through a Fisher Vectors encoding mode. As an example, the dimensionality reduction processing may be performed by a PCA (Principal Component Analysis) dimensionality reduction method.
As an example, the first convolutional neural network may be an AlexNet network (a type of convolutional neural network) with which convolutional feature extraction is performed, for example, the AlexNet network may include 5 convolutional layers, 3 pooling layers, and 2 fully-connected layers connected in sequence for feature extraction. The first convolutional layer has 96 convolution kernels, the size of the convolution kernels is 11 x 11, and the convolution step size is 4. The second convolutional layer has 256 convolutional kernels, the convolutional kernel size is 5, and the convolution step size is 1. The convolution kernel sizes of the next three convolutional layers are all 3, and the convolution step sizes are all 1. The number of convolution kernels for the third and fourth convolutional layers is 384, and the number of convolution kernels for the fifth convolutional layer is 256. And finally, the number of the two full connection layer nodes is 4096, and the number of the output layer nodes is 1000. The AlexNet network is selected to extract the features, and the effect of the network in image classification is superior to that of the similar network, so that the accuracy of the extracted convolution features can be improved.
Optionally, in a case that the first image is detected to include the target subject, before recognizing text content in the target subject and/or extracting feature information of the target subject, the method further includes: extracting a subimage to be identified of the first image; carrying out angle correction on the subimage to be identified to obtain a corrected subimage;
under the condition that the first image is detected to comprise the target main body, recognizing the text content in the target main body and/or extracting the characteristic information of the target main body comprises the following steps:
and under the condition that the corrected image is detected to comprise the target main body, recognizing the text content in the target main body and/or extracting the characteristic information of the target main body.
The subimage to be identified can be understood as the image of the main body to be identified, and because other backgrounds or interferences may exist in the process of shooting the main body to be identified, after the first image is acquired, the main body to be identified needs to be detected on the first image, the main body to be identified is separated from the background, and the extraction of the main body to be identified is realized, namely the subimage to be identified is extracted. The detection of the main body to be identified mainly comprises the steps of extracting the main body part to be identified in the first image through an algorithm and removing the interference of a background. The model applied to the extraction of the subject to be identified of the first image may be an image semantic segmentation model, that is, a sub-image to be identified of the first image is extracted through the image semantic segmentation model. The algorithm applied to the image semantic segmentation can be PaddleSeg (a kind of image segmentation library), and the algorithm can more accurately detect the subject to be identified.
After the sub-image to be identified of the first image is extracted, angle correction is needed to be carried out on the sub-image to be identified, due to the problem of the shooting angle, the first image obtained by shooting may have a certain inclination, and the extracted sub-image to be identified may have a certain inclination, so that the angle correction can be carried out on the sub-image to be identified, the corrected sub-image can be obtained, and the accuracy of subsequent character content identification and feature extraction is improved. As one example, rectification of the image can be accomplished using a rectification algorithm in OpenCV.
The process of the above-mentioned authentication method is explained in detail with an embodiment.
As shown in fig. 4, in order to realize the principle diagram of the authentication system of the authentication method, the authentication system includes: the system comprises an image access module, an image preprocessing module, a seal detection and verification module, an invoice number verification module, a signature verification module, an interactive interface and a storage module. The image access module receives the first image, the image preprocessing module extracts and corrects the sub-image to be identified, the seal detection and verification module is used for detecting the seal and identifying the authenticity of the seal, the invoice number verification module is used for identifying the invoice number and identifying the authenticity of the invoice number, and the signature verification module is used for detecting the signature and identifying the authenticity of the signature.
As shown in fig. 5, which is a flowchart of the method for authenticating authenticity according to an embodiment, taking an example of a sequence of first detecting a stamp, then detecting an invoice number, and finally detecting a signature, a flow of the method for authenticating authenticity through the system for authenticating authenticity is as follows:
firstly, a user obtains a first image through modes of scanning, shooting or screenshot and the like, and the first image is input into the authenticity identification system.
The authenticity identification system firstly detects whether the first image contains a seal through a seal detection module.
If the first image contains the seal, performing seal authenticity identification on the first image, for example, identifying the text content in the seal through a seal content identification module, and comparing the text content with the real text content through a seal comparison module; if the first image does not contain the seal, the step is skipped, and whether the first image contains the invoice number or not is detected.
The authenticity identification system detects whether the first image contains the invoice number through an Optical Character Recognition (OCR) module. If the first image contains the invoice number, the invoice is verified through calling an invoice query interface. If the first image does not contain the invoice number, the system skips this step and proceeds to detect if the first image contains a handwritten signature.
The authentication system then detects whether the first image includes a handwritten signature. If the first image contains the handwritten signature, the system extracts the topological features through a topological structure-based identification module in a signature verification module, performs authenticity identification by using the topological features, extracts the convolution features through an image analysis identification module, and performs authenticity identification based on the convolution features, so that authenticity of the signature is determined.
The identification results are displayed on the interactive interface, the identification results can be checked manually through the interactive interface, and handwritten signatures and seal images can be input through the interactive interface.
And finally, a storage module of the authenticity identification system stores all the identification results for system training. The storage module is mainly used for storing the first image, the seal, the signature and the like received by the system. And the subsequent data becomes training data of the system after being labeled. In order to more accurately verify the signature, the signature remaining in the system needs to be trained. The training process mainly aims at a classifier for performing identification based on topological features, an AlexNet model and a classifier for performing classification based on image information. For data augmentation, the embodiment of the application generates more signatures by using the saved signatures. For example, the signature of size 70 x 60 is first scaled up to 76 x 66, i.e., 6 pixels each increase in width and height. Then 70 x 60 sub-regions are acquired in the magnified image in the form of sliding windows with a step size of 3, so that a total of 9 homogeneous samples are obtained. In addition, the present application can perform signature amplification by rotating a signature.
The application completes the verification work through the artificial intelligence technology (including the technologies of handwritten content identification, printed character identification, invoice number verification, handwritten signature verification, seal detection and the like). The whole verification process is carried out automatically, the efficiency is high, and the cost is low. The method can improve the efficiency of authenticity identification, has high response speed, can form stable service, and has strong generalization capability. The training data generation module is included in the application, and the cost of manually collecting data and marking data can be reduced.
As shown in fig. 6, according to an embodiment of the present application, the present application further provides an apparatus 600 for authenticating authenticity, which is applicable to an electronic device, the apparatus 600 including:
an obtaining module 601, configured to obtain a first image;
the processing module 602 is configured to, in a case that it is detected that the first image includes a target subject, identify text content in the target subject and/or extract feature information of the target subject, where the target subject includes at least one of a stamp and a signature, and the feature information includes at least one of a shape feature, a topology feature and a convolution feature;
the determining module 603 is configured to determine an object authenticity identification result of the object main body based on the text content of the object main body and/or the feature information of the object main body.
Optionally, the target main body includes a stamp, and the characteristic information of the target main body includes shape characteristics of the stamp;
a determination module comprising:
the comparison module is used for comparing the text content of the seal with the real text content of the seal to obtain a first authenticity identification result of the seal;
the first sub-determination module is used for determining a second authenticity identification result of the seal based on the shape characteristics of the seal;
and the second sub-determining module is used for determining the target authenticity result according to the first authenticity identification result and the second authenticity identification result.
That is, in this embodiment, as shown in fig. 7, the present application further provides an authentication apparatus 700 applicable to an electronic device, the apparatus 700 including:
an obtaining module 701, configured to obtain a first image;
a processing module 702, configured to, when it is detected that the first image includes a target main body, identify text content in the target main body and/or extract feature information of the target main body, where the target main body includes the stamp, and the feature information of the target main body includes a shape feature of the stamp;
a comparison module 703, configured to compare the text content of the stamp with the real text content of the stamp to obtain a first authenticity identification result of the stamp;
a first sub-determining module 704, configured to determine a second authenticity identification result of the stamp based on the shape characteristic of the stamp;
a second sub-determining module 705, configured to determine the target authenticity result according to the first authenticity identification result and the second authenticity identification result.
Optionally, the feature information includes at least one of a topological feature and a convolution feature;
a determination module comprising:
the probability determination module is used for inputting the topological characteristic of the target subject and the real topological characteristic of the real subject corresponding to the target subject into the first neural network and outputting the first real probability of the target subject through the first neural network; and/or, the convolution characteristic of the target subject and the real convolution characteristic of the real subject corresponding to the target subject are input into a second neural network, and a second real probability of the target subject is output through the second neural network;
and the third sub-determination module is used for determining the authenticity identification result of the target based on the first real probability and/or the second real probability.
That is, in this embodiment, as shown in fig. 8, the present application further provides an authentication apparatus 800 applicable to an electronic device, where the apparatus 800 includes:
an obtaining module 801, configured to obtain a first image;
a processing module 802, configured to, when it is detected that the first image includes a target main body, recognize text content in the target main body and/or extract feature information of the target main body, where the target main body includes the stamp, and the feature information of the target main body includes a shape feature of the stamp;
a probability determining module 803, configured to input the topological feature of the target subject and a real topological feature of a real subject corresponding to the target subject into a first neural network, and output a first real probability of the target subject through the first neural network; and/or, the convolution characteristic of the target subject and the real convolution characteristic of the real subject corresponding to the target subject are input into a second neural network, and a second real probability of the target subject is output through the second neural network;
a third sub-determining module 804, configured to determine the target authenticity identifying result based on the first reality probability and/or the second reality probability.
Optionally, the probability determining module 803 includes:
the characteristic vector determining module is used for obtaining a first characteristic vector based on the topological characteristic and the real topological characteristic of the target main body;
the first probability obtaining module is used for inputting the first feature vector to the first neural network to obtain a first true probability.
Optionally, the probability determining module 803 includes:
the binarization module is used for carrying out binarization processing on the target main body to obtain a first binarization main body;
the cutting module is used for carrying out character cutting on the first binarized main body to obtain cut characters of the first binarized main body;
the characteristic extraction module is used for extracting the convolution characteristics of the cut characters of the first binary main body through a first convolution neural network;
and the second real probability determining module is used for inputting the convolution characteristics of the cut characters of the first binarized main body and the real convolution characteristics of the cut characters of the real main body into a second neural network to obtain a second real probability.
Optionally, the apparatus for authenticating a genuine may further include:
the subimage extraction module is used for extracting subimages to be identified of the first image;
the correction module is used for carrying out angle correction on the subimage to be identified to obtain a corrected subimage;
under the condition that the first image is detected to comprise the target main body, recognizing the text content in the target main body and/or extracting the characteristic information of the target main body comprises the following steps:
and under the condition that the corrected image is detected to comprise the target main body, recognizing the text content in the target main body and/or extracting the characteristic information of the target main body.
The authenticity identification device of each embodiment is a device for implementing the authenticity identification method of each embodiment, and the technical features and technical effects correspond to each other, and are not described herein again.
According to an embodiment of the present application, an electronic device and a readable storage medium are also provided.
Fig. 9 is a block diagram of an electronic device according to the authentication method of the embodiment of the present application. Electronic devices are intended to represent various forms of digital computers, such as laptops, desktops, workstations, personal digital assistants, servers, blade servers, mainframes, and other appropriate computers. The electronic device may also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions, are meant to be examples only, and are not meant to limit implementations of the present application that are described and/or claimed herein.
As shown in fig. 9, the electronic apparatus includes: one or more processors 901, memory 902, and interfaces for connecting the various components, including a high-speed interface and a low-speed interface. The various components are interconnected using different buses and may be mounted on a common motherboard or in other manners as desired. The processor may process instructions for execution within the electronic device, including instructions stored in or on the memory to display graphical information of the GUM on an external input/output device (such as a display device coupled to the interface). In other embodiments, multiple processors and/or multiple buses may be used, along with multiple memories and multiple memories, as desired. Also, multiple electronic devices may be connected, with each device providing portions of the necessary operations (e.g., as a server array, a group of blade servers, or a multi-processor system). Fig. 9 illustrates an example of a processor 901.
The memory 902, which is a non-transitory computer readable storage medium, may be used to store non-transitory software programs, non-transitory computer executable programs, and modules, such as program instructions/modules corresponding to the authentication method in the embodiment of the present application (for example, the obtaining module 601, the processing module 602, and the determining module 603 shown in fig. 6). The processor 901 executes various functional applications of the server and data processing by running non-transitory software programs, instructions and modules stored in the memory 902, so as to implement the authentication method in the above method embodiments.
The memory 902 may include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application program required for at least one function; the storage data area may store data created according to use of the electronic device displayed by the keyboard, and the like. Further, the memory 902 may include high speed random access memory, and may also include non-transitory memory, such as at least one magnetic disk storage device, flash memory device, or other non-transitory solid state storage device. In some embodiments, the memory 902 may optionally include memory located remotely from the processor 901, which may be connected to a keyboard-displayed electronic device via a network. Examples of such networks include, but are not limited to, the internet, intranets, local area networks, mobile communication networks, and combinations thereof.
The electronic device of the authentication method may further include: an input device 903 and an output device 904. The processor 901, the memory 902, the input device 903 and the output device 904 may be connected by a bus or other means, and fig. 9 illustrates the connection by a bus as an example.
The input device 903 may receive input numeric or character information and generate key signal inputs related to user settings and function controls of the electronic device displayed by the keyboard, such as a touch screen, keypad, mouse, track pad, touch pad, pointer stick, one or more mouse buttons, track ball, joystick, or other input device. The output devices 904 may include a display device, auxiliary lighting devices (e.g., LEDs), tactile feedback devices (e.g., vibrating motors), and the like. The display device may include, but is not limited to, a Liquid Crystal Display (LCD), a Light Emitting Diode (LED) display, and a plasma display. In some implementations, the display device can be a touch screen.
Various implementations of the systems and techniques described here can be realized in digital electronic circuitry, integrated circuitry, special-purpose ASMC (application specific integrated circuit), computer hardware, firmware, software, and/or combinations thereof. These various embodiments may include: implemented in one or more computer programs that are executable and/or interpretable on a programmable system including at least one programmable processor, which may be special or general purpose, receiving data and instructions from, and transmitting data and instructions to, a storage system, at least one input device, and at least one output device.
These computing programs (also known as programs, software applications, or code) include machine instructions for a programmable processor, and may be implemented using procedural and/or object oriented programming languages, and/or assembly/machine languages. As used herein, the terms "machine-readable medium" and "computer-readable medium" refer to any computer program product, apparatus, and/or device (e.g., magnetic discs, optical disks, memory, Programmable Logic Devices (PLDs)) used to provide machine instructions and/or data to a programmable processor, including a machine-readable medium that receives machine instructions as a machine-readable signal. The term "machine-readable signal" refers to any signal used to provide machine instructions and/or data to a programmable processor.
To provide for interaction with a user, the systems and techniques described here can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to a user; and a keyboard and a pointing device (e.g., a mouse or a trackball) by which a user can provide input to the computer. Other kinds of devices may also be used to provide for interaction with a user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form, including acoustic, speech, or tactile input.
The systems and techniques described here can be implemented in a computing system that includes a back-end component (e.g., as a data server), or that includes a middleware component (e.g., an application server), or that includes a front-end component (e.g., a user computer having a graphical user interface or a web browser through which a user can interact with an implementation of the systems and techniques described here), or any combination of such back-end, middleware, or front-end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: local Area Networks (LANs), Wide Area Networks (WANs), and the Internet.
The computer system may include clients and servers. A client and server are generally remote from each other and typically interact through a communication network. The relationship of client and server arises by virtue of computer programs running on the respective computers and having a client-server relationship to each other.
According to the technical scheme of the embodiment of the application, under the condition that the first image is obtained and the target main body is detected in the first image, the character content in the target main body is identified and/or the characteristic information of the target main body is extracted, and the target authenticity identification result of the target main body is determined based on the character content of the target main body and/or the characteristic information of the target main body. The target main body comprises at least one of a seal and a signature, the characteristic information can comprise at least one of a shape characteristic, a topology characteristic and a convolution characteristic, namely in the embodiment, the authenticity of the target main body can be determined by utilizing the text content of the target main body and/or at least one of the shape characteristic, the topology characteristic and the convolution characteristic of the target main body, the authenticity does not need to be identified through manual checking, and the authenticity identification efficiency can be improved. Meanwhile, the authenticity identification method of the embodiment can be used for checking the situation that errors are easy to generate when authenticity is identified through manual checking, and can improve identification accuracy.
It should be understood that various forms of the flows shown above may be used, with steps reordered, added, or deleted. For example, the steps described in the present application may be executed in parallel, sequentially, or in different orders, and the present invention is not limited thereto as long as the desired results of the technical solutions disclosed in the present application can be achieved.
The above-described embodiments should not be construed as limiting the scope of the present application. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions may be made in accordance with design requirements and other factors. Any modification, equivalent replacement, and improvement made within the spirit and principle of the present application shall be included in the protection scope of the present application.
Claims (14)
1. A method for authenticating genuineness, wherein the method comprises:
acquiring a first image;
under the condition that the first image is detected to comprise a target main body, identifying text content in the target main body and/or extracting feature information of the target main body, wherein the target main body comprises at least one of a seal and a signature, and the feature information comprises at least one of a shape feature, a topological feature and a convolution feature;
and determining the target authenticity identification result of the target main body based on the text content of the target main body and/or the characteristic information of the target main body.
2. The method according to claim 1, wherein the target body includes the stamp, the characteristic information of the target body includes a shape characteristic of the stamp;
the determining the target authenticity identification result of the target main body based on the text content of the target main body and/or the characteristic information of the target main body comprises the following steps:
comparing the text content of the seal with the real text content of the seal to obtain a first authenticity identification result of the seal;
determining a second authenticity identification result of the seal based on the shape characteristics of the seal;
and determining the target authenticity result according to the first authenticity identification result and the second authenticity identification result.
3. The method of claim 1, wherein the feature information comprises at least one of the topological feature and the convolution feature;
the determining the target authenticity identification result of the target main body based on the text content of the target main body and/or the characteristic information of the target main body comprises the following steps:
inputting the topological features of the target subject and the real topological features of the real subject corresponding to the target subject into a first neural network, and outputting a first real probability of the target subject through the first neural network; and/or the presence of a gas in the gas,
inputting the convolution characteristics of the target subject and the real convolution characteristics of the real subject corresponding to the target subject into a second neural network, and outputting a second real probability of the target subject through the second neural network;
and determining the target authenticity identification result based on the first real probability and/or the second real probability.
4. The method of claim 3, wherein the inputting the topological feature of the target subject and the true topological feature of the true subject to which the target subject corresponds to a first neural network through which a first true probability of the target subject is output comprises:
obtaining a first feature vector based on the topological feature of the target subject and the real topological feature;
and inputting the first feature vector into the first neural network to obtain the first true probability.
5. The method of claim 3, wherein the inputting the convolution features of the target subject and the true convolution features of the true subject to which the target subject corresponds to a second neural network through which a second true probability of the target subject is output comprises:
carrying out binarization processing on the target main body to obtain a first binarization main body;
performing character cutting on the first binarized main body to obtain a cut character of the first binarized main body;
extracting convolution characteristics of the cut character of the first binarized body through a first convolution neural network;
inputting the convolution characteristic of the cut character of the first binarized main body and the real convolution characteristic of the cut character of the real main body into the second neural network to obtain the second real probability.
6. The method of claim 1, wherein, in the case that the first image is detected to include a target subject, identifying text content in the target subject and/or extracting feature information of the target subject, further comprising:
extracting a subimage to be identified of the first image;
carrying out angle correction on the subimage to be identified to obtain a corrected subimage;
wherein, in the case that it is detected that the first image includes a target subject, recognizing text content in the target subject and/or extracting feature information of the target subject includes:
and under the condition that the corrected image is detected to comprise a target main body, recognizing the text content in the target main body and/or extracting the characteristic information of the target main body.
7. An authentication apparatus, wherein the apparatus comprises:
the acquisition module is used for acquiring a first image;
the processing module is used for identifying text content in a target main body and/or extracting feature information of the target main body under the condition that the first image is detected to comprise the target main body, wherein the target main body comprises at least one of a seal and a signature, and the feature information comprises at least one of a shape feature, a topological feature and a convolution feature;
and the determining module is used for determining the target authenticity identification result of the target main body based on the text content of the target main body and/or the characteristic information of the target main body.
8. The apparatus according to claim 7, wherein the target body includes the stamp, the characteristic information of the target body includes a shape characteristic of the stamp;
the determining module includes:
the comparison module is used for comparing the text content of the seal with the real text content of the seal to obtain a first authenticity identification result of the seal;
the first sub-determination module is used for determining a second authenticity identification result of the seal based on the shape characteristics of the seal;
and the second sub-determining module is used for determining the target authenticity result according to the first authenticity identification result and the second authenticity identification result.
9. The apparatus of claim 7, wherein the feature information comprises at least one of the topological feature and the convolution feature;
the determining module includes:
a probability determination module, configured to input the topological feature of the target subject and a real topological feature of a real subject corresponding to the target subject into a first neural network, and output a first real probability of the target subject through the first neural network; and/or, the convolution feature of the target subject and the real convolution feature of the real subject corresponding to the target subject are input into a second neural network, and a second real probability of the target subject is output through the second neural network;
and the third sub-determination module is used for determining the target authenticity identification result based on the first real probability and/or the second real probability.
10. The apparatus of claim 9, wherein the probability determination module comprises:
the feature vector determination module is used for obtaining a first feature vector based on the topological feature of the target subject and the real topological feature;
a first probability obtaining module, configured to input the first feature vector to the first neural network, so as to obtain the first true probability.
11. The apparatus of claim 9, wherein the probability determination module comprises:
the binarization module is used for carrying out binarization processing on the target main body to obtain a first binarization main body;
the cutting module is used for performing character cutting on the first binarized main body to obtain cut characters of the first binarized main body;
the characteristic extraction module is used for extracting the convolution characteristics of the cut characters of the first binary main body through a first convolution neural network;
a second true probability determining module, configured to input the convolution feature of the cut character of the first binarized main body and the true convolution feature of the cut character of the true main body into the second neural network, so as to obtain the second true probability.
12. The apparatus of claim 7, further comprising:
the subimage extraction module is used for extracting subimages to be identified of the first image;
the correction module is used for carrying out angle correction on the subimage to be identified to obtain a corrected subimage;
wherein, in the case that it is detected that the first image includes a target subject, recognizing text content in the target subject and/or extracting feature information of the target subject includes:
and under the condition that the corrected image is detected to comprise a target main body, recognizing the text content in the target main body and/or extracting the characteristic information of the target main body.
13. An electronic device, comprising:
at least one processor; and
a memory communicatively coupled to the at least one processor; wherein the content of the first and second substances,
the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the method of any one of claims 1-6.
14. A non-transitory computer readable storage medium having stored thereon computer instructions for causing the computer to perform the method of any one of claims 1-6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011065166.3A CN112115921A (en) | 2020-09-30 | 2020-09-30 | True and false identification method and device and electronic equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011065166.3A CN112115921A (en) | 2020-09-30 | 2020-09-30 | True and false identification method and device and electronic equipment |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112115921A true CN112115921A (en) | 2020-12-22 |
Family
ID=73798418
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011065166.3A Pending CN112115921A (en) | 2020-09-30 | 2020-09-30 | True and false identification method and device and electronic equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112115921A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113434912A (en) * | 2021-08-24 | 2021-09-24 | 共道网络科技有限公司 | Material compliance verification method and device |
| CN114255334A (en) * | 2021-12-13 | 2022-03-29 | 中国中医科学院中药研究所 | Shape feature acquisition device, database and identification system for traditional Chinese medicine |
| CN115063804A (en) * | 2022-06-29 | 2022-09-16 | 支付宝(杭州)信息技术有限公司 | Seal identification and anti-counterfeiting detection method and system |
| CN116109914A (en) * | 2023-04-07 | 2023-05-12 | 平安银行股份有限公司 | Method and device for identifying authenticity of bank running water image, electronic equipment and medium |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1178962A (en) * | 1996-10-04 | 1998-04-15 | 星友科技股份有限公司 | Automatic seal comparing and identifying system |
| CN101593324A (en) * | 2009-06-17 | 2009-12-02 | 浙江师范大学 | The network multi-level measures and procedures for the examination and approval and system based on dependable computing application technique |
| CN103617673A (en) * | 2013-12-10 | 2014-03-05 | 江苏国光信息产业股份有限公司 | Ultraviolet image characteristic-based check true and false identification system and method |
| CN106778586A (en) * | 2016-12-08 | 2017-05-31 | 武汉理工大学 | Offline handwriting signature verification method and system |
| CN108491870A (en) * | 2018-03-14 | 2018-09-04 | 平安科技（深圳）有限公司 | Electronic device, physical seal recognition methods and computer readable storage medium |
| CN109977750A (en) * | 2019-01-15 | 2019-07-05 | 平安科技（深圳）有限公司 | Seal true and false method of calibration, device and computer readable storage medium |
| WO2019174130A1 (en) * | 2018-03-14 | 2019-09-19 | 平安科技（深圳）有限公司 | Bill recognition method, server, and computer readable storage medium |
| CN110378351A (en) * | 2019-07-26 | 2019-10-25 | 中国工商银行股份有限公司 | Seal discrimination method and device |
| CN110765992A (en) * | 2019-09-23 | 2020-02-07 | 厦门大学 | Seal identification method, medium, equipment and device |
| CN111401372A (en) * | 2019-12-20 | 2020-07-10 | 国家电网有限公司 | Method for extracting and identifying image-text information of scanned document |
-
2020
- 2020-09-30 CN CN202011065166.3A patent/CN112115921A/en active Pending
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1178962A (en) * | 1996-10-04 | 1998-04-15 | 星友科技股份有限公司 | Automatic seal comparing and identifying system |
| CN101593324A (en) * | 2009-06-17 | 2009-12-02 | 浙江师范大学 | The network multi-level measures and procedures for the examination and approval and system based on dependable computing application technique |
| CN103617673A (en) * | 2013-12-10 | 2014-03-05 | 江苏国光信息产业股份有限公司 | Ultraviolet image characteristic-based check true and false identification system and method |
| CN106778586A (en) * | 2016-12-08 | 2017-05-31 | 武汉理工大学 | Offline handwriting signature verification method and system |
| CN108491870A (en) * | 2018-03-14 | 2018-09-04 | 平安科技（深圳）有限公司 | Electronic device, physical seal recognition methods and computer readable storage medium |
| WO2019174130A1 (en) * | 2018-03-14 | 2019-09-19 | 平安科技（深圳）有限公司 | Bill recognition method, server, and computer readable storage medium |
| CN109977750A (en) * | 2019-01-15 | 2019-07-05 | 平安科技（深圳）有限公司 | Seal true and false method of calibration, device and computer readable storage medium |
| CN110378351A (en) * | 2019-07-26 | 2019-10-25 | 中国工商银行股份有限公司 | Seal discrimination method and device |
| CN110765992A (en) * | 2019-09-23 | 2020-02-07 | 厦门大学 | Seal identification method, medium, equipment and device |
| CN111401372A (en) * | 2019-12-20 | 2020-07-10 | 国家电网有限公司 | Method for extracting and identifying image-text information of scanned document |
Non-Patent Citations (2)
| Title |
|---|
| 刘丰威;潘炜;韩丽丽;: "稽查中印章真伪识别智能算法", 中国高新科技, no. 13, 1 July 2020 (2020-07-01) * |
| 胥玉龙;张永梅;滑瑞敏;: "基于深度学习的离线手写签名真伪识别方法", 电脑知识与技术, no. 15, 25 May 2019 (2019-05-25) * |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113434912A (en) * | 2021-08-24 | 2021-09-24 | 共道网络科技有限公司 | Material compliance verification method and device |
| CN114255334A (en) * | 2021-12-13 | 2022-03-29 | 中国中医科学院中药研究所 | Shape feature acquisition device, database and identification system for traditional Chinese medicine |
| CN115063804A (en) * | 2022-06-29 | 2022-09-16 | 支付宝(杭州)信息技术有限公司 | Seal identification and anti-counterfeiting detection method and system |
| CN116109914A (en) * | 2023-04-07 | 2023-05-12 | 平安银行股份有限公司 | Method and device for identifying authenticity of bank running water image, electronic equipment and medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US10762376B2 (en) | Method and apparatus for detecting text | |
| CN111753727B (en) | Method, apparatus, device and readable storage medium for extracting structured information | |
| CN112115921A (en) | True and false identification method and device and electronic equipment | |
| CN109343920B (en) | Image processing method and device, equipment and storage medium thereof | |
| EP3274921B1 (en) | Multi-layer skin detection and fused hand pose matching | |
| US8483440B2 (en) | Methods and systems for verifying automatic license plate recognition results | |
| EP2783328B1 (en) | Text detection using multi-layer connected components with histograms | |
| CN106980856B (en) | Formula identification method and system and symbolic reasoning calculation method and system | |
| US20230005257A1 (en) | Illegal building identification method and apparatus, device, and storage medium | |
| EP1519302A2 (en) | System and method for detecting a hand-drawn object in electronic ink input | |
| CN112381183B (en) | Target detection method and device, electronic equipment and storage medium | |
| CN110717497B (en) | Image similarity matching method, device and computer readable storage medium | |
| CN113033543B (en) | Curve text recognition method, device, equipment and medium | |
| CN112380566A (en) | Method, apparatus, electronic device, and medium for desensitizing document image | |
| CN113627439A (en) | Text structuring method, processing device, electronic device and storage medium | |
| Lahiani et al. | Hand pose estimation system based on Viola-Jones algorithm for android devices | |
| CN103336579A (en) | Input method of wearable device and wearable device | |
| CN110147785B (en) | Image recognition method, related device and equipment | |
| CN113610809A (en) | Fracture detection method, fracture detection device, electronic device, and storage medium | |
| EP3979129A1 (en) | Object recognition method and apparatus, and electronic device and storage medium | |
| US20150186718A1 (en) | Segmentation of Overwritten Online Handwriting Input | |
| CN111552829A (en) | Method and apparatus for analyzing image material | |
| CN108255298B (en) | Infrared gesture recognition method and device in projection interaction system | |
| CN114663886A (en) | Text recognition method, model training method and device | |
| Uzair et al. | Electronet: An enhanced model for small-scale object detection in electrical schematic diagrams |
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 |