CN114241463A - Signature verification method and device, computer equipment and storage medium - Google Patents
Signature verification method and device, computer equipment and storage medium Download PDFInfo
- Publication number
- CN114241463A CN114241463A CN202111338188.7A CN202111338188A CN114241463A CN 114241463 A CN114241463 A CN 114241463A CN 202111338188 A CN202111338188 A CN 202111338188A CN 114241463 A CN114241463 A CN 114241463A
- Authority
- CN
- China
- Prior art keywords
- signature
- image
- feature
- matching
- target
- 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
- 238000012795 verification Methods 0.000 title claims abstract description 100
- 238000000034 method Methods 0.000 title claims abstract description 51
- 238000000605 extraction Methods 0.000 claims abstract description 35
- 238000012545 processing Methods 0.000 claims description 32
- 238000004590 computer program Methods 0.000 claims description 28
- 238000012549 training Methods 0.000 claims description 24
- 238000001914 filtration Methods 0.000 claims description 4
- 239000003550 marker Substances 0.000 claims 1
- 238000010586 diagram Methods 0.000 description 7
- 230000008569 process Effects 0.000 description 4
- 230000011218 segmentation Effects 0.000 description 3
- 238000013475 authorization Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000007781 pre-processing Methods 0.000 description 2
- OKTJSMMVPCPJKN-UHFFFAOYSA-N Carbon Chemical compound [C] OKTJSMMVPCPJKN-UHFFFAOYSA-N 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000015572 biosynthetic process Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 229910021389 graphene Inorganic materials 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 230000001788 irregular Effects 0.000 description 1
- 238000002372 labelling Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000003062 neural network model Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
Images
Landscapes
- Collating Specific Patterns (AREA)
Abstract
The application relates to a signature verification method, a signature verification device, a computer device and a storage medium. The method comprises the following steps: acquiring a signature file image to be processed; the signature file image comprises a target signature to be verified; identifying a region corresponding to the target signature from the signature file image based on preset character information corresponding to the target signature to obtain a signature region image; performing feature extraction on the character outline in the signature area image to obtain a character feature image corresponding to the target signature; inputting the character characteristic image into a pre-trained signature characteristic matching model to obtain a characteristic matching result; the feature matching result is used for determining a matching probability value corresponding to the target signature; and obtaining a signature verification result of the target signature based on the matching probability value and a preset verification matching threshold. By adopting the method, whether the signature to be verified is the real signature of the corresponding signer can be efficiently judged, the accuracy of the verification result can be ensured, and the verification efficiency of the validity of the signature is improved.
Description
Technical Field
The present application relates to the field of computer technologies, and in particular, to a signature verification method, apparatus, computer device, storage medium, and computer program product.
Background
At present, when the user signature in the entity file is judged, no matter the user signature is a paper signature or a signature of an electronic document, because the matching judgment of the signature handwriting is involved, a manual identification mode is usually adopted to verify the authenticity of the user signature in the entity file, but the accuracy of manual verification is low, the validity of the user signature cannot be ensured, and a large amount of manpower is consumed.
Therefore, the related art has a problem that manual authentication of a signature is inefficient.
Disclosure of Invention
In view of the above, it is necessary to provide a signature verification method, an apparatus, a computer device, a storage medium, and a computer program product that can solve the above problems.
In a first aspect, the present application provides a signature verification method, including:
acquiring a signature file image to be processed; the signature file image comprises a target signature to be verified;
identifying a region corresponding to the target signature from the signature file image based on preset character information corresponding to the target signature to obtain a signature region image;
performing feature extraction on the character outline in the signature area image to obtain a character feature image corresponding to the target signature;
inputting the character feature image into a pre-trained signature feature matching model to obtain a feature matching result; the feature matching result is used for determining a matching probability value corresponding to the target signature; the matching probability value and the signature truth degree form a positive correlation;
and obtaining a signature verification result of the target signature based on the matching probability value and a preset verification matching threshold.
In one embodiment, the identifying, based on the preset text information corresponding to the target signature, an area corresponding to the target signature from the signature file image to obtain a signature area image includes:
acquiring preset text information corresponding to the target signature; the preset text information is obtained based on a real signature image corresponding to the target signature;
based on the preset text information, identifying the image position of the target signature in the signature file image to determine the area corresponding to the target signature and the area mark coordinates thereof;
and obtaining the signature area image according to the area and the area mark coordinates thereof.
In one embodiment, the obtaining the signature region image according to the region and the region mark coordinates thereof includes:
performing preset image processing on the signature file image after the area calibration to obtain a processed signature file image; the preset image processing comprises graying processing, noise point removing processing and binarization processing;
and cutting the processed signature file image to obtain the signature area image based on the area mark coordinates.
In one embodiment, the performing feature extraction on the text outline in the signature region image to obtain a text feature image corresponding to the target signature includes:
determining a background image area in the signature area image;
filtering the background image area from the signature area image to obtain a character image area in the signature area image;
and performing feature extraction on the character outline in the character image area to obtain the character feature image.
In one embodiment, before the step of inputting the text feature image into a pre-trained signature feature matching model to obtain a feature matching result, the method further includes:
acquiring training sample data; the training sample data comprises a sample real signature image and a sample character characteristic image corresponding to at least one sample signature; the sample text feature image comprises a plurality of sample local feature images; the plurality of sample local characteristic images are obtained by extracting the sample real signature images;
and training a signature feature matching model to be trained based on the sample real signature image and the sample character feature image to obtain the pre-trained signature feature matching model.
In one embodiment, the text feature image includes a plurality of local feature images, and after the step of inputting the text feature image to a pre-trained signature feature matching model to obtain a feature matching result, the method further includes:
acquiring the number of images of the successfully matched local feature images according to the feature matching result;
obtaining the matching probability value based on the number of the images and the total number of the images of the plurality of local feature images;
the obtaining of the signature verification result of the target signature based on the matching probability value and a preset verification matching threshold includes:
if the matching probability value is larger than or equal to the verification matching threshold value, determining that the target signature is a real signature;
or, if the matching probability value is smaller than the verification matching threshold, determining that the target signature is a fake signature.
In a second aspect, the present application further provides a signature verification apparatus, the apparatus comprising:
the signature file image acquisition module is used for acquiring a signature file image to be processed; the signature file image comprises a target signature to be verified;
a signature region image obtaining module, configured to identify a region corresponding to the target signature from the signature file image based on preset text information corresponding to the target signature, so as to obtain a signature region image;
the character outline feature extraction module is used for extracting features of character outlines in the signature area images to obtain character feature images corresponding to the target signatures;
the model matching module is used for inputting the character feature image into a pre-trained signature feature matching model to obtain a feature matching result; the feature matching result is used for determining a matching probability value corresponding to the target signature; the matching probability value and the signature truth degree form a positive correlation
And the signature verification result obtaining module is used for obtaining the signature verification result of the target signature based on the matching probability value and a preset verification matching threshold value.
In a third aspect, the present application also provides a computer device. The computer device comprises a memory storing a computer program and a processor implementing the steps of the signature verification method as described above when executing the computer program.
In a fourth aspect, the present application further provides a computer-readable storage medium. The computer-readable storage medium, on which a computer program is stored, which computer program, when being executed by a processor, carries out the steps of the signature verification method as described above.
In a fifth aspect, the present application further provides a computer program product. The computer program product comprising a computer program which, when executed by a processor, carries out the steps of the signature verification method as described above.
The signature verification method, the device, the computer equipment, the storage medium and the computer program product are characterized in that a signature file image to be processed is obtained, the signature file image comprises a target signature to be verified, a region corresponding to the target signature is identified from the signature file image based on preset character information corresponding to the target signature to obtain a signature region image, then character outline in the signature region image is subjected to feature extraction to obtain a character feature image corresponding to the target signature, the character feature image is input to a pre-trained signature feature matching model to obtain a feature matching result, the feature matching result is used for determining a matching probability value corresponding to the target signature, the matching probability value is in positive correlation with the signature true degree, and then a signature verification result of the target signature is obtained based on the matching probability value and a preset verification matching threshold value, whether the signature to be verified is the real signature of the corresponding signer or not is efficiently judged, the accuracy of the verification result can be ensured, and the verification efficiency of the validity of the signature is improved.
Drawings
FIG. 1 is a schematic flow chart diagram illustrating a signature verification method in one embodiment;
FIG. 2 is a schematic flow chart diagram illustrating the model training steps in one embodiment;
FIG. 3 is a flow diagram of another signature verification method in one embodiment;
FIG. 4 is a block diagram of a signature verification device in one embodiment;
FIG. 5 is a diagram of the internal structure of a computer device, in one embodiment.
Detailed Description
In order to make the objects, technical solutions and advantages of the present application more apparent, the present application is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the present application and are not intended to limit the present application.
It should be noted that, the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data for presentation, analyzed data, etc.) referred to in this application are information and data authorized by the user or sufficiently authorized by each party; correspondingly, the application also provides a corresponding user authorization entrance for the user to select authorization or to select denial.
In an embodiment, as shown in fig. 1, a signature verification method is provided, and this embodiment is illustrated by applying the method to a server, and it is to be understood that the method may also be applied to a terminal, and may also be applied to a system including the terminal and the server, and is implemented by interaction between the terminal and the server. In this embodiment, the method includes the steps of:
as an example, the signature file image may be an image corresponding to an entity file of a signature to be verified, and may be obtained by scanning a paper entity file, or may be obtained by format conversion for an electronic entity file.
In practical application, an image corresponding to an entity file to be verified and signed may be acquired as a signature file image to be processed, where the signature file image may include a target signature to be verified.
102, identifying a region corresponding to the target signature from the signature file image based on preset character information corresponding to the target signature to obtain a signature region image;
the preset text information may be signature text information obtained based on a real signature image corresponding to the target signature, for example, text information corresponding to the real signature may be obtained in advance.
After the signature file image is obtained, an area corresponding to the target signature can be identified from the signature file image based on the preset text information corresponding to the target signature, the area can be marked as an interested area, and then a signature area image can be obtained according to the interested area, namely the image obtained after the interested area is marked in the signature file image.
In an example, based on machine vision and an image processing technology, according to preset text information corresponding to a target signature, an image area to be processed is outlined in a signature file image in a manner of a square frame, a circle, an ellipse, an irregular polygon, and the like, and is used as an area of interest (ROI) to further perform signature verification on the ROI.
103, extracting the characteristics of the character outline in the signature area image to obtain a character characteristic image corresponding to the target signature;
in specific implementation, feature extraction may be performed on the text outline in the signature region image to obtain text outline features corresponding to the target signature, and then the feature extraction result may be divided into a plurality of local feature images as text feature images corresponding to the target signature.
For example, the extracted character outline features of the signature to be verified can be divided into a plurality of feature image blocks (i.e., a plurality of local feature images) with the same size, so as to obtain a character feature image corresponding to the target signature.
after the character feature image is obtained, the character feature image can be input into a signature feature matching model which is pre-trained, a sample character feature image extracted according to a real signature image corresponding to a target signature can be pre-stored in the signature feature matching model, and then a feature matching result can be obtained by performing feature matching on the character feature image and the sample character feature image.
Specifically, a feature matching result obtained by performing feature matching on the character feature image and the sample character feature image can be used for determining a matching probability value corresponding to the target signature, the matching probability value and the signature authenticity degree form a positive correlation relationship, that is, the larger the matching probability value is, the higher the signature authenticity degree corresponding to the target signature is, and the target label can be judged to be a real signature or a forged signature based on the matching probability value.
And 105, obtaining a signature verification result of the target signature based on the matching probability value and a preset verification matching threshold.
In practical application, the matching probability value can be calculated according to the feature matching result, the probability that the signature to be verified is the real signature can be represented, the matching probability value can be compared with a preset verification matching threshold value, the signature verification result of the target signature is obtained, and the target signature is judged to be the real signature or the fake signature.
In one example, a feature extraction mode is adopted to perform character outline recognition on a target signature to be verified, then the extracted character feature image can be input to a signature feature matching model pre-trained on the basis of a real signature image of a signer corresponding to the target signature, and whether the target signature is the real signature of the corresponding signer can be judged by calculating the probability that the target signature to be verified is the real signature, so that the validity of the signature can be timely and accurately verified. The character outline characteristics based on the character image are identified, the identification speed can be accelerated during signature verification by utilizing the identification capability of the neural network model to be more sensitive to the overall shape of the character, the signature verification accuracy is improved, a manual identification mode is not needed, the manpower is saved, and the objectivity of a verification result is ensured.
According to the signature verification method, the region corresponding to the target signature is identified from the signature file image by obtaining the signature file image to be processed and based on the preset character information corresponding to the target signature, the signature region image is obtained, then the character outline in the signature region image is subjected to feature extraction, the character feature image corresponding to the target signature is obtained, the character feature image is input to the pre-trained signature feature matching model, the feature matching result is obtained, and then the signature verification result of the target signature is obtained based on the matching probability value and the preset verification matching threshold value, so that whether the signature to be verified is the real signature of the corresponding signer or not is judged efficiently, the accuracy of the verification result can be ensured, and the signature validity verification efficiency is improved.
In an embodiment, the identifying, based on the preset text information corresponding to the target signature, an area corresponding to the target signature from the signature file image to obtain a signature area image may include:
acquiring preset text information corresponding to the target signature; the preset text information is obtained based on a real signature image corresponding to the target signature; based on the preset text information, identifying the image position of the target signature in the signature file image to determine the area corresponding to the target signature and the area mark coordinates thereof; and obtaining the signature area image according to the area and the area mark coordinates thereof.
In practical application, for a target signature to be verified, a real signature image corresponding to the target signature can be obtained in advance, signature text information can be obtained by performing text recognition processing on the real signature image to serve as preset text information corresponding to the target signature, then, a signature file image acquired based on the preset text information can be traversed, a region to be processed containing the text information of the target signature to be verified is defined from the signature file image to serve as an area of interest, namely, the image position of the target signature is identified, and then, a signature region image can be obtained according to the area of interest and region mark coordinates thereof.
In an example, an OCR character recognition mode may be adopted to recognize character information (i.e., preset character information) corresponding to a real signature in a pre-acquired real signature image, and by screening and comparing the character information in a signature file image to be verified, a signature region including the character information may be found out as an area of interest, and then a coordinate point marking may be performed on the area of interest in the signature file image, so as to obtain an area marking image (i.e., a signature region image) converted from the signature file image.
In the embodiment, the preset text information corresponding to the target signature is obtained, the image position of the target signature is identified in the signature file image based on the preset text information to determine the area corresponding to the target signature and the area mark coordinates thereof, and then the signature area image is obtained according to the area and the area mark coordinates thereof.
In one embodiment, the obtaining the signature region image according to the region and the region mark coordinates thereof may include the following steps:
performing preset image processing on the signature file image after the area calibration to obtain a processed signature file image; the preset image processing comprises graying processing, noise point removing processing and binarization processing; and cutting the processed signature file image to obtain the signature area image based on the area mark coordinates.
After the image is calibrated, the signature file image after the area calibration can be subjected to graying processing to obtain a grayscale image, then noise point removal processing can be performed on the grayscale image to obtain a blurred image, the blurred image can be subjected to binarization to obtain a binary image (namely a processed signature file image), and then a region of interest (namely an area mark coordinate) marked in a coordinate point mode can be subjected to cutting processing to obtain a signature area image containing a target signature to be verified.
For example, because the signature file image can be a color image, a grayscale image, or a black-and-white image, the color features in the signature file image can be directly grayed to obtain the grayscale image without distinguishing the color features, and thus, after the graying, the interference of the color difference in the signature file image on the signature feature extraction process can be reduced.
For another example, noise in the signature file image can be removed, the gray-scale image is blurred by adopting a gaussian blur processing mode, the pixel value of the adjacent pixel of each pixel can be obtained, the pixel value of the pixel is corrected based on the pixel value of the adjacent pixel, and then the blurred image is obtained through image blur processing, so that the noise pixel in the signature handwriting can be removed, and the accuracy of signature verification and result judgment can be improved.
For another example, after the blurred image is obtained through the image blurring process, the blurred image may be subjected to binarization process, and may be cut according to the marked region of interest, so as to obtain a signature region image only including the target signature to be verified.
In the embodiment, the signature file image after the area calibration is subjected to preset image processing to obtain the processed signature file image, and then the signature area image is obtained by cutting from the processed signature file image based on the area marking coordinates, so that the interference condition of color difference in the signature file image on the signature characteristic extraction process can be reduced, noise pixels in signature handwriting are removed, and the accuracy of signature verification and result judgment is improved.
In an embodiment, the performing feature extraction on the text outline in the signature region image to obtain a text feature image corresponding to the target signature may include the following steps:
determining a background image area in the signature area image; filtering the background image area from the signature area image to obtain a character image area in the signature area image; and performing feature extraction on the character outline in the character image area to obtain the character feature image.
In practical application, for a preprocessed signature region image, a foreground region and a background region can be determined from the signature region image, and then a foreground image (namely a character image region) containing character outline features of a target signature can be extracted through soft segmentation matting processing in the signature region image, and the character outline features of the target signature in the foreground image can be subjected to feature extraction to obtain a character feature image.
In an example, image soft segmentation matting processing is performed on a signature region image, an alpha matching matting algorithm can be adopted, and a foreground region containing a character outline is segmented from a background region by performing finer foreground edge soft segmentation on a gray unknown region, so that a foreground image containing signature character outline features can be extracted, and character outline features in the foreground image can be further identified.
In an optional embodiment, a feature extraction result is obtained by performing feature extraction on a text outline of a target signature, and the feature extraction result can be used for signature feature matching identification to determine whether a signature to be verified is consistent with a real signature, and the feature extraction result includes, but is not limited to, signature text track features, density features of signature text in a signature image area, posture features of the signature text outline, text feature outline overlap ratio features, a starting point and an ending point of the signature text based on handwriting formation, and handwriting continuous stagnation point features, an inclination angle and an arc of the signature text in the signature area image, and spacing features between adjacent handwriting.
In the embodiment, the background image area in the signature area image is determined, the background image area is then filtered from the signature area image, the character image area in the signature area image is obtained, the character outline in the character image area is further subjected to feature extraction, the character feature image is obtained, the character outline feature of the target signature can be further identified based on the segmented character image area, and the accuracy of the signature validity identification result is improved.
In an embodiment, as shown in fig. 2, before the step of inputting the text feature image into a pre-trained signature feature matching model to obtain a feature matching result, the method may include the following steps:
in an example, a real signature image (i.e., a sample real signature image) of a signer corresponding to at least one group of signatures to be verified may be obtained, and by preprocessing the real signature image and extracting a text outline feature of the real signature, a plurality of local feature images with the same size may be selected as a plurality of sample local feature images based on a feature extraction result.
202, training a signature feature matching model to be trained based on the sample real signature image and the sample character feature image to obtain the pre-trained signature feature matching model;
after training sample data is obtained, a sample real signature image and a sample character feature image comprising a plurality of sample local feature images can be used as a training set, the training set is input to a neural network convolution model to be trained (namely, a signature feature matching model to be trained) for training, model verification can be performed by inputting any sample local feature image, a signature feature matching model to be trained can be further obtained, and therefore a plurality of sample local feature images are selected as sample data for model training according to a feature extraction result of the real signature image, and a signature feature matching model corresponding to the real signature image can be obtained.
In the embodiment, training is performed on the signature feature matching model to be trained by obtaining the training sample data and further based on the sample real signature image and the sample character feature image to obtain the pre-trained signature feature matching model, so that the signature feature matching model corresponding to the real signature image can be obtained, the recognition speed can be increased, and the accuracy of signature verification can be improved.
In one embodiment, the text feature image may include a plurality of local feature images, and after the step of inputting the text feature image to a pre-trained signature feature matching model to obtain a feature matching result, the method may include the following steps:
acquiring the number of images of the successfully matched local feature images according to the feature matching result; obtaining the matching probability value based on the number of the images and the total number of the images of the plurality of local feature images;
in practical application, the character outline features are divided into a plurality of feature image blocks (namely a plurality of local feature images) and input into a signature feature matching model for identification, the number of feature image blocks matched between a plurality of local feature images to be verified and a plurality of sample local feature images of a real signature image of a corresponding signer can be counted, namely the number of images of the local feature images which are successfully matched, and then the ratio of the number of matched feature image blocks to the total number of feature image blocks of the plurality of local feature images to be verified can be calculated to obtain the probability of a real signature, namely the matching probability value.
The obtaining of the signature verification result of the target signature based on the matching probability value and a preset verification matching threshold includes:
if the matching probability value is larger than or equal to the verification matching threshold value, determining that the target signature is a real signature; or, if the matching probability value is smaller than the verification matching threshold, determining that the target signature is a fake signature.
After the matching probability value is obtained, the calculated probability of the real signature can be compared with a preset verification matching threshold, when the matching probability value is larger than or equal to the verification matching threshold, the real signature can be judged, and when the matching probability value is smaller than the verification matching threshold, the forged signature can be judged.
In this embodiment, if the matching probability value is greater than or equal to the verification matching threshold, the target label name is determined to be a real signature, or if the matching probability value is less than the verification matching threshold, the target label name is determined to be a fake signature, objectivity of a verification result can be ensured, the real signature and the fake signature can be verified, and accuracy of signature verification is improved.
In one embodiment, as shown in FIG. 3, a flow diagram of another method of signature verification is provided. In this embodiment, the method includes the steps of:
in step 301, training sample data is acquired; the training sample data comprises a sample real signature image and a sample character characteristic image corresponding to at least one sample signature; the sample text feature image comprises a plurality of sample local feature images; the plurality of sample local characteristic images are obtained by extracting the sample real signature images. In step 302, a signature feature matching model to be trained is trained based on the sample real signature image and the sample text feature image, so as to obtain the pre-trained signature feature matching model. In step 303, a signature file image to be processed is acquired; the signature file image includes a target signature to be verified. In step 304, based on the preset text information corresponding to the target signature, an area corresponding to the target signature is identified from the signature file image, so as to obtain a signature area image. In step 305, feature extraction is performed on the character outline in the signature region image, so as to obtain a character feature image corresponding to the target signature. In step 306, the text feature image is input to a pre-trained signature feature matching model to obtain a feature matching result. In step 307, the number of images of the successfully matched local feature images is obtained according to the feature matching result. In step 308, the matching probability value is obtained based on the number of images and the total number of images of the local feature images. In step 309, if the matching probability value is greater than or equal to the verification matching threshold, determining that the target signature is a real signature; or, if the matching probability value is smaller than the verification matching threshold, determining that the target signature is a fake signature. It should be noted that, for the specific limitations of the above steps, reference may be made to the above specific limitations of a signature verification method, which are not described herein again.
It should be understood that, although the steps in the flowcharts related to the embodiments as described above are sequentially displayed as indicated by arrows, the steps are not necessarily performed sequentially as indicated by the arrows. The steps are not performed in the exact order shown and described, and may be performed in other orders, unless explicitly stated otherwise. Moreover, at least a part of the steps in the flowcharts related to the embodiments described above may include multiple steps or multiple stages, which are not necessarily performed at the same time, but may be performed at different times, and the execution order of the steps or stages is not necessarily sequential, but may be rotated or alternated with other steps or at least a part of the steps or stages in other steps.
Based on the same inventive concept, the embodiment of the present application further provides a signature verification apparatus for implementing the above-mentioned signature verification method. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme described in the above method, so specific limitations in one or more embodiments of the signature verification device provided below can be referred to the above limitations on the signature verification method, and are not described herein again.
In one embodiment, as shown in fig. 4, there is provided a signature verification apparatus including:
a signature file image obtaining module 401, configured to obtain a signature file image to be processed; the signature file image comprises a target signature to be verified;
a signature region image obtaining module 402, configured to identify a region corresponding to the target signature from the signature file image based on preset text information corresponding to the target signature, so as to obtain a signature region image;
a character outline feature extraction module 403, configured to perform feature extraction on a character outline in the signature region image to obtain a character feature image corresponding to the target signature;
a model matching module 404, configured to input the text feature image to a pre-trained signature feature matching model to obtain a feature matching result; the feature matching result is used for determining a matching probability value corresponding to the target signature; the matching probability value and the signature truth degree form a positive correlation
A signature verification result obtaining module 405, configured to obtain a signature verification result of the target signature based on the matching probability value and a preset verification matching threshold.
In one embodiment, the signature region image obtaining module 402 includes:
the character information acquisition submodule is used for acquiring preset character information corresponding to the target signature; the preset text information is obtained based on a real signature image corresponding to the target signature;
the region determining submodule is used for identifying the image position of the target signature in the signature file image based on the preset text information so as to determine a region corresponding to the target signature and region mark coordinates thereof;
and the region marking submodule is used for obtaining the signature region image according to the region and the region marking coordinates thereof.
In one embodiment, the region labeling submodule includes:
the image preprocessing unit is used for carrying out preset image processing on the signature file image after the area calibration to obtain a processed signature file image; the preset image processing comprises graying processing, noise point removing processing and binarization processing;
and the image cutting unit is used for cutting the signature area image from the processed signature file image based on the area mark coordinates.
In one embodiment, the text contour feature extraction module 403 includes:
a background image area determining submodule for determining a background image area in the signature area image;
the character image area obtaining submodule is used for filtering the background image area from the signature area image to obtain a character image area in the signature area image;
and the feature extraction submodule is used for performing feature extraction on the character outline in the character image area to obtain the character feature image.
In one embodiment, the apparatus further comprises:
the training sample data acquisition module is used for acquiring training sample data; the training sample data comprises a sample real signature image and a sample character characteristic image corresponding to at least one sample signature; the sample text feature image comprises a plurality of sample local feature images; the plurality of sample local characteristic images are obtained by extracting the sample real signature images;
and the model training module is used for training a signature feature matching model to be trained based on the sample real signature image and the sample character feature image to obtain the pre-trained signature feature matching model.
In one embodiment, the text feature image comprises a plurality of local feature images, and the apparatus further comprises:
the matching success image determining module is used for acquiring the image quantity of the successfully matched local feature images according to the feature matching result;
a matching probability value obtaining module, configured to obtain the matching probability value based on the number of images and the total number of images of the plurality of local feature images;
the signature verification result obtaining module 405 includes:
the signature judgment submodule is used for determining that the target signature is a real signature if the matching probability value is greater than or equal to the verification matching threshold; or, if the matching probability value is smaller than the verification matching threshold, determining that the target signature is a fake signature.
The modules in the signature verification device can be wholly or partially implemented by software, hardware and a combination thereof. The modules can be embedded in a hardware form or independent from a processor in the computer device, and can also be stored in a memory in the computer device in a software form, so that the processor can call and execute operations corresponding to the modules.
In one embodiment, a computer device is provided, which may be a server, the internal structure of which may be as shown in fig. 5. The computer device includes a processor, a memory, and a network interface connected by a system bus. Wherein the processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of an operating system and computer programs in the non-volatile storage medium. The database of the computer device is for storing signature verification data. The network interface of the computer device is used for communicating with an external terminal through a network connection. The computer program is executed by a processor to implement a signature verification method.
Those skilled in the art will appreciate that the architecture shown in fig. 5 is merely a block diagram of some of the structures associated with the disclosed aspects and is not intended to limit the computing devices to which the disclosed aspects apply, as particular computing devices may include more or less components than those shown, or may combine certain components, or have a different arrangement of components.
In one embodiment, a computer device is provided, comprising a memory and a processor, the memory having a computer program stored therein, the processor implementing the following steps when executing the computer program:
acquiring a signature file image to be processed; the signature file image comprises a target signature to be verified;
identifying a region corresponding to the target signature from the signature file image based on preset character information corresponding to the target signature to obtain a signature region image;
performing feature extraction on the character outline in the signature area image to obtain a character feature image corresponding to the target signature;
inputting the character feature image into a pre-trained signature feature matching model to obtain a feature matching result; the feature matching result is used for determining a matching probability value corresponding to the target signature; the matching probability value and the signature truth degree form a positive correlation;
and obtaining a signature verification result of the target signature based on the matching probability value and a preset verification matching threshold.
In one embodiment, the steps of the signature verification method in the other embodiments described above are also implemented when the computer program is executed by a processor.
In one embodiment, a computer-readable storage medium is provided, having a computer program stored thereon, which when executed by a processor, performs the steps of:
acquiring a signature file image to be processed; the signature file image comprises a target signature to be verified;
identifying a region corresponding to the target signature from the signature file image based on preset character information corresponding to the target signature to obtain a signature region image;
performing feature extraction on the character outline in the signature area image to obtain a character feature image corresponding to the target signature;
inputting the character feature image into a pre-trained signature feature matching model to obtain a feature matching result; the feature matching result is used for determining a matching probability value corresponding to the target signature; the matching probability value and the signature truth degree form a positive correlation;
and obtaining a signature verification result of the target signature based on the matching probability value and a preset verification matching threshold.
In one embodiment, the computer program when executed by the processor further implements the steps of the signature verification method in the other embodiments described above.
In one embodiment, a computer program product is provided, comprising a computer program which, when executed by a processor, performs the steps of:
acquiring a signature file image to be processed; the signature file image comprises a target signature to be verified;
identifying a region corresponding to the target signature from the signature file image based on preset character information corresponding to the target signature to obtain a signature region image;
performing feature extraction on the character outline in the signature area image to obtain a character feature image corresponding to the target signature;
inputting the character feature image into a pre-trained signature feature matching model to obtain a feature matching result; the feature matching result is used for determining a matching probability value corresponding to the target signature; the matching probability value and the signature truth degree form a positive correlation;
and obtaining a signature verification result of the target signature based on the matching probability value and a preset verification matching threshold.
In one embodiment, the computer program when executed by the processor further implements the steps of the signature verification method in the other embodiments described above.
It will be understood by those skilled in the art that all or part of the processes of the methods of the embodiments described above can be implemented by hardware instructions of a computer program, which can be stored in a non-volatile computer-readable storage medium, and when executed, can include the processes of the embodiments of the methods described above. Any reference to memory, database, or other medium used in the embodiments provided herein may include at least one of non-volatile and volatile memory. The nonvolatile Memory may include Read-Only Memory (ROM), magnetic tape, floppy disk, flash Memory, optical Memory, high-density embedded nonvolatile Memory, resistive Random Access Memory (ReRAM), Magnetic Random Access Memory (MRAM), Ferroelectric Random Access Memory (FRAM), Phase Change Memory (PCM), graphene Memory, and the like. Volatile Memory can include Random Access Memory (RAM), external cache Memory, and the like. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM), among others. The databases referred to in various embodiments provided herein may include at least one of relational and non-relational databases. The non-relational database may include, but is not limited to, a block chain based distributed database, and the like. The processors referred to in the embodiments provided herein may be general purpose processors, central processing units, graphics processors, digital signal processors, programmable logic devices, quantum computing based data processing logic devices, etc., without limitation.
The technical features of the above embodiments can be arbitrarily combined, and for the sake of brevity, all possible combinations of the technical features in the above embodiments are not described, but should be considered as the scope of the present specification as long as there is no contradiction between the combinations of the technical features.
The above-mentioned embodiments only express several embodiments of the present application, and the description thereof is more specific and detailed, but not construed as limiting the scope of the present application. It should be noted that, for a person skilled in the art, several variations and modifications can be made without departing from the concept of the present application, which falls within the scope of protection of the present application. Therefore, the protection scope of the present application shall be subject to the appended claims.
Claims (10)
1. A method of signature verification, the method comprising:
acquiring a signature file image to be processed; the signature file image comprises a target signature to be verified;
identifying a region corresponding to the target signature from the signature file image based on preset character information corresponding to the target signature to obtain a signature region image;
performing feature extraction on the character outline in the signature area image to obtain a character feature image corresponding to the target signature;
inputting the character feature image into a pre-trained signature feature matching model to obtain a feature matching result; the feature matching result is used for determining a matching probability value corresponding to the target signature; the matching probability value and the signature truth degree form a positive correlation;
and obtaining a signature verification result of the target signature based on the matching probability value and a preset verification matching threshold.
2. The method according to claim 1, wherein the identifying a region corresponding to the target signature from the signature file image based on preset text information corresponding to the target signature to obtain a signature region image comprises:
acquiring preset text information corresponding to the target signature; the preset text information is obtained based on a real signature image corresponding to the target signature;
based on the preset text information, identifying the image position of the target signature in the signature file image to determine the area corresponding to the target signature and the area mark coordinates thereof;
and obtaining the signature area image according to the area and the area mark coordinates thereof.
3. The method of claim 2, wherein obtaining the signature region image according to the region and its region marker coordinates comprises:
performing preset image processing on the signature file image after the area calibration to obtain a processed signature file image; the preset image processing comprises graying processing, noise point removing processing and binarization processing;
and cutting the processed signature file image to obtain the signature area image based on the area mark coordinates.
4. The method according to claim 1, wherein the performing feature extraction on the text outline in the signature region image to obtain a text feature image corresponding to the target signature comprises:
determining a background image area in the signature area image;
filtering the background image area from the signature area image to obtain a character image area in the signature area image;
and performing feature extraction on the character outline in the character image area to obtain the character feature image.
5. The method of claim 1, wherein before the step of inputting the text feature image into a pre-trained signature feature matching model to obtain a feature matching result, the method further comprises:
acquiring training sample data; the training sample data comprises a sample real signature image and a sample character characteristic image corresponding to at least one sample signature; the sample text feature image comprises a plurality of sample local feature images; the plurality of sample local characteristic images are obtained by extracting the sample real signature images;
and training a signature feature matching model to be trained based on the sample real signature image and the sample character feature image to obtain the pre-trained signature feature matching model.
6. The method according to any one of claims 1 to 5, wherein the text feature image comprises a plurality of local feature images, and after the step of inputting the text feature image into a pre-trained signature feature matching model to obtain a feature matching result, the method further comprises:
acquiring the number of images of the successfully matched local feature images according to the feature matching result;
obtaining the matching probability value based on the number of the images and the total number of the images of the plurality of local feature images;
the obtaining of the signature verification result of the target signature based on the matching probability value and a preset verification matching threshold includes:
if the matching probability value is larger than or equal to the verification matching threshold value, determining that the target signature is a real signature;
or, if the matching probability value is smaller than the verification matching threshold, determining that the target signature is a fake signature.
7. A signature verification apparatus, the apparatus comprising:
the signature file image acquisition module is used for acquiring a signature file image to be processed; the signature file image comprises a target signature to be verified;
a signature region image obtaining module, configured to identify a region corresponding to the target signature from the signature file image based on preset text information corresponding to the target signature, so as to obtain a signature region image;
the character outline feature extraction module is used for extracting features of character outlines in the signature area images to obtain character feature images corresponding to the target signatures;
the model matching module is used for inputting the character feature image into a pre-trained signature feature matching model to obtain a feature matching result; the feature matching result is used for determining a matching probability value corresponding to the target signature; the matching probability value and the signature truth degree form a positive correlation
And the signature verification result obtaining module is used for obtaining the signature verification result of the target signature based on the matching probability value and a preset verification matching threshold value.
8. A computer device comprising a memory and a processor, the memory storing a computer program, characterized in that the processor, when executing the computer program, implements the steps of the method of any of claims 1 to 6.
9. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out the steps of the method of any one of claims 1 to 6.
10. A computer program product comprising a computer program, characterized in that the computer program realizes the steps of the method of any one of claims 1 to 6 when executed by a processor.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111338188.7A CN114241463A (en) | 2021-11-12 | 2021-11-12 | Signature verification method and device, computer equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111338188.7A CN114241463A (en) | 2021-11-12 | 2021-11-12 | Signature verification method and device, computer equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114241463A true CN114241463A (en) | 2022-03-25 |
Family
ID=80749200
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111338188.7A Pending CN114241463A (en) | 2021-11-12 | 2021-11-12 | Signature verification method and device, computer equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114241463A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113378609A (en) * | 2020-03-10 | 2021-09-10 | 中国移动通信集团辽宁有限公司 | Method and device for identifying agent signature |
| CN115083024A (en) * | 2022-06-17 | 2022-09-20 | 平安银行股份有限公司 | Signature identification method, device, medium and equipment based on region division |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6424728B1 (en) * | 1999-12-02 | 2002-07-23 | Maan Ammar | Method and apparatus for verification of signatures |
| CN110619274A (en) * | 2019-08-14 | 2019-12-27 | 深圳壹账通智能科技有限公司 | Identity verification method and device based on seal and signature and computer equipment |
| CN111178290A (en) * | 2019-12-31 | 2020-05-19 | 上海眼控科技股份有限公司 | Signature verification method and device |
| CN112036323A (en) * | 2020-09-01 | 2020-12-04 | 中国银行股份有限公司 | Signature handwriting identification method, client and server |
| US20210124905A1 (en) * | 2019-10-28 | 2021-04-29 | Microsoft Technology Licensing, Llc | Signature verification |
| CN113095156A (en) * | 2021-03-23 | 2021-07-09 | 西安深信科创信息技术有限公司 | Double-current network signature identification method and device based on inverse gray scale mode |
| WO2021164481A1 (en) * | 2020-02-18 | 2021-08-26 | 深圳壹账通智能科技有限公司 | Neural network model-based automatic handwritten signature verification method and device |
| CN113313092A (en) * | 2021-07-29 | 2021-08-27 | 太平金融科技服务(上海)有限公司深圳分公司 | Handwritten signature recognition method, and claims settlement automation processing method, device and equipment |
| CN113378609A (en) * | 2020-03-10 | 2021-09-10 | 中国移动通信集团辽宁有限公司 | Method and device for identifying agent signature |
-
2021
- 2021-11-12 CN CN202111338188.7A patent/CN114241463A/en active Pending
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6424728B1 (en) * | 1999-12-02 | 2002-07-23 | Maan Ammar | Method and apparatus for verification of signatures |
| CN110619274A (en) * | 2019-08-14 | 2019-12-27 | 深圳壹账通智能科技有限公司 | Identity verification method and device based on seal and signature and computer equipment |
| US20210124905A1 (en) * | 2019-10-28 | 2021-04-29 | Microsoft Technology Licensing, Llc | Signature verification |
| CN111178290A (en) * | 2019-12-31 | 2020-05-19 | 上海眼控科技股份有限公司 | Signature verification method and device |
| WO2021164481A1 (en) * | 2020-02-18 | 2021-08-26 | 深圳壹账通智能科技有限公司 | Neural network model-based automatic handwritten signature verification method and device |
| CN113378609A (en) * | 2020-03-10 | 2021-09-10 | 中国移动通信集团辽宁有限公司 | Method and device for identifying agent signature |
| CN112036323A (en) * | 2020-09-01 | 2020-12-04 | 中国银行股份有限公司 | Signature handwriting identification method, client and server |
| CN113095156A (en) * | 2021-03-23 | 2021-07-09 | 西安深信科创信息技术有限公司 | Double-current network signature identification method and device based on inverse gray scale mode |
| CN113313092A (en) * | 2021-07-29 | 2021-08-27 | 太平金融科技服务(上海)有限公司深圳分公司 | Handwritten signature recognition method, and claims settlement automation processing method, device and equipment |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113378609A (en) * | 2020-03-10 | 2021-09-10 | 中国移动通信集团辽宁有限公司 | Method and device for identifying agent signature |
| CN113378609B (en) * | 2020-03-10 | 2023-07-21 | 中国移动通信集团辽宁有限公司 | Agent proxy signature identification method and device |
| CN115083024A (en) * | 2022-06-17 | 2022-09-20 | 平安银行股份有限公司 | Signature identification method, device, medium and equipment based on region division |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110414507B (en) | License plate recognition method and device, computer equipment and storage medium | |
| CN106156766B (en) | Method and device for generating text line classifier | |
| US8750619B2 (en) | Character recognition | |
| CN111680690B (en) | Character recognition method and device | |
| CN110619274A (en) | Identity verification method and device based on seal and signature and computer equipment | |
| JP2020525947A (en) | Manipulated image detection | |
| CN108197644A (en) | A kind of image-recognizing method and device | |
| CN114241463A (en) | Signature verification method and device, computer equipment and storage medium | |
| CN109389115B (en) | Text recognition method, device, storage medium and computer equipment | |
| CN110909809A (en) | Card image identification method based on deep learning | |
| CN112507782A (en) | Text image recognition method and device | |
| CN113158895A (en) | Bill identification method and device, electronic equipment and storage medium | |
| CN111027545A (en) | Card picture mark detection method and device, computer equipment and storage medium | |
| CN112200789B (en) | Image recognition method and device, electronic equipment and storage medium | |
| CN113435219B (en) | Anti-counterfeiting detection method and device, electronic equipment and storage medium | |
| CN108921006B (en) | Method for establishing handwritten signature image authenticity identification model and authenticity identification method | |
| CN111191584B (en) | Face recognition method and device | |
| US6694059B1 (en) | Robustness enhancement and evaluation of image information extraction | |
| CN115410191A (en) | Text image recognition method, device, equipment and storage medium | |
| CN111986176B (en) | Crack image identification method, system, terminal and readable storage medium | |
| CN116152079A (en) | Image processing method and image processing model training method | |
| CN113239738B (en) | Image blurring detection method and blurring detection device | |
| CN110895849A (en) | Method and device for cutting and positioning crown word number, computer equipment and storage medium | |
| CN114529570A (en) | Image segmentation method, image identification method, user certificate subsidizing method and system | |
| CN111626244B (en) | Image recognition method, device, electronic equipment and 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 |