Signature detection method and device, equipment, storage medium and program product

Abstract

The embodiment of the invention provides a signature detection method and device, equipment, a storage medium and a program product, and relates to the field of financial science and technology or other related fields. The method comprises the steps of obtaining a to-be-detected signature of a target object, wherein the to-be-detected signature comprises at least one to-be-detected word; performing detection processing on each to-be-detected character to obtain stroke information and structure information of each to-be-detected character; determining a pre-stored signature of the target object in a signature library, wherein the pre-stored signature comprises pre-stored words corresponding to the to-be-detected words; according to the stroke information of each to-be-detected character, the structure information of each to-be-detected character, the stroke information of the corresponding pre-stored character and the structure information of the corresponding pre-stored character, a detection result of the to-be-detected signature is determined, and the detection result indicates that the to-be-detected signature is a real signature or an unreal signature of the target object. According to the scheme, the signature detection accuracy is improved.

Description

Technical Field

[0001] The present application relates to the field of financial technology or other related fields, and in particular to a signature detection method, device, equipment, storage medium and program product. Background Art

[0002] Detecting handwritten signatures is an important way to authenticate and confirm identity. In some scenarios, it is necessary to verify whether the handwritten signature is a genuine signature written by the user.

[0003] In the related art, the gesture corresponding to each stroke in the handwritten signature is usually compared with the gesture corresponding to the corresponding stroke in the real signature to determine the authenticity of the handwritten signature. However, the accuracy of the above method in detecting handwritten signatures is low. Summary of the invention

[0004] The embodiments of the present application provide a signature detection method, apparatus, device, storage medium and program product to improve the accuracy of handwritten signature detection.

[0005] In a first aspect, an embodiment of the present application provides a signature detection method, including:

[0006] Obtaining a signature to be detected of a target object, wherein the signature to be detected includes at least one word to be detected;

[0007] Perform detection processing on each character to be detected respectively to obtain the stroke information and structure information of each character to be detected;

[0008] Determine a pre-stored signature of the target object in a preset signature library, wherein the pre-stored signature includes pre-stored words corresponding to each word to be detected;

[0009] Based on the stroke information of each character to be detected, the structural information of each character to be detected, the stroke information of the corresponding pre-stored characters and the structural information of the corresponding pre-stored characters, the detection result of the signature to be detected is determined, and the detection result indicates whether the signature to be detected is a real signature or a non-real signature of the target object.

[0010] In a second aspect, an embodiment of the present application provides a signature detection device, the device comprising an acquisition module, a processing module, a first determination module and a second determination module, wherein:

[0011] An acquisition module, used for acquiring a signature to be detected of a target object, wherein the signature to be detected includes at least one word to be detected;

[0012] A processing module is used to perform detection processing on each character to be detected, and obtain stroke information and structure information of each character to be detected;

[0013] A first determination module is used to determine a pre-stored signature of a target object in a preset signature library, wherein the pre-stored signature includes pre-stored words corresponding to each word to be detected;

[0014] The second determination module is used to determine the detection result of the signature to be detected based on the stroke information of each character to be detected, the structural information of each character to be detected, the stroke information of the corresponding pre-stored characters, and the structural information of the corresponding pre-stored characters, and the detection result indicates whether the signature to be detected is a real signature or a non-real signature of the target object.

[0015] In a third aspect, an embodiment of the present application provides a signature detection device, including:

[0016] at least one processor; and

[0017] a memory communicatively connected to at least one processor; wherein,

[0018] The memory stores instructions that can be executed by at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the signature detection method involved in the first aspect and any possible implementation manner.

[0019] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores computer execution instructions, which, when executed by a processor, are used to implement the signature detection method involved in the first aspect and any possible implementation method.

[0020] In a fifth aspect, an embodiment of the present application provides a computer program product, including a computer program, which, when executed by a processor, implements the signature detection method involved in the first aspect and any possible implementation method.

[0021] The signature detection method, device, equipment, storage medium and program product provided in the embodiment of the present application, after obtaining the signature to be detected of the target object, detects and processes each character to be detected included in the signature to be detected, obtains the stroke information and structure information of each character to be detected, and compares the stroke information and structure information of each character to be detected with the stroke information and structure information of the corresponding pre-stored characters in the preset signature library, to obtain the detection result of the signature to be detected. The above method compares each character to be detected in the signature to be detected with the corresponding pre-stored characters from the two dimensions of the stroke and structure of the signature to be detected, determines the detection result of the signature to be detected, and improves the accuracy of signature detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0023] Figure 1A schematic diagram of applicable scenarios provided for embodiments of the present application;

[0024] Figure 2 A flow chart of a signature detection method provided in an embodiment of the present application;

[0025] Figure 3 A schematic diagram of a flow chart for performing detection processing on each word to be detected provided in an embodiment of the present application;

[0026] Figure 4 A schematic diagram of a word to be detected provided in an embodiment of the present application;

[0027] Figure 5 A schematic diagram of the structure of a component of a word to be detected provided in an embodiment of the present application;

[0028] Figure 6 A schematic diagram of the overall distribution of characters to be detected provided in an embodiment of the present application;

[0029] Figure 7 A schematic diagram of a process for determining a detection result of a signature to be detected provided in an embodiment of the present application;

[0030] Figure 8 A schematic diagram of the structure of a signature detection device provided in an embodiment of the present application;

[0031] Fig. 9 A schematic diagram of the structure of a signature detection device provided in an embodiment of the present application.

[0032] The above drawings have shown clear embodiments of the present application, which will be described in more detail later. These drawings and text descriptions are not intended to limit the scope of the present application in any way, but to illustrate the concept of the present application to those skilled in the art by referring to specific embodiments. DETAILED DESCRIPTION

[0033] Exemplary embodiments will be described in detail herein, examples of which are shown in the accompanying drawings. When the following description refers to the drawings, the same numbers in different drawings represent the same or similar elements unless otherwise indicated. The implementations described in the following exemplary embodiments do not represent all implementations consistent with the present application. Instead, they are merely examples of devices and methods consistent with some aspects of the present application as detailed in the appended claims.

[0034] 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 used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, storage, use, processing, transmission, provision, disclosure and application of the relevant data comply with the relevant laws, regulations and standards of the relevant countries and regions, take necessary confidentiality measures, do not violate public order and good morals, and provide corresponding operation entrances for users to choose to authorize or refuse.

[0035] In addition, this application involves big data analysis of user information (including but not limited to personal biometrics, identity data, consumption data, asset data, electronic terminal operation data, etc.), and the use of artificial intelligence technology for automated decision-making, and a technical solution for making decisions that have a significant impact on personal rights and interests based on the results of automated decision-making. The application provides users with corresponding operation entrances for them to choose to agree or reject the results of automated decision-making; if the user chooses to reject, the expert decision-making process will be entered.

[0036] It should be noted that the signature detection method, device, equipment, storage medium and program product provided in this application can be used in the field of financial technology or other related fields, and can also be used in any field other than the field of financial technology or other related fields. The application field of the signature detection method, device, equipment, storage medium and program product in this application is not limited.

[0037] For ease of understanding, the following Figure 1 , briefly describe the application scenarios to which the embodiments of the present application are applicable.

[0038] Figure 1 A schematic diagram of an applicable scenario provided for an embodiment of the present application, such as Figure 1 As shown, it includes an electronic screen 11 and a processor 12, wherein the electronic screen 11 can be an electronic screen or a handwriting board of a terminal such as a mobile phone, a computer, or a tablet, and the target object can sign on the electronic screen 11. The electronic screen 11 and the processor 12 can belong to the same device or two different devices.

[0039] In the case where the electronic screen 11 and the processor 12 belong to the same device, the processor 12 obtains the signature to be detected of the target object in response to the touch operation of the target object on the electronic screen 11 .

[0040] In the case where the electronic screen 11 and the processor 12 belong to two different devices, such as Figure 1 As shown, the electronic screen 11 may be the screen of the terminal 13 . The terminal 13 obtains the signature to be detected of the target object in response to the touch operation of the target object on the electronic screen 11 , and sends the signature to be detected to the processor 12 .

[0041] The signature to be detected can be a user's handwritten signature. Detecting the authenticity of handwritten signatures is of great significance and can be applied to different scenarios and fields. For example, in the scenario where a user handles a business, it is usually necessary for the user to write his or her own signature for business confirmation. By detecting the user's handwritten signature, it can be confirmed whether the user handling the business is the real person, reducing the impact caused by the unauthentic handwritten signature. Relevant businesses may include, for example, bank card application services, document signing and confirmation services, and so on.

[0042] It should be noted that Figure 1 It is just an example to illustrate a system architecture diagram, and it is not a limitation of the system architecture diagram.

[0043] In the related art, the signature to be detected is usually detected to obtain the direction of the stroke corresponding to at least one stroke in the signature to be detected, and then the direction of the stroke corresponding to at least one stroke is compared with the direction of the stroke corresponding to at least one stroke in the real signature to obtain the number of strokes with the same direction. When the number of strokes with the same direction is greater than or equal to the stroke number threshold, the authenticity of the signature to be detected is determined. However, the accuracy of signature detection achieved by the above method is low.

[0044] In response to the above technical problems, in an embodiment of the present application, each character to be detected in the signature to be detected of the target object can be detected to obtain the stroke information and structure information of each character to be detected, and based on the stroke information and structure information of each character to be detected, each character to be detected and the corresponding pre-stored character are compared from two different perspectives of stroke and structure, thereby improving the accuracy of signature detection.

[0045] The technical solution of the present application and how the technical solution of the present application solves the above-mentioned technical problems are described in detail below with specific embodiments. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be repeated in some embodiments. The embodiments of the present application will be described below in conjunction with the accompanying drawings.

[0046] Figure 2 A flow chart of a signature detection method provided in an embodiment of the present application. Figure 2 , the method may include the following steps:

[0047] S21. Obtain a signature to be detected of the target object, where the signature to be detected includes at least one word to be detected.

[0048] The execution entity of the embodiment of this application can be the background server of the signature detection system, or a signature detection device set in the server. The signature detection device can be implemented by software, or by a combination of software and hardware.

[0049] Among them, the signature to be detected can be a handwritten signature to be subjected to signature detection, and the target object can be the signature object corresponding to the signature to be detected. The word to be detected is a word included in the signature to be detected. For example, assuming that the signature to be detected is "eleven", at least one word to be detected included in the signature to be detected is respectively "ten" and "one".

[0050] S22. Perform detection processing on each word to be detected to obtain the stroke information and structure information of each word to be detected.

[0051] Each word to be detected includes at least one stroke. The stroke information of the word to be detected is used to indicate the positions of at least one stroke in the word to be detected, the timings of at least one stroke, and the positional relationship between any two strokes in at least one stroke, etc. The structure information of the word to be detected is used to indicate the overall distribution of the word to be detected, etc.

[0052] For example, assuming that the words to be detected are respectively "ten", performing detection processing on the word to be detected "ten", the obtained stroke information is used to indicate the positions and timings of the two strokes, namely the horizontal stroke and the vertical stroke, in the word to be detected "ten", and the positional relationship between the horizontal stroke and the vertical stroke, etc., and the obtained structure information is used to indicate the overall distribution of the word to be detected "ten", etc.

[0053] S23. Determine the pre-stored signature of the target object in the preset signature library. The pre-stored signature includes the pre-stored words corresponding to each word to be detected.

[0054] The preset signature library includes the basic information and pre-stored signatures of at least one object respectively. The pre-stored signature of the target object is the true signature of the target object pre-stored in the preset signature library.

[0055] The pre-stored signature of the target object can be determined in the preset signature library in the following manner: Obtain the basic information of the target object. The basic information of the target object is used to uniquely identify the target object. Determine the object in the preset signature library with the same basic information as the target object, and determine the pre-stored signature of this object as the pre-stored signature of the target object.

[0056] For example, it is assumed that the preset signature library includes basic information and pre-stored signatures of object 1, object 2, and object 3, and the basic information of object 1 is xxx, and the pre-stored signature of object 1 is "eleven", the basic information of object 2 is xxy, and the pre-stored signature of object 2 is "twelve", and the basic information of object 3 is yyx, and the pre-stored signature of object 3 is "twelve". The basic information of the target object is xxy, and the signature to be detected of the target object is "twelve". Then, according to the basic information of the target object, the object with the same basic information as the target object is determined in the preset signature library to be object 2, and the pre-stored signature "twelve" of object 2 is determined as the pre-stored signature of the target object. At this time, the "ten" in the pre-stored signature "twelve" of the target object is the pre-stored word corresponding to the word "ten" to be detected, and the "two" in the pre-stored signature "twelve" of the target object is the pre-stored word corresponding to the word "two" to be detected.

[0057] S24. Determine the detection result of the signature to be detected based on the stroke information of each character to be detected, the structural information of each character to be detected, the stroke information of the corresponding pre-stored characters, and the structural information of the corresponding pre-stored characters. The detection result indicates whether the signature to be detected is a real signature or a non-real signature of the target object.

[0058] Each pre-stored character includes at least one stroke, and the stroke information of the pre-stored character is used to indicate the position of at least one stroke in the pre-stored character, the time sequence of at least one stroke, and the position relationship between any two strokes in at least one stroke, etc. The structural information of the pre-stored character is used to indicate the overall distribution of the pre-stored character, etc.

[0059] It should be noted that the preset signature library includes stroke information and structure information of pre-stored characters corresponding to each detected character.

[0060] According to the stroke information of each character to be detected, the structural information of each character to be detected, the stroke information of the corresponding pre-stored character, and the structural information of the corresponding pre-stored character, the method of determining the detection result of the signature to be detected is as follows: the stroke information of each character to be detected is compared with the stroke information of the corresponding pre-stored character to obtain the comparison result of the stroke information; the structural information of each character to be detected is compared with the structural information of the corresponding pre-stored character to obtain the comparison result of the structural information. Finally, according to the comparison result of the stroke information between each character to be detected and the corresponding pre-stored character, and the comparison result of the structural information, the detection result of the signature to be detected is determined.

[0061] For example, assuming that the signature to be detected is "twelve", the stroke information of the character "ten" to be detected is compared with the stroke information of the corresponding pre-stored character "ten", and the structural information of the character "ten" to be detected is compared with the structural information of the corresponding pre-stored character "ten", and the comparison result of the stroke information and structural information between the character "ten" to be detected and the corresponding pre-stored character "ten" is obtained; the stroke information of the character "two" to be detected is compared with the stroke information of the corresponding pre-stored character "two", and the structural information of the character "two" to be detected is compared with the structural information of the corresponding pre-stored character "two", and the comparison result of the stroke information and structural information between the character "two" to be detected and the corresponding pre-stored character "two" is obtained; according to the comparison result of the stroke information and structural information between the character "ten" to be detected and the corresponding pre-stored character "ten", as well as the comparison result of the stroke information and structural information between the character "two" to be detected and the corresponding pre-stored character "two", the detection result of the signature "twelve" to be detected is determined.

[0062] exist Figure 2 In the embodiment shown, for the signature to be detected of the target object, each character to be detected in the signature to be detected is detected and processed, and the stroke information and structure information of each character to be detected are obtained respectively, and the stroke information and structure information of each character to be detected are respectively compared with the stroke information and structure information of the corresponding pre-stored characters in the preset signature library, and the detection result of the signature to be detected is determined according to the comparison result of the stroke information and structure information between each character to be detected and the corresponding pre-stored characters. In the above method, each character to be detected is compared with the corresponding pre-stored characters from the two aspects of the strokes and structure of each character to be detected in the signature to be detected, thereby improving the accuracy of signature detection.

[0063] exist Figure 2 Based on the embodiment shown below, combined with Figure 3 The detailed process of detecting and processing each character to be detected and obtaining the stroke information and structure information of each character to be detected is further introduced.

[0064] Figure 3 For a flow chart of a detection process for each word to be detected provided in the embodiment of the present application, please refer to Figure 3 , the method may include the following steps:

[0065] S31. Perform timing detection processing and coordinate detection processing on at least one stroke to obtain writing timing and coordinate information of each stroke, and at least one stroke is a stroke included in the character to be detected.

[0066] For any one of at least one stroke included in each character to be detected, performing timing detection processing on this stroke can be to record the writing timing of this stroke during the writing process, and the writing timing of this stroke can be the writing order of this stroke during the writing process. Performing coordinate detection processing on this stroke can be to detect at least one coordinate information of this stroke on the writing plane, and the coordinate information of this stroke is used to indicate at least one coordinate of this stroke on the writing plane. The at least one coordinate includes the coordinates of at least one point on this stroke, and the at least one point on this stroke includes the starting point and the ending point of this stroke, and can also include other points. The signature trajectory of the character to be detected can be determined through the coordinate information of each stroke in the character to be detected.

[0067] For example, assume that the character to be detected is "人" (person), and the strokes included in the character to be detected are "撇" (left-falling stroke) and "捺" (right-falling stroke) respectively. Performing timing detection processing on these two strokes of "撇" and "捺", it is obtained that the writing timing of "撇" is the first stroke, and the writing timing of "捺" is the second stroke; performing coordinate detection processing on these two strokes of "撇" and "捺" respectively, it is obtained that at least one coordinate information of "撇" on the writing plane is (x1, y1), (x2, y2), and at least one coordinate information of "捺" on the writing plane is (x3, y3), (x4, y4). According to the coordinate information of these two strokes of "撇" and "捺" in the character to be detected, the signature trajectory of the character to be detected is determined as U = {u1, u2}, where u1 represents the stroke coordinates of the stroke with the writing timing of the first stroke, that is, the stroke coordinates of the "撇" stroke, and u1 = {(x1, y1), (x2, y2)}; u2 represents the coordinates of the stroke with the writing timing of the second stroke, that is, the coordinates of the "捺" stroke, and u2 = {(x3, y3), (x4, y4)}.

[0068] S32. Obtain the stroke information of the character to be detected according to the writing timing and coordinate information of each stroke. The stroke information of the character to be detected includes the stroke timing vector of the character to be detected, the stroke direction of each stroke, and the relative position relationship between each stroke.

[0069] The stroke timing vector of the character to be detected is used to represent the writing timing of each stroke in the character to be detected. The stroke timing vector of the character to be detected can be a one-dimensional vector, and each element in the stroke timing vector of the character to be detected is used to represent the writing timing of each stroke in the detection. For example, assume that the character to be detected is "人" (person), and the strokes included in the character to be detected are "撇" (left-falling stroke) and "捺" (right-falling stroke). The stroke timing vector of the character to be detected is [1, 2], where 1 is used to represent that the writing timing of the "撇" stroke is the first stroke, and 2 is used to represent that the writing timing of the "捺" stroke is the second stroke.

[0070] The stroke direction of each stroke is used to represent the stroke trend of each stroke in the character to be detected during the writing process, and the relative position relationship between each stroke is used to represent the relative position relationship between any one stroke and other strokes in the character to be detected.

[0071] According to the writing timing and coordinate information of each stroke, the method for obtaining the stroke information of the character to be detected is as follows: Generate a stroke timing vector of the character to be detected according to the writing timing and type of each stroke; Determine the stroke segment of each stroke according to the starting coordinate and ending coordinate of each stroke; Determine the stroke direction of each stroke as the angle between the stroke segment of each stroke and the preset direction; Determine the relative position relationship according to the stroke segment, starting coordinate and ending coordinate of each stroke.

[0072] The type of each stroke is used to represent the shape of each stroke. For example, the type of each stroke can be types such as horizontal, vertical, left-falling stroke, right-falling stroke, etc. The starting coordinate of each stroke can be the coordinate of the first point of each stroke on the writing plane, the ending coordinate of each stroke can be the coordinate of the last point of each stroke on the writing plane, and the stroke segment of each stroke can be the connection line between the starting point and the ending point of each stroke.

[0073] It can be combined with Figure 4 for understanding. Figure 4 For a schematic diagram of a character to be detected provided by an embodiment of the present application, please refer to Figure 4 , the character to be detected is "人" which includes strokes with types of "left-falling stroke" and "right-falling stroke" respectively. Among them, the starting point of the "left-falling stroke" is O, the ending point is P, and the coordinate of point O is (x1, y1), the coordinate of point P is (x2, y2), that is, the starting coordinate of the "left-falling stroke" is (x1, y1), the ending coordinate is (x2, y2). Connect the starting point and the ending point of the "left-falling stroke" to obtain the stroke segment of the "left-falling stroke" as OP; the starting point of the "right-falling stroke" is M, the ending point is N, and the coordinate of point M is (x3, y3), the coordinate of point N is (x4, y4), that is, the starting coordinate of the "right-falling stroke" is (x3, y3), the ending coordinate is (x4, y4). Connect the starting point and the ending point of the "right-falling stroke" to obtain the stroke segment of the "right-falling stroke" as MN; the preset direction is direction S, then the stroke direction of the "left-falling stroke" is the angle a between the stroke segment OP and the preset direction S, and the stroke direction of the "right-falling stroke" is the angle b between the stroke segment MN and the preset direction S.

[0074] According to the writing sequence of each stroke and the type of each stroke, the method of generating the stroke sequence vector of the word to be detected can be as follows: Determine the type of each stroke in the word to be detected, determine the number of elements in the stroke sequence vector according to the number of strokes included in the word to be detected, determine the stroke type corresponding to each element in the stroke sequence vector according to the type of each stroke in the word to be detected, and generate the stroke sequence vector according to the writing sequence of each stroke, the number of elements in the stroke sequence vector, and the stroke type corresponding to each element in the stroke sequence vector.

[0075] For example, assume that the word to be detected is "十" (ten). The types of the strokes in the word to be detected "十" are "horizontal" and "vertical" respectively. That is, it is determined that the number of elements in the stroke sequence vector is 2. The stroke type corresponding to the first element in the stroke sequence vector is "horizontal", and the stroke type corresponding to the second element is "vertical". The writing sequence of the "horizontal" stroke in the word to be detected "十" is the first stroke, and the writing sequence of the "vertical" stroke is the second stroke. Then the generated stroke sequence vector is [1, 2], where 1 represents the writing sequence of the "horizontal" stroke in the word to be detected, and 2 represents the writing sequence of the "vertical" stroke in the word to be detected.

[0076] The method of determining the relative position relationship according to the stroke segments, starting coordinates, and ending coordinates of each stroke is as follows: Determine the relative position relationship between each stroke according to the stroke segments of each stroke. The relative position relationship between each stroke can be position relationships such as parallel, intersecting, and perpendicular. The relative position between each stroke can be, for example, the relative position relationship between the i-th stroke and the (i - 1)-th stroke.

[0077] Determine the relative position relationship between the starting points of each stroke according to the starting coordinates of each stroke. The relative position relationship between the starting points of each stroke can be position relationships such as directly above, directly below, 10 degrees to the lower left, etc. The relative position relationship between the starting points of each stroke can be, for example, the relative position relationship between the starting coordinate of the i-th stroke relative to the starting coordinate of the (i - 1)-th stroke and / or the relative position relationship between the starting coordinate of the i-th stroke relative to the starting coordinate of the (i - 2)-th stroke, where the value of i is less than or equal to the total number of strokes in the word to be detected.

[0078] Determine the relative position relationship between the ending points of each stroke according to the ending coordinates of each stroke. The relative position relationship between the ending points of each stroke can be position relationships such as directly above, directly below, 10 degrees to the lower left, etc. The relative position relationship between the ending points of each stroke can be, for example, the relative position relationship between the ending coordinate of the i-th stroke relative to the ending coordinate of the (i - 1)-th stroke and / or the relative position relationship between the ending coordinate of the i-th stroke relative to the ending coordinate of the (i - 2)-th stroke, where the value of i is less than or equal to the number of strokes in the word to be detected.

[0079] S33. Perform a structural detection process on the word to be detected to determine the components that make up the word to be detected and at least one detection block corresponding to the word to be detected.

[0080] A component refers to the smallest unit that makes up a character. For example, it can be understood as the radical, the radical of a character, etc. The word to be detected includes at least one component that makes up the word to be detected. When the number of components that make up the word to be detected in the word to be detected is greater than 1, the word to be detected is a compound character. Compound characters can be, for example, "good", "of", "Han", "forest", etc. Taking the word to be detected as "good" as an example, the components that make up the word to be detected are "female" and "son".

[0081] For each word to be detected, after dividing the word to be detected into grids of a fixed size, the detection block corresponding to the word to be detected can be the grid that divides the word to be detected.

[0082] For example, assume that the word to be detected is "of", and the word to be detected is divided into 4 grids of a fixed size, namely Grid 1, Grid 2, Grid 3, and Grid 4. Perform a structural detection process on the word to be detected "of" to determine that the components of the word to be detected are "white" and "spoon", and at least one detection block corresponding to the word to be detected "of" are Grid 1, Grid 2, Grid 3, and Grid 4 respectively.

[0083] S34. Obtain the structural information of the word to be detected according to the components that make up the word to be detected and at least one detection block corresponding to the word to be detected. The structural information includes the component distribution information and the overall distribution information of the word to be detected.

[0084] The method for obtaining the structural information of the word to be detected is as follows: Obtain at least one of the number, center position, and size of the components that make up the word to be detected; obtain the component distribution information of the word to be detected according to at least one of the number, center position, and size of the components that make up the word to be detected; obtain the number of sampling points in each detection block and the total number of sampling points of the word to be detected; obtain the overall distribution information of the word to be detected according to the number of sampling points in each detection block and the total number of sampling points.

[0085] The center position of the component that makes up the word to be detected can be the center coordinates of the component that makes up the word to be detected, and the size of the component that makes up the word to be detected can be the product of the width and height of the component that makes up the word to be detected. Among them, the width of the component that makes up the word to be detected is the horizontal distance between the rightmost point and the leftmost point of the component that makes up the word to be detected, and the height of the component that makes up the word to be detected is the vertical distance between the highest point and the lowest point of the component that makes up the word to be detected.

[0086] It can be combined with Figure 5 for understanding. Figure 5 This is a schematic diagram of the component structure of a word to be detected provided by an embodiment of the present application. Please refer to Figure 5 , the character to be detected is "的". Among them, the center point of the component "白" is E, and the center point of the component "勺" is F. The coordinates of point E are (s1, m1), and the coordinates of point F are (s2, m2). Then the center position of the component "白" is (s1, m1), and the center position of the component "勺" is (s2, m2). Taking the component "白" as an example, the highest point of the component "白" is R1(s3, m3), the lowest point is R2(s4, m4), the leftmost point is V1(s5, m5), and the rightmost point is V2(s6, m6). Then the height of the component "白" is the vertical distance H between R1 and R2 = |m3 - m4|, and the width of the component "白" is the horizontal distance L between V1 and V2 = |s6 - s5|. Thus, the size of the component "白" is H * L.

[0087] The way to obtain the component distribution information of the character to be detected according to at least one of the number, center position, and size of the components constituting the character to be detected can be as follows: When the number of components constituting the character to be detected is greater than 1, according to the center positions of the components constituting the character to be detected, determine the relative positions between the components and the distances between the centers of each pair of components. The relative positions between the components can be left - right positions, up - down positions, etc. When the relative position between the components is a left - right position, determine the proportion of the distance between the centers of each pair of components in the overall width of the character to be detected. When the relative position between the components is an up - down position, determine the proportion of the distance between the centers of each pair of components in the overall height of the character to be detected. According to the size of each component constituting the character to be detected, determine the proportion of the size of each component constituting the character to be detected in the overall size of the character to be detected.

[0088] For example, assume the character to be detected is "的", the overall length of the character to be detected is 6, and the overall height of the character to be detected is 4. Then the overall size of the character to be detected is 4 * 6. The components constituting the character to be detected are "白" and "勺" respectively. It is determined that the number of components constituting the character to be detected is 2. Among them, the center position of the component "白" is (2, 2), the size of the component "白" is 3 * 4, the center position of the component "勺" is (5, 2), and the size of the component "勺" is 3 * 4. Since the ordinates of the center positions of the two components "白" and "勺" are the same, the abscissa of the component "白" is 2, and the abscissa of the component "勺" is 5. Then it is determined that the relative position between the component "白" and the component "勺" is a left - right position, and the distance between the centers of the component "白" and the component "勺" is ||5 - 2|| = 3. It is determined that the proportion of the distance between the centers of each pair of components in the overall width of the character to be detected is 50%, the proportion of the size of the component "白" in the overall size of the character to be detected is 50%, and the proportion of the size of the component "勺" in the overall size of the character to be detected is 50%.

[0089] Sample the characters to be detected on the writing plane. The sampling process can be random sampling, uniform sampling, etc. Obtain the number of sampling points in each detection block and the total number of sampling points of the characters to be detected. Among them, the sum of the number of sampling points in all detection blocks corresponding to the characters to be detected is the same as the total number of sampling points of the characters to be detected. For example, assume that the detection blocks corresponding to the characters to be detected are detection block A, detection block B, and detection block C, the number of sampling points in detection block A is a, the number of sampling points in detection block B is b, the number of sampling points in detection block C is c, and the total number of sampling points of the characters to be detected is d, then a + b + c = d.

[0090] The method for obtaining the overall distribution information of the characters to be detected based on the number of sampling points in each detection block and the total number of sampling points is as follows: Determine the proportion of the number of sampling points in each detection block to the total number of sampling points according to the number of sampling points in the detection block and the total number of sampling points, and generate the overall distribution information of the characters to be detected. The overall distribution information of the characters to be detected can be represented by a one-dimensional vector, and each element in the vector corresponds to the proportion of the number of sampling points in each block to the total number of sampling points.

[0091] It can be combined with Figure 6 for understanding. Figure 6 For an overall distribution schematic diagram of the characters to be detected provided by an embodiment of this application, please refer to Figure 6 , the character to be detected is "的". The detection blocks corresponding to the character "的" are detection block A, detection block B, and detection block C respectively. Among them, the sampling points in detection block A are sampling point 1, sampling point 2, and sampling point 3, the sampling points in detection block B are sampling point 4, sampling point 4, sampling point 6, sampling point 7, and sampling point 8, and the sampling points in detection block C are sampling point 9 and sampling point 10. That is, the number of sampling points in detection block A is 3, the number of sampling points in detection block B is 5, the number of sampling points in detection block C is 2, and the total number of sampling points of the characters to be detected is 10. Then the proportion of the number of sampling points in detection block A to the total number of sampling points is 30%, the proportion of the number of sampling points in detection block B to the total number of sampling points is 50%, and the proportion of the number of sampling points in detection block C to the total number of sampling points is 20%. Then the overall distribution information of the characters to be detected can be represented by the vector [30%, 50%, 20%].

[0092] In Figure 3In the embodiment shown, at least one stroke of each character to be detected is detected and processed to obtain the stroke information of the character to be detected, and the structure of each character to be detected is detected and processed to obtain the overall distribution information of the character to be detected. Secondly, when the number of components constituting the character to be detected is greater than 1, the components constituting the character to be detected are detected to obtain the component distribution information of the character to be detected. In the above method, each character to be detected is detected and analyzed from three aspects: the strokes, components and overall structure of the character to be detected, to obtain the stroke information, component distribution information and overall distribution information of the character to be detected. Combining the stroke information, component distribution information and overall distribution information of the character to be detected can better understand and identify the writing habits, structural distribution, etc. of the character to be detected, thereby further improving the accuracy of signature detection.

[0093] Based on the above embodiments, Figure 7 The process of comparing the stroke information and structure information of the character to be detected with the stroke information and structure information of the corresponding pre-stored characters to determine the detection result of the signature to be detected is described in detail.

[0094] Figure 7 For a flow chart of determining the detection result of a signature to be detected provided in an embodiment of the present application, see Figure 7 , the method may include the following steps:

[0095] S71 . Determine the stroke similarity between each character to be detected and the corresponding pre-stored character according to the stroke information of each character to be detected and the stroke information of the corresponding pre-stored character.

[0096] The stroke information of the pre-stored characters is pre-stored in a preset signature library, and the method for determining the stroke information of the pre-stored characters is consistent with the method for determining the stroke information of the character to be detected.

[0097] The method for determining the stroke similarity between each character to be detected and the corresponding pre-stored character is as follows: determine the vector distance between the stroke timing vector of each character to be detected and the stroke timing vector of the corresponding pre-stored character; determine the number of strokes with different stroke directions between each character to be detected and the corresponding pre-stored character according to the stroke direction of at least one stroke included in each character to be detected and the stroke direction of at least one stroke included in the corresponding pre-stored character; determine the number of different relative position relationships between each character to be detected and the corresponding pre-stored character according to the relative position relationship between at least one stroke included in each character to be detected and the relative position relationship between at least one stroke included in the corresponding pre-stored character; the stroke similarity includes the vector distance, the number of strokes with different stroke directions, and the number of different relative position relationships.

[0098] The Hausdorff distance between the stroke timing vector of the character to be detected and the stroke timing vector of the corresponding pre-stored character can be calculated to determine it as the vector distance between the stroke timing vector of the character to be detected and the stroke timing vector of the corresponding pre-stored character. For each character to be detected, the vector distance of the stroke timing vector between the character to be detected and the corresponding pre-stored character is used to represent the difference in the stroke writing timing dimension between the character to be detected and the pre-stored character.

[0099] For any stroke in each stroke of each character to be detected, if the angle between the stroke segment of the stroke and the preset direction, and the product of the angle between the stroke segment of the stroke in the corresponding pre-stored character and the preset direction is greater than 0, it means that the stroke directions of the character to be detected and the corresponding pre-stored character are the same; if the angle between the stroke segment of the stroke and the preset direction, and the product of the angle between the stroke segment of the stroke in the corresponding pre-stored character and the preset direction is less than or equal to 0, it means that the stroke directions of the character to be detected and the corresponding pre-stored character are different. That is, for any stroke in each character to be detected, assuming that the angle between the stroke segment of the stroke and the preset direction is θ1, and the angle between the stroke segment of the stroke in the corresponding pre-stored character and the preset direction is θ2, if θ1*θ2>0, it means that the stroke directions of the stroke between the character to be detected and the corresponding pre-stored character are the same; if θ1*θ2≤0, it means that the stroke directions of the stroke between the character to be detected and the corresponding pre-stored character are different.

[0100] Through the above method, the number of strokes with different stroke directions between each character to be detected and the corresponding pre-stored character can be determined. For each character to be detected, the number of strokes with different stroke directions between the character to be detected and the corresponding pre-stored character is used to represent the difference in stroke direction dimension between the character to be detected and the corresponding pre-stored character.

[0101] The method for determining the number of different relative position relationships between each character to be detected and the corresponding pre-stored character can be as follows: compare the relative position relationship between each stroke in each character to be detected with the relative position relationship between each stroke in the corresponding pre-stored character, and obtain the number of different relative position relationships between each stroke between the character to be detected and the corresponding pre-stored character; compare the relative position relationship between the starting point coordinates of each stroke in the character to be detected with the relative position relationship between the starting point coordinates of each stroke in the corresponding pre-stored character, and obtain the number of different relative position relationships between the starting point coordinates of each stroke between the character to be detected and the corresponding pre-stored character; compare the relative position relationship between the end point coordinates of each stroke in the character to be detected with the relative position relationship between the end point coordinates of each stroke in the corresponding pre-stored character, and obtain the number of different relative position relationships between the end point coordinates of each stroke between the character to be detected and the corresponding pre-stored character.

[0102] Finally, the number of different relative position relationships between the strokes of the character to be detected and the corresponding pre-stored character, the number of different relative position relationships between the starting coordinates of the strokes of the character to be detected and the corresponding pre-stored character, and the number of different relative position relationships between the end coordinates of the strokes of the character to be detected and the corresponding pre-stored character are added together to obtain the number of different relative position relationships. For each character to be detected, the number of different relative positions between the character to be detected and the corresponding pre-stored character is used to represent the difference in the relative position dimension between the character to be detected and the corresponding pre-stored character.

[0103] S72: Determine the structural similarity between each to-be-detected word and the corresponding pre-stored word according to the structural information of each to-be-detected word and the structural information of the corresponding pre-stored word.

[0104] The structural information of the pre-stored word is pre-stored in a preset signature library, and the method of determining the structural information of the pre-stored word is consistent with the method of determining the structural information of the word to be detected.

[0105] The method of determining the structural similarity between each word to be detected and the corresponding pre-stored word is as follows: according to the component distribution information of each word to be detected and the component distribution information of the corresponding pre-stored word, the number of inconsistent components between each word to be detected and the corresponding pre-stored word is determined; according to the overall distribution information of each word to be detected and the overall distribution information of the corresponding pre-stored word, the difference between each word to be detected and the corresponding pre-stored word is determined; the structural similarity includes the number of inconsistent components between each word to be detected and the corresponding pre-stored word, and the difference between each word to be detected and the corresponding pre-stored word.

[0106] The method for determining the number of inconsistent components between each word to be detected and the corresponding pre-stored word can be as follows: compare the relative positions of the components in each word to be detected with the relative positions of the components in the corresponding pre-stored word, and obtain the number of different relative positions between the components; compare the ratio of the distance between the centers of two components in each word to be detected to the overall height or width of the word to be detected, and the ratio of the distance between the centers of two components in the corresponding pre-stored word to the overall height or width of the corresponding pre-stored word, and obtain the number of different ratios of the distance between the centers of two components to the overall height or width; compare the ratio of the size of each component in each word to be detected to the overall size of the word to be detected, and the ratio of the size of each component in the corresponding pre-stored word to the overall size of the word to be detected, and obtain the number of different ratios of the size of each component to the overall size of the word to be detected.

[0107] Finally, the number of different relative positions between components, the number of different ratios of the distance between the centers of two components to the overall height or length, and the number of different ratios of the size of each component to the overall size of the word to be detected are added to obtain the number of inconsistent components between each word to be detected and the corresponding pre-stored word. For each word to be detected, the number of inconsistent components between the word to be detected and the corresponding pre-stored word is used to represent the difference in component distribution dimension between the word to be detected and the corresponding pre-stored word.

[0108] The difference between each word to be detected and the corresponding pre-stored word is the difference between the overall distribution information of each word to be detected and the overall distribution information of the corresponding pre-stored word, and the difference between the overall distribution information of each word to be detected and the corresponding pre-stored word d5=||Q w -Q t || / (||Q w ||+||Q t ||), where d5 is the difference between the overall distribution information of the word to be detected and the corresponding pre-stored word, Q w is the overall distribution information of the word to be detected, Q t is the overall distribution information of the pre-stored words corresponding to the word to be detected, ||Q w -Q t || means Q w -Q t The norm of ||Q w || means Q w The norm of ||Q t || means Q t The norm may be a Manhattan (L1) norm, a Euclidean (L2) norm or an infinite norm, etc. For each word to be detected, the difference between the overall distribution information of the word to be detected and the overall distribution information of the corresponding pre-stored word is used to represent the difference between the word to be detected and the corresponding pre-stored word in terms of overall distribution.

[0109] S73, determining a detection result according to the stroke similarity between each character to be detected and the corresponding pre-stored character, and the structure similarity between each character to be detected and the corresponding pre-stored character.

[0110] The detection result can be determined in the following manner: respectively multiply the vector distance between the stroke timing vector of each character to be detected and the stroke timing vector of the corresponding pre-stored character, the number of strokes with different stroke directions between each character to be detected and the corresponding pre-stored character, the number of different relative positional relationships between each character to be detected and the corresponding pre-stored character, the number of inconsistent components between each character to be detected and the corresponding pre-stored character, and the difference between each character to be detected and the corresponding pre-stored character by the corresponding preset weights, and then subtract the above products from the preset similarity full score value to obtain the similarity score between each character to be detected and the corresponding pre-stored character. Finally, the similarity score between each character to be detected and the corresponding pre-stored character is compared with the similarity score threshold. When the similarity score between each character to be detected and the corresponding pre-stored character in the signature to be detected is greater than or equal to the similarity score threshold, it is determined that the detection result indicates that the signature to be detected is a true signature of the target object; when the similarity score between any one or more characters to be detected and the corresponding pre-stored character in the signature to be detected is less than the similarity score threshold, it is determined that the detection result indicates that the signature to be detected is a non-authentic signature of the target object.

[0111] For example, assume that the full score of the preset similarity is 100. The signature to be detected is "好的" (hǎo de). The vector distance between the stroke timing vectors of the character "好" (hǎo) to be detected and the corresponding pre-stored character is d1, the number of strokes with different stroke directions between the character "好" (hǎo) to be detected and the corresponding pre-stored character is d2, the number of different relative position relationships between the character "好" (hǎo) to be detected and the corresponding pre-stored character is d3, the number of inconsistent components between the character "好" (hǎo) to be detected and the corresponding pre-stored character is d4, and the difference between the character "好" (hǎo) to be detected and the corresponding pre-stored character is d5. The vector distance between the stroke timing vectors of the character "的" (de) to be detected and the corresponding pre-stored character is f1, the number of strokes with different stroke directions between the character "的" (de) to be detected and the corresponding pre-stored character is f2, the number of different relative position relationships between the character "的" (de) to be detected and the corresponding pre-stored character is f3, the number of inconsistent components between the character "的" (de) to be detected and the corresponding pre-stored character is f4, and the difference between the character "的" (de) to be detected and the corresponding pre-stored character is f5. And the preset weight of the vector distance between the stroke timing vectors of the character to be detected and the corresponding pre-stored character is k1, the preset weight of the number of strokes with different stroke directions between the character to be detected and the corresponding pre-stored character is k2, the preset weight of the number of different relative position relationships between the character to be detected and the corresponding pre-stored character is k3, the preset weight of the number of inconsistent components between the character to be detected and the corresponding pre-stored character is k4, and the preset weight of the difference between the character to be detected and the corresponding pre-stored character is k5. Then the similarity score F1 between the character "好" (hǎo) to be detected and the corresponding pre-stored character is F1 = 100 - k1 * d1 - k2 * d2 - k3 * d3 - k4 * d4 - k5 * d5, and the similarity score F2 between the character "的" (de) to be detected and the corresponding pre-stored character is F2 = 100 - k1 * f1 - k2 * f2 - k3 * f3 - k4 * f4 - k5 * f5.

[0112] Compare the similarity score F1 between the character "好" (hǎo) to be detected and the corresponding pre-stored character and the similarity score F2 between the character "的" (de) to be detected and the corresponding pre-stored character with the similarity score threshold δ respectively. When both the similarity score F1 and the similarity score F2 are greater than or equal to the similarity score threshold δ, determine that the detection result of the signature to be detected is used to indicate that the signature to be detected is the true signature of the target object; when the similarity score F1 is less than the similarity score threshold δ, and / or, the similarity score F2 is less than the similarity score threshold δ, determine that the detection result of the signature to be detected is used to indicate that the signature to be detected is a non-true signature of the target object.

[0113] At Figure 7In the embodiment shown, the stroke information and structural information of the character to be detected are compared with the stroke information and structural information of the corresponding pre-stored character, respectively, to obtain the stroke similarity and structural similarity between the character to be detected and the corresponding pre-stored character, and the similarity score between the character to be detected and the corresponding pre-stored character is determined based on the stroke similarity and structural similarity between the character to be detected and the corresponding pre-stored character. When the similarity score between each character to be detected and the corresponding pre-stored character in the signature to be detected is greater than or equal to the similarity score threshold, the detection result of the signature to be detected is determined to indicate that the signature to be detected is the real signature of the target object. In the above method, each character to be detected and the corresponding pre-stored character in each signature to be detected are compared from the two aspects of the stroke and structural distribution of the signature to be detected, thereby improving the accuracy of signature detection.

[0114] Figure 8 This is a schematic diagram of the structure of a signature detection device provided in an embodiment of the present application. Figure 8 The signature detection device 80 may include: an acquisition module 81, a processing module 82, a first determination module 83 and a second determination module 84, wherein:

[0115] An acquisition module 81 is used to acquire a signature to be detected of a target object, wherein the signature to be detected includes at least one word to be detected;

[0116] The processing module 82 is used to perform detection processing on each character to be detected to obtain stroke information and structure information of each character to be detected;

[0117] A first determination module 83 is used to determine a pre-stored signature of a target object in a preset signature library, wherein the pre-stored signature includes pre-stored words corresponding to each word to be detected;

[0118] The second determination module 84 is used to determine the detection result of the signature to be detected based on the stroke information of each character to be detected, the structural information of each character to be detected, the stroke information of the corresponding pre-stored characters, and the structural information of the corresponding pre-stored characters, and the detection result indicates whether the signature to be detected is a real signature or a non-real signature of the target object.

[0119] In a possible implementation, the processing module 82 is specifically configured to:

[0120] Performing timing detection processing and coordinate detection processing on at least one stroke respectively to obtain writing timing and coordinate information of each stroke, wherein at least one stroke is a stroke included in the character to be detected;

[0121] According to the writing time sequence and coordinate information of each stroke, the stroke information of the character to be detected is obtained, and the stroke information of the character to be detected includes the stroke time sequence vector of the character to be detected, the stroke direction of each stroke, and the relative position relationship between the strokes;

[0122] Performing structural detection processing on the word to be detected to determine the components constituting the word to be detected and at least one detection block corresponding to the word to be detected;

[0123] According to the components constituting the word to be detected and at least one detection block corresponding to the word to be detected, structural information of the word to be detected is obtained, and the structural information includes component distribution information and overall distribution information of the word to be detected.

[0124] In a possible implementation, the processing module 82 is specifically configured to:

[0125] Generate a stroke timing vector of the character to be detected according to the writing timing of each stroke and the type of each stroke;

[0126] Determine the stroke line segment of each stroke according to the starting point coordinates and the end point coordinates of each stroke;

[0127] Determine the angle between the stroke line segment of each stroke and the preset direction as the stroke direction of each stroke;

[0128] The relative position relationship is determined according to the stroke line segment, the starting point coordinates and the ending point coordinates of each stroke.

[0129] In a possible implementation, the processing module 82 is specifically configured to:

[0130] Obtain at least one of the number, center position, and size of components constituting the character to be detected;

[0131] Obtaining component distribution information of the word to be detected according to at least one of the number, center position, and size of the components constituting the word to be detected;

[0132] Obtain the number of sampling points in each detection block and the total number of sampling points of the word to be detected;

[0133] According to the number of sampling points in each detection block and the total number of sampling points, the overall distribution information of the word to be detected is obtained.

[0134] In a possible implementation manner, the second determining module 84 is specifically configured to:

[0135] Determine the stroke similarity between each character to be detected and the corresponding pre-stored character according to the stroke information of each character to be detected and the stroke information of the corresponding pre-stored character;

[0136] Determine the structural similarity between each to-be-detected word and the corresponding pre-stored word according to the structural information of each to-be-detected word and the structural information of the corresponding pre-stored word;

[0137] The detection result is determined according to the stroke similarity between each character to be detected and the corresponding pre-stored character, and the structural similarity between each character to be detected and the corresponding pre-stored character.

[0138] In a possible implementation manner, the second determining module 84 is specifically configured to:

[0139] Determine the vector distance between the stroke timing vector of each character to be detected and the stroke timing vector of the corresponding pre-stored character;

[0140] Determine the number of strokes having different stroke directions between each character to be detected and the corresponding pre-stored character according to the stroke direction of at least one stroke included in each character to be detected and the stroke direction of at least one stroke included in the corresponding pre-stored character;

[0141] Determining the number of different relative positional relationships between each to-be-detected character and the corresponding pre-stored character according to the relative positional relationship between at least one stroke included in each to-be-detected character and the relative positional relationship between at least one stroke included in the corresponding pre-stored character;

[0142] The stroke similarity includes vector distance, the number of strokes with different stroke directions, and the number of strokes with different relative position relationships.

[0143] In a possible implementation manner, the second determining module 84 is specifically configured to:

[0144] Determine the number of inconsistent components between each word to be detected and the corresponding pre-stored word according to the component distribution information of each word to be detected and the component distribution information of the corresponding pre-stored word;

[0145] Determine the difference between each to-be-detected word and the corresponding pre-stored word according to the overall distribution information of each to-be-detected word and the overall distribution information of the corresponding pre-stored word;

[0146] The structural similarity includes the number of inconsistent components between each to-be-detected word and the corresponding pre-stored word, and the difference between each to-be-detected word and the corresponding pre-stored word.

[0147] The signature detection device 80 provided in the embodiment of the present application can execute the technical solution in the above method embodiment, and its implementation principle and beneficial effects are similar, which will not be repeated here.

[0148] Fig. 9 This is a schematic diagram of the structure of a signature detection device provided in an embodiment of the present application. Fig. 9 As shown, the signature detection device 90 may include: a memory 91 and at least one processor 92. Exemplarily, the memory 91 and at least one processor 92 are interconnected via a bus 93.

[0149] The memory 91 is used to store program instructions;

[0150] At least one processor 92 is used to execute the program instructions stored in the memory, so as to enable the signature detection device 90 to perform the signature detection method in the above method embodiment.

[0151] Optionally, the processor may be a central processing unit (CPU), a graphics processing unit (GPU), other general-purpose processors, a digital signal processor (DSP), or an application-specific integrated circuit (ASIC). A general-purpose processor may be a microprocessor or any conventional processor. The steps of the method disclosed in the embodiments of the present application may be directly implemented as being executed by a hardware processor, or may be implemented by a combination of hardware and software modules in the processor.

[0152] The signature detection device 90 provided in the embodiment of the present application can execute the technical solution in the above method embodiment, and its implementation principle and beneficial effects are similar, which will not be repeated here.

[0153] An embodiment of the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores computer-executable instructions, and when the computer-executable instructions are executed by a processor, they are used to implement the signature detection method involved in the above method embodiment.

[0154] An embodiment of the present application provides a computer program product, including a computer program, which, when executed by a processor, implements the signature detection method involved in the above method embodiment.

[0155] An embodiment of the present application provides a chip having a computer program stored thereon. When the computer program is executed by the chip, the signature detection method in the above method embodiment is implemented.

[0156] An embodiment of the present application provides a chip module, on which a computer program is stored. When the computer program is executed by the chip module, the signature detection method in the above method embodiment is implemented.

[0157] It should be noted that, for the aforementioned method embodiments, for the sake of simplicity, they are all expressed as a series of action combinations, but those skilled in the art should be aware that the present application is not limited by the described order of actions, because according to the present application, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in the specification are all optional embodiments, and the actions and modules involved are not necessarily required by the present application.

[0158] It should be further noted that, although the various steps in the flowchart are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence in the order indicated by the arrows. Unless there is a clear description in this article, the execution of these steps is not strictly limited in order, and these steps can be executed in other orders. Moreover, at least a portion of the steps in the flowchart may include multiple sub-steps or multiple stages, and these sub-steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these sub-steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a portion of the sub-steps or stages of other steps.

[0159] It should be understood that the above-mentioned device embodiments are only illustrative, and the device of the present application can also be implemented in other ways. For example, the division of units / modules in the above-mentioned embodiments is only a logical function division, and there may be other division methods in actual implementation. For example, multiple units, modules or components can be combined, or can be integrated into another system, or some features can be ignored or not executed.

[0160] In addition, unless otherwise specified, each functional unit / module in each embodiment of the present application may be integrated into one unit / module, each unit / module may exist physically separately, or two or more units / modules may be integrated together. The above-mentioned integrated unit / module may be implemented in the form of hardware or in the form of a software program module.

[0161] If the integrated unit / module is implemented in the form of hardware, the hardware may be a digital circuit, an analog circuit, etc. The physical implementation of the hardware structure includes but is not limited to transistors, memristors, etc. If not specifically stated, the processor may be any appropriate hardware processor, such as a CPU, a GPU, a Field Programmable Gate Array (FPGA), a DSP, and an ASIC, etc. If not specifically stated, the storage unit may be any appropriate magnetic storage medium or magneto-optical storage medium, such as a resistive random access memory RRAM (Resistive Random Access Memory), a dynamic random access memory DRAM (Dynamic Random Access Memory), a static random access memory SRAM (Static Random-Access Memory), an enhanced dynamic random access memory EDRAM (Enhanced Dynamic Random Access Memory), a high-bandwidth memory HBM (High-Bandwidth Memory), a hybrid memory cube HMC (Hybrid Memory Cube), etc.

[0162] If the integrated unit / module is implemented in the form of a software program module and sold or used as an independent product, it can be stored in a computer-readable memory. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art or all or part of the technical solution, can be embodied in the form of a software product, which is stored in a memory and includes several instructions for a computer device (which can be a personal computer, a server or a network device, etc.) to perform all or part of the steps of the various embodiments of the present application. The aforementioned memory includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, disk or optical disk and other media that can store program codes.

[0163] In the above embodiments, the description of each embodiment has its own emphasis. For the part not described in detail in a certain embodiment, please refer to the relevant description of other embodiments. The technical features of the above embodiments can be combined arbitrarily. In order to make the description concise, all possible combinations of the technical features in the above embodiments are not described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0164] Those skilled in the art will readily appreciate other embodiments of the present application after considering the specification and practicing the invention disclosed herein. The present application is intended to cover any modification, use or adaptation of the present application, which follows the general principles of the present application and includes common knowledge or customary techniques in the art that are not disclosed in the present application. The specification and examples are intended to be exemplary only, and the true scope and spirit of the present application are indicated by the following claims.

[0165] It should be understood that the present application is not limited to the precise structures that have been described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof. The scope of the present application is limited only by the appended claims.

Claims

1. A signature detection method, characterized in that: The method comprises: Acquire a signature to be detected of a target object, wherein the signature to be detected includes at least one word to be detected; Performing detection processing on each of the characters to be detected respectively to obtain stroke information and structure information of each of the characters to be detected; Determine a pre-stored signature of the target object in a preset signature library, wherein the pre-stored signature includes pre-stored words corresponding to each of the words to be detected; Based on the stroke information of each character to be detected, the structural information of each character to be detected, the stroke information of the corresponding pre-stored characters and the structural information of the corresponding pre-stored characters, the detection result of the signature to be detected is determined, and the detection result indicates that the signature to be detected is a real signature or a non-real signature of the target object.

2. The method according to claim 1, characterized in that For each character to be detected, the detection process is performed on each character to be detected to obtain stroke information and structure information of each character to be detected, including: Performing timing detection processing and coordinate detection processing on at least one stroke respectively to obtain writing timing and coordinate information of each stroke, wherein the at least one stroke is a stroke included in the character to be detected; According to the writing time sequence and coordinate information of each stroke, the stroke information of the character to be detected is obtained, wherein the stroke information of the character to be detected includes the stroke time sequence vector of the character to be detected, the stroke direction of each stroke, and the relative position relationship between each stroke; Performing structure detection processing on the word to be detected to determine components constituting the word to be detected and at least one detection block corresponding to the word to be detected; According to the components constituting the word to be detected and at least one detection block corresponding to the word to be detected, structural information of the word to be detected is obtained, and the structural information includes component distribution information and overall distribution information of the word to be detected.

3. The method according to claim 2, characterized in that The step of obtaining the stroke information of the character to be detected according to the writing timing and coordinate information of each stroke comprises: Generating a stroke timing vector of the character to be detected according to the writing timing of each stroke and the type of each stroke; Determining the stroke segment of each stroke according to the starting point coordinates and the ending point coordinates of each stroke; Determining the angle between the stroke line segment of each stroke and the preset direction as the stroke direction of each stroke; The relative position relationship is determined according to the stroke line segment, the starting point coordinates and the ending point coordinates of each stroke.

4. The method according to claim 2, characterized in that: The obtaining structural information of the word to be detected according to the components constituting the word to be detected and at least one detection block corresponding to the word to be detected includes: Obtain at least one of the number, center position, and size of components constituting the character to be detected; Obtaining component distribution information of the word to be detected according to at least one of the number, center position, and size of the components constituting the word to be detected; Obtaining the number of sampling points in each of the detection blocks and the total number of sampling points of the word to be detected; According to the number of sampling points in each detection block and the total number of sampling points, the overall distribution information of the word to be detected is obtained.

5. The method according to any one of claims 2 to 4, characterized in that: Determining the detection result of the signature to be detected according to the stroke information of each of the characters to be detected, the structural information of each of the characters to be detected, the stroke information of the corresponding pre-stored characters, and the structural information of the corresponding pre-stored characters, comprises: Determining the stroke similarity between each of the characters to be detected and the corresponding pre-stored characters according to the stroke information of each of the characters to be detected and the stroke information of the corresponding pre-stored characters; Determining the structural similarity between each of the to-be-detected words and the corresponding pre-stored words according to the structural information of each of the to-be-detected words and the structural information of the corresponding pre-stored words; The detection result is determined according to the stroke similarity between each of the characters to be detected and the corresponding pre-stored characters, and the structural similarity between each of the characters to be detected and the corresponding pre-stored characters.

6. The method according to claim 5, characterized in that Determining the stroke similarity between each of the characters to be detected and the corresponding pre-stored characters according to the stroke information of each of the characters to be detected and the stroke information of the corresponding pre-stored characters includes: Determine the vector distance between the stroke timing vector of each of the characters to be detected and the stroke timing vector of the corresponding pre-stored character; Determining the number of strokes having different stroke directions between each of the characters to be detected and the corresponding pre-stored characters according to the stroke direction of at least one stroke included in each of the characters to be detected and the stroke direction of at least one stroke included in the corresponding pre-stored characters; Determining the number of different relative positional relationships between each of the characters to be detected and the corresponding pre-stored characters according to the relative positional relationship between at least one stroke included in each of the characters to be detected and the relative positional relationship between at least one stroke included in the corresponding pre-stored characters; The stroke similarity includes the vector distance, the number of strokes with different stroke directions, and the number of strokes with different relative position relationships.

7. The method according to claim 5, characterized in that Determining the structural similarity between each of the to-be-detected words and the corresponding pre-stored words according to the structural information of each of the to-be-detected words and the structural information of the corresponding pre-stored words includes: Determining the number of inconsistent components between each of the to-be-detected words and the corresponding pre-stored words according to the component distribution information of each of the to-be-detected words and the component distribution information of the corresponding pre-stored words; Determining the difference between each of the to-be-detected words and the corresponding pre-stored words according to the overall distribution information of each of the to-be-detected words and the overall distribution information of the corresponding pre-stored words; The structural similarity includes the number of inconsistent components between each of the to-be-detected words and the corresponding pre-stored words, and the difference between each of the to-be-detected words and the corresponding pre-stored words.

8. A signature detection device, characterized in that: The device comprises: An acquisition module, used for acquiring a signature to be detected of a target object, wherein the signature to be detected includes at least one word to be detected; A processing module, used for performing detection processing on each of the characters to be detected to obtain stroke information and structure information of each of the characters to be detected; A first determination module is used to determine a pre-stored signature of the target object in a preset signature library, wherein the pre-stored signature includes pre-stored words corresponding to each of the words to be detected; The second determination module is used to determine the detection result of the signature to be detected based on the stroke information of each character to be detected, the structural information of each character to be detected, the stroke information of the corresponding pre-stored characters, and the structural information of the corresponding pre-stored characters, and the detection result indicates that the signature to be detected is a real signature or a non-real signature of the target object.

9. A signature detection device, characterized in that: include: at least one processor; as well as a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the signature detection method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-executable instructions, which are used to implement the signature detection method according to any one of claims 1 to 7 when executed by a processor.

11. A computer program product, characterized in that The invention comprises a computer program, which implements the signature detection method according to any one of claims 1 to 7 when executed by a processor.

Images