阅读说明：本技术 一种基于多识别参量的网上银行服务方法 (Online banking service method based on multiple identification parameters ) 是由 程玉 丁扬 杨成林 陈骁 邓日晓 于 2020-06-30 设计创作，主要内容包括：本发明涉及一种基于多识别参量的网上银行服务方法,包括：对待识别证件进行预扫描；根据预扫描图像选取对应的扫描分辨率；对图像文件中的文字信息进行识别并生成文本信息；判定文本信息中文本所属语种；判定文本信息所属证件的种类；输出模块将文本信息和判定结果。本发明先使用文本分析模块对待识别证件的文本信息进行识别和比对,判定文本信息所属语种,再通过证件分析模块对文本信息中的特征进行提取,从而完成对待识别证件的判定,最后通过信息输出模块将文本信息和判定过程交由人工审核,能够有效提高对不同种类证件进行分类解析的速率,从而使所述方法能够对不同类型的证件进行高效识别和判定。(The invention relates to an online banking service method based on multiple identification parameters, which comprises the following steps: pre-scanning the certificate to be identified; selecting corresponding scanning resolution according to the pre-scanning image; identifying character information in the image file and generating text information; judging the language of the text in the text information; judging the type of the certificate to which the text information belongs; and the output module is used for outputting the text information and the judgment result. The method comprises the steps of firstly identifying and comparing the text information of the certificate to be identified by using the text analysis module, judging the language of the text information, then extracting the characteristics in the text information by using the certificate analysis module so as to complete the judgment of the certificate to be identified, and finally handing the text information and the judgment process to manual examination and verification by using the information output module, so that the speed of classifying and analyzing different types of certificates can be effectively improved, and the method can efficiently identify and judge different types of certificates.)

1. An online banking service method based on multiple identification parameters is characterized by comprising the following steps:

step 1: pre-scanning a certificate to be identified by using an image collector to generate a pre-scanned image;

step 2: the method comprises the steps that a pre-scanned image is conveyed to a resolution adjusting module, the resolution adjusting module identifies characters in the pre-scanned image and judges the character size D of the smallest character in the pre-scanned image, after judgment is finished, the resolution adjusting module calls a preset character size matrix D0 from a storage module, D and parameters in D0 are compared, and an image collector is controlled according to a comparison result to select the corresponding resolution to scan a certificate to be identified;

and step 3: the scanned image file is conveyed to an OCR recognition module, the OCR recognition module recognizes character information in the image file, and text information with the same layout as that in the picture information is generated after recognition is completed;

And 4, step 4: the text information is transmitted to a text analysis module, the text module calls a preset language text matrix group L0 from a storage module, after the calling is completed, the text recognition module compares characters in the text information with texts in various language matrices in an L0 matrix group, and judges the language to which the text in the text information output by the OCR recognition module belongs according to a comparison result;

and 5: after the analysis is finished, the text analysis module transmits the judgment result and the text information to the certificate analysis module, the certificate analysis module calls a preset judgment characteristic matrix group R0 from the storage module, after the calling is finished, the certificate analysis module inspects the text information according to the specified characteristic points in the corresponding languages in the R0 matrix group, and finishes the judgment of the certificate type of the text information according to the type and the number of the characteristic points when the inspection is finished;

step 6: when the certificate analysis module finishes the judgment of the certificate to which the text information belongs, the text information and the judgment result are transmitted to the information output module, and the information output module outputs the text information and the judgment result to the system so that workers can manually check the text information and the judgment result.

2. The internet banking service method based on multiple identification parameters of claim 1, wherein the storage module is provided with a preset text size matrix D0, and the resolution adjustment module is provided with a preset resolution matrix I0; for the preset character size matrixes D0, D0(D1, D2, D3, D4, D5), wherein D1 is a first preset character size, D2 is a second preset character size, D3 is a third preset character size, D4 is a fourth preset character size, and D5 is a fifth preset character size, the size values of the preset character sizes are gradually reduced in sequence; for the preset resolution matrixes I0, I0(I1, I2, I3, I4, I5), wherein I1 is a first preset resolution, I2 is a second preset resolution, I3 is a third preset resolution, I4 is a fourth preset resolution, I5 is a fifth preset resolution, and the numerical values of the preset resolutions are gradually increased in sequence;

when the system identifies the certificate to be identified, the image collector pre-scans the certificate to be identified by using I1 resolution, and conveys the pre-scanned image which is pre-scanned to the resolution adjusting module, the resolution adjusting module can identify the pre-scanned image and extract character information from the pre-scanned image, after the extraction is completed, the resolution adjusting module sequentially measures the size of each character and selects the minimum character size D from the pre-scanned image, and compares the D with each numerical value in a D0 matrix:

When D > D1, the resolution adjustment module adjusts the scan resolution of the image recognition module to I1;

when D1 is more than or equal to D and more than D2, the resolution adjusting module adjusts the scanning resolution of the image identification module to I2;

when D2 is more than or equal to D and more than D3, the resolution adjusting module adjusts the scanning resolution of the image identification module to I3;

when D3 is more than or equal to D and more than D4, the resolution adjusting module adjusts the scanning resolution of the image identification module to I4;

when D4 is more than or equal to D and more than D5, the resolution adjusting module adjusts the scanning resolution of the image identification module to I5;

when the resolution adjusting module finishes adjusting the scanning resolution of the image recognition module, the image recognition module scans the certificate to be recognized and transmits the scanned image file to the OCR recognition module when the scanning is finished.

3. The internet banking service method based on multiple identification parameters of claim 2, wherein the storage module further comprises a preset language text matrix set L0, L0(L1, L2, L3, L4), wherein L1 is a first preset language text matrix, L2 is a second preset language text matrix, L3 is a third preset language text matrix, and L4 is a fourth preset language text matrix; when the text recognition module recognizes the text information transmitted by the OCR recognition module, the characters in the text information can be extracted, and the characters are sequentially compared with the preset characters in each preset language text matrix:

When the similarity between the characters in the text information and the characters in the L1 matrix is more than or equal to 80%, the language of the characters in the text information is judged as a first language by the text identification module;

when the similarity between the characters in the text information and the characters in the L2 matrix is more than or equal to 80%, the language of the characters in the text information is judged as a second language by the text identification module;

when the similarity between the characters in the text information and the characters in the L3 matrix is more than or equal to 80%, the language of the characters in the text information is judged as a third language by the text identification module;

when the similarity between the characters in the text information and the characters in the L4 matrix is more than or equal to 80%, the language of the characters in the text information is judged to be a fourth language by the text identification module;

when the similarity between the characters in the text information and the characters in the text matrixes of the preset languages is more than or equal to 80%, the text recognition module judges the language to which the text matrix of the preset language with the highest similarity to the characters in the text information belongs as the language to which the characters in the text information belong;

and after the judgment is finished, the text recognition module transmits the text information and the judgment result to the text analysis module together.

4. The internet banking service method based on multiple identification parameters, according to claim 3, wherein the storage module is further provided with preset feature matrix groups R0 and R0(R1, R2, R3 and R4), wherein R1 is a feature matrix group for determination in a first language, R2 is a feature matrix group for determination in a second language, R3 is a feature matrix group for determination in a third language, and R4 is a feature matrix group for determination in a fourth language;

When the text recognition module finishes the judgment of the language of the text information, the certificate analysis module selects a corresponding language judgment feature matrix Ri from an R0 matrix group prestored in the storage module according to the judgment result of the text recognition module to judge the type of the certificate to which the text information belongs, wherein i is 1, 2, 3 and 4.

5. The method of claim 4, wherein the i-th language decision feature matrix set Ri, Ri (Ri1, Ri2, Ri3, Ri4) is a Ri1 first-class certificate decision feature matrix, Ri2 is an i-th language second-class certificate decision feature matrix, Ri3 is an i-th language third-class certificate decision feature matrix, and Ri4 is an i-th language fourth-class certificate decision feature matrix; for j-th certificate determination feature matrixes Rij and Rij (Rij1, Rij2, Rij3.. Rijn) of the ith language, wherein Rij1 is a first determination feature of j-th certificate of the ith language, Rij2 is a second determination feature of j-th certificate of the ith language, Rij3 is a third determination feature of j-th certificate of the ith language, and Rijn is an nth determination feature of j-th certificate of the ith language;

when the certificate analysis module analyzes the text information, the text information is subjected to full-text investigation, and in the investigation process, the certificate analysis module establishes a feature statistical matrix N (N1, N2, N3 and N4), wherein N1 is the frequency of occurrence of the first type of certificate judgment features of the ith language in the investigation process, N2 is the frequency of occurrence of the second type of certificate judgment features of the ith language in the investigation process, N3 is the frequency of occurrence of the third type of certificate judgment features of the ith language in the investigation process, and N4 is the frequency of occurrence of the fourth type of certificate judgment features of the ith language in the investigation process; before investigation, N1 ═ N2 ═ N3 ═ N4 ═ 0; when the first certificate judgment feature appears in the checking process, N1 is 0+1 is 1; when the second type of certificate judgment features appear in the checking process, N2 is 0+1 is 1; when the third type of certificate judgment features appear in the checking process, N3 is 0+1 is 1; when the fourth certificate judgment feature appears in the checking process, N4 is 0+1 is 1; when the investigation is finished, the certificate analysis module counts the values of N1, N2, N3 and N4 in turn:

When the numerical value of N1 is maximum, the certificate analysis module judges that the certificate to which the text information belongs to a first type of certificate;

when the numerical value of N2 is maximum, the certificate analysis module judges that the certificate to which the text information belongs to a second type of certificate;

when the numerical value of N3 is maximum, the certificate analysis module judges that the certificate to which the text information belongs to a third type certificate;

when the numerical value of N4 is maximum, the certificate analysis module judges that the certificate to which the text information belongs to a fourth type certificate;

after the judgment is finished, the certificate analysis module marks the searched characteristics on the text information in the troubleshooting process and transmits the judgment result and the text information with the marks to the information output module.

6. The method as claimed in claim 5, wherein the nth determination feature Rijn of the jth document in the ith language comprises a single text feature, a single word feature, a single sentence feature and a single number feature.

7. The internet banking service method based on multiple recognition parameters in claim 1, wherein the OCR module in step 3, when recognizing the text in the image file, comprises the following steps:

step 3-1: the OCR recognition module preprocesses the received image file and extracts character information in the image file;

Step 3-2: the OCR recognition module segments and divides the extracted character information into lines according to the positions of characters in the initial image file;

step 3-3: the OCR recognition module cuts the characters to reduce the influence of character adhesion and broken strokes in the character information on the character information recognition precision;

step 3-4: the OCR recognition module recognizes the extracted character information;

step 3-5: after the recognition is finished, the OCR recognition module typesets the recognized characters according to the original image file, and generates text information with the same layout as the original image file after the typesetting is finished.

8. The internet banking service method based on multiple identification parameters as claimed in claim 7, wherein when the OCR module preprocesses the received image file, it first distinguishes foreground information and background information by using binarization, removes noise from the foreground information, and then performs tilt correction on the image information according to the tilt angle of the image file to complete preprocessing of five files.

9. The method for internet banking based on multiple identification parameters as claimed in claim 1, wherein the manner of collecting the document to be identified by the image collector comprises scanning and shooting.

Technical Field

The invention relates to the technical field of certificate identification, in particular to an online banking service method based on multiple identification parameters.

Background

When a bank handles a loan, a plurality of personal certificate materials are required to be provided, and the variety of materials is wide, so that the bank needs to examine and verify pictures and characters in a plurality of certificates when handling the loan. The manual method for inputting the certificate information has high cost, and the traditional manual operation cannot be broken through.

The existing certificates are of various types, so that the existing identification or verification devices are also numerous, and the related technical scheme in the prior art adopts a certificate or a class of certificates to apply a special identification and verification machine. Like this, some occasions need a plurality of windows to carry out the same or different certificate to the outside simultaneously when in actual use and verify, this just needs every window to place a plurality of relevant discernment and verifies the machine, and this not only makes material, resource waste, and the expense of purchasing increases, has occupied various places office position moreover, and the during operation is very inconvenient.

Disclosure of Invention

Therefore, the invention provides an online banking service method based on multiple identification parameters, which is used for solving the problem that a single system cannot be used for efficiently analyzing multiple certificates of different types in the prior art.

In order to achieve the above object, the present invention provides an internet banking service method based on multiple identification parameters, which comprises:

step 1: pre-scanning a certificate to be identified by using an image collector to generate a pre-scanned image;

step 2: the method comprises the steps that a pre-scanned image is conveyed to a resolution adjusting module, the resolution adjusting module identifies characters in the pre-scanned image and judges the character size D of the smallest character in the pre-scanned image, after judgment is finished, the resolution adjusting module calls a preset character size matrix D0 from a storage module, D and parameters in D0 are compared, and an image collector is controlled according to a comparison result to select the corresponding resolution to scan a certificate to be identified;

and step 3: the scanned image file is conveyed to an OCR recognition module, the OCR recognition module recognizes character information in the image file, and text information with the same layout as that in the picture information is generated after recognition is completed;

and 4, step 4: the text information is transmitted to a text analysis module, the text module calls a preset language text matrix group L0 from a storage module, after the calling is completed, the text recognition module compares characters in the text information with texts in various language matrices in an L0 matrix group, and judges the language to which the text in the text information output by the OCR recognition module belongs according to a comparison result;

And 5: after the analysis is finished, the text analysis module transmits the judgment result and the text information to the certificate analysis module, the certificate analysis module calls a preset judgment characteristic matrix group R0 from the storage module, after the calling is finished, the certificate analysis module inspects the text information according to the specified characteristic points in the corresponding languages in the R0 matrix group, and finishes the judgment of the certificate type of the text information according to the type and the number of the characteristic points when the inspection is finished;

step 6: when the certificate analysis module finishes the judgment of the certificate to which the text information belongs, the text information and the judgment result are transmitted to the information output module, and the information output module outputs the text information and the judgment result to the system so that workers can manually check the text information and the judgment result.

Furthermore, a preset character size matrix D0 is arranged in the storage module, and a preset resolution matrix I0 is arranged in the resolution adjusting module; for the preset character size matrixes D0, D0(D1, D2, D3, D4, D5), wherein D1 is a first preset character size, D2 is a second preset character size, D3 is a third preset character size, D4 is a fourth preset character size, and D5 is a fifth preset character size, the size values of the preset character sizes are gradually reduced in sequence; for the preset resolution matrixes I0, I0(I1, I2, I3, I4, I5), wherein I1 is a first preset resolution, I2 is a second preset resolution, I3 is a third preset resolution, I4 is a fourth preset resolution, I5 is a fifth preset resolution, and the numerical values of the preset resolutions are gradually increased in sequence;

When the system identifies the certificate to be identified, the image collector pre-scans the certificate to be identified by using I1 resolution, and conveys the pre-scanned image which is pre-scanned to the resolution adjusting module, the resolution adjusting module can identify the pre-scanned image and extract character information from the pre-scanned image, after the extraction is completed, the resolution adjusting module sequentially measures the size of each character and selects the minimum character size D from the pre-scanned image, and compares the D with each numerical value in a D0 matrix:

when D > D1, the resolution adjustment module adjusts the scan resolution of the image recognition module to I1;

when D1 is more than or equal to D and more than D2, the resolution adjusting module adjusts the scanning resolution of the image identification module to I2;

when D2 is more than or equal to D and more than D3, the resolution adjusting module adjusts the scanning resolution of the image identification module to I3;

when D3 is more than or equal to D and more than D4, the resolution adjusting module adjusts the scanning resolution of the image identification module to I4;

when D4 is more than or equal to D and more than D5, the resolution adjusting module adjusts the scanning resolution of the image identification module to I5;

when the resolution adjusting module finishes adjusting the scanning resolution of the image recognition module, the image recognition module scans the certificate to be recognized and transmits the scanned image file to the OCR recognition module when the scanning is finished.

Further, the storage module is further provided with preset language text matrix groups L0 and L0(L1, L2, L3, and L4), where L1 is a first preset language text matrix, L2 is a second preset language text matrix, L3 is a third preset language text matrix, and L4 is a fourth preset language text matrix; when the text recognition module recognizes the text information transmitted by the OCR recognition module, the characters in the text information can be extracted, and the characters are sequentially compared with the preset characters in each preset language text matrix:

when the similarity between the characters in the text information and the characters in the L1 matrix is more than or equal to 80%, the language of the characters in the text information is judged as a first language by the text identification module;

when the similarity between the characters in the text information and the characters in the L2 matrix is more than or equal to 80%, the language of the characters in the text information is judged as a second language by the text identification module;

when the similarity between the characters in the text information and the characters in the L3 matrix is more than or equal to 80%, the language of the characters in the text information is judged as a third language by the text identification module;

when the similarity between the characters in the text information and the characters in the L4 matrix is more than or equal to 80%, the language of the characters in the text information is judged to be a fourth language by the text identification module;

When the similarity between the characters in the text information and the characters in the text matrixes of the preset languages is more than or equal to 80%, the text recognition module judges the language to which the text matrix of the preset language with the highest similarity to the characters in the text information belongs as the language to which the characters in the text information belong;

and after the judgment is finished, the text recognition module transmits the text information and the judgment result to the text analysis module together.

Further, the storage module is further provided with preset judgment feature matrix groups R0 and R0(R1, R2, R3 and R4), wherein R1 is a first language judgment feature matrix group, R2 is a second language judgment feature matrix group, R3 is a third language judgment feature matrix group, and R4 is a fourth language judgment feature matrix group;

when the text recognition module finishes the judgment of the language of the text information, the certificate analysis module selects a corresponding language judgment feature matrix Ri from an R0 matrix group prestored in the storage module according to the judgment result of the text recognition module to judge the type of the certificate to which the text information belongs, wherein i is 1, 2, 3 and 4.

Further, for the ith language judgment feature matrix group Ri, Ri (Ri1, Ri2, Ri3, Ri4), where Ri1 is the ith language first-class certificate judgment feature matrix, Ri2 is the ith language second-class certificate judgment feature matrix, Ri3 is the ith language third-class certificate judgment feature matrix, and Ri4 is the ith language fourth-class certificate judgment feature matrix; for j-th certificate determination feature matrixes Rij and Rij (Rij1, Rij2, Rij3.. Rijn) of the ith language, wherein Rij1 is a first determination feature of j-th certificate of the ith language, Rij2 is a second determination feature of j-th certificate of the ith language, Rij3 is a third determination feature of j-th certificate of the ith language, and Rijn is an nth determination feature of j-th certificate of the ith language;

When the certificate analysis module analyzes the text information, the text information is subjected to full-text investigation, and in the investigation process, the certificate analysis module establishes a feature statistical matrix N (N1, N2, N3 and N4), wherein N1 is the frequency of occurrence of the first type of certificate judgment features of the ith language in the investigation process, N2 is the frequency of occurrence of the second type of certificate judgment features of the ith language in the investigation process, N3 is the frequency of occurrence of the third type of certificate judgment features of the ith language in the investigation process, and N4 is the frequency of occurrence of the fourth type of certificate judgment features of the ith language in the investigation process; before investigation, N1 ═ N2 ═ N3 ═ N4 ═ 0; when the first certificate judgment feature appears in the checking process, N1 is 0+1 is 1; when the second type of certificate judgment features appear in the checking process, N2 is 0+1 is 1; when the third type of certificate judgment features appear in the checking process, N3 is 0+1 is 1; when the fourth certificate judgment feature appears in the checking process, N4 is 0+1 is 1; when the investigation is finished, the certificate analysis module counts the values of N1, N2, N3 and N4 in turn:

when the numerical value of N1 is maximum, the certificate analysis module judges that the certificate to which the text information belongs to a first type of certificate;

when the numerical value of N2 is maximum, the certificate analysis module judges that the certificate to which the text information belongs to a second type of certificate;

When the numerical value of N3 is maximum, the certificate analysis module judges that the certificate to which the text information belongs to a third type certificate;

when the numerical value of N4 is maximum, the certificate analysis module judges that the certificate to which the text information belongs to a fourth type certificate;

after the judgment is finished, the certificate analysis module marks the searched characteristics on the text information in the troubleshooting process and transmits the judgment result and the text information with the marks to the information output module.

Further, the nth judgment feature Rijn of the jth certificate of the ith language comprises a single character feature, a single word feature, a single sentence feature and a single number feature.

Further, when the OCR recognition module recognizes the text in the image file in the step 3, the method includes the following steps:

step 3-1: the OCR recognition module preprocesses the received image file and extracts character information in the image file;

step 3-2: the OCR recognition module segments and divides the extracted character information into lines according to the positions of characters in the initial image file;

step 3-3: the OCR recognition module cuts the characters to reduce the influence of character adhesion and broken strokes in the character information on the character information recognition precision;

Step 3-4: the OCR recognition module recognizes the extracted character information;

step 3-5: after the recognition is finished, the OCR recognition module typesets the recognized characters according to the original image file, and generates text information with the same layout as the original image file after the typesetting is finished.

Furthermore, when the OCR module preprocesses the received image file, the binarization method distinguishes foreground information and background information, carries out noise removal on the foreground information, and then carries out tilt correction on the image information according to the tilt angle of the image file so as to complete preprocessing of the five files.

Further, the manner in which the image collector collects the document to be identified includes scanning and photographing.

Compared with the prior art, the certificate identification system has the advantages that the text analysis module is used for identifying and comparing the text information of the certificate to be identified, the language of the text information is judged, the certificate analysis module is used for extracting the characteristics in the text information, so that the judgment of the certificate to be identified is completed, and finally the text information and the judgment process are manually checked through the information output module, so that the speed of classifying and analyzing different types of certificates can be effectively improved, and the system can efficiently identify and judge different types of certificates.

Furthermore, an OCR recognition module is further arranged in the system, characters in the image file can be rapidly and accurately recognized through the private OCR recognition module, high-precision character track/shape information is provided for subsequent language identification and certificate type judgment, and therefore the analysis efficiency of the system on different certificates is further improved.

Furthermore, the system is also provided with a storage module, the identification standard and the judgment standard required by the analysis of the certificate can be stored in the system in advance through the storage module, and when the system carries out scanning resolution adjustment, language analysis and certificate type judgment on the certificate to be identified, each appointed module can extract the corresponding identification standard or judgment standard from the storage module so as to carry out rapid and accurate analysis on the certificate, so that the analysis efficiency of the system on different types of certificates is further improved.

Furthermore, the system is also provided with a resolution adjusting module, a preset resolution matrix I0(I1, I2, I3, I4, I5) is arranged in the adjusting module, when the system analyzes a document to be identified, the image collector pre-scans an image and conveys the pre-scanned image to the resolution adjusting module when the scanning is completed, the resolution adjusting module measures the size D of the smallest character in the pre-scanned image and calls a preset character size matrix D0(D1, D2, D3, D4, D5) from the storage module, the values in the D and D0 matrices are sequentially compared, and the corresponding resolution is selected from the I0 matrix according to the comparison result and the image collector is controlled to scan the document to be identified by using the specified resolution. By using the pre-scanning, the scanning resolution used by the system during analysis is determined according to the size of the minimum character in the pre-scanned image, the definition of an image file can be effectively ensured, and the precision of a subsequent module during extraction, identification and judgment is improved, so that the certificate analysis efficiency of the system is further improved.

Furthermore, a preset language text matrix group L0(L1, L2, L3 and L4) is further arranged in the storage module, when the text recognition module recognizes the language of the text information, characters in the text information are extracted and are sequentially compared with preset characters in each matrix in the L0 matrix group, the language of the characters in the text information is judged according to a comparison result, and the system can analyze the multi-language certificate by presetting judgment standards of a plurality of language characters, so that the analysis range of the system is improved.

Furthermore, a preset judgment feature matrix group R0(R1, R2, R3, R4) is further arranged in the storage module, the certificate analysis module can select a corresponding judgment feature matrix from an R0 matrix group prestored in the storage module according to the judgment result of the text recognition module to judge the type of the certificate to which the text information belongs, and the judgment feature matrix of the specified language is used for judging the type of the certificate to which the text information of the corresponding language belongs, so that the accuracy of judging the type of the certificate can be further improved, and the analysis efficiency of the system on the certificate is further improved.

Further, for the i-th language, the feature matrix groups Ri, Ri (Ri1, Ri2, Ri3, Ri4) are determined, and for the i-th language, the j-th certificate, the feature matrices Rij, Rij (Rij1, Rij2, Rij3.. Rijn) are determined; in the process of investigation, the certificate analysis module establishes a characteristic statistical matrix N (N1, N2, N3 and N4), corresponding numerical values in the matrix are updated according to the types of the characteristics appearing in the text information, when the investigation is completed, the judgment of the types of the certificates to be recognized is completed according to the statistical result, the types of the certificates to be recognized are classified by using a plurality of preset characteristic words/sentences, the analysis of the certificates to be recognized can be completed more quickly and accurately, and the analysis efficiency of the system on the certificates is further improved.

Further, the OCR recognition module can perform inclination correction on the image, so that the selection of specific equipment of the image collector can be more diversified, and the application range of the system is widened.

Drawings

Fig. 1 is a flow chart of the multi-identification parameter-based online banking service method of the present invention.

Detailed Description

In order that the objects and advantages of the invention will be more clearly understood, the invention is further described below with reference to examples; it should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

Preferred embodiments of the present invention are described below with reference to the accompanying drawings. It should be understood by those skilled in the art that these embodiments are only for explaining the technical principle of the present invention, and do not limit the scope of the present invention.

It should be noted that in the description of the present invention, the terms of direction or positional relationship indicated by the terms "upper", "lower", "left", "right", "inner", "outer", etc. are based on the directions or positional relationships shown in the drawings, which are only for convenience of description, and do not indicate or imply that the device or element must have a specific orientation, be constructed in a specific orientation, and be operated, and thus, should not be construed as limiting the present invention.

Furthermore, it should be noted that, in the description of the present invention, unless otherwise explicitly specified or limited, the terms "mounted," "connected," and "connected" are to be construed broadly, and may be, for example, fixedly connected, detachably connected, or integrally connected; can be mechanically or electrically connected; they may be connected directly or indirectly through intervening media, or they may be interconnected between two elements. The specific meanings of the above terms in the present invention can be understood by those skilled in the art according to specific situations.

Fig. 1 is a flowchart illustrating a method for internet banking based on multiple identification parameters according to the present invention. The invention relates to an online banking service method based on multiple identification parameters, which comprises the following steps:

step 1: pre-scanning a certificate to be identified by using an image collector to generate a pre-scanned image;

step 2: the method comprises the steps that a pre-scanned image is conveyed to a resolution adjusting module, the resolution adjusting module identifies characters in the pre-scanned image and judges the character size D of the smallest character in the pre-scanned image, after judgment is finished, the resolution adjusting module calls a preset character size matrix D0 from a storage module, D and parameters in D0 are compared, and an image collector is controlled according to a comparison result to select the corresponding resolution to scan a certificate to be identified;

And step 3: the scanned image file is conveyed to an OCR recognition module, the OCR recognition module recognizes character information in the image file, and text information with the same layout as that in the picture information is generated after recognition is completed;

and 4, step 4: the text information is transmitted to a text analysis module, the text module calls a preset language text matrix group L0 from a storage module, after the calling is completed, the text recognition module compares characters in the text information with texts in various language matrices in an L0 matrix group, and judges the language to which the text in the text information output by the OCR recognition module belongs according to a comparison result;

and 5: after the analysis is finished, the text analysis module transmits the judgment result and the text information to the certificate analysis module, the certificate analysis module calls a preset judgment characteristic matrix group R0 from the storage module, after the calling is finished, the certificate analysis module inspects the text information according to the specified characteristic points in the corresponding languages in the R0 matrix group, and finishes the judgment of the certificate type of the text information according to the type and the number of the characteristic points when the inspection is finished;

step 6: when the certificate analysis module finishes the judgment of the certificate to which the text information belongs, the text information and the judgment result are transmitted to the information output module, and the information output module outputs the text information and the judgment result to the system so that workers can manually check the text information and the judgment result.

Specifically, a preset character size matrix D0 is arranged in the storage module, and a preset resolution matrix I0 is arranged in the resolution adjustment module; for the preset character size matrixes D0, D0(D1, D2, D3, D4, D5), wherein D1 is a first preset character size, D2 is a second preset character size, D3 is a third preset character size, D4 is a fourth preset character size, and D5 is a fifth preset character size, the size values of the preset character sizes are gradually reduced in sequence; for the preset resolution matrixes I0, I0(I1, I2, I3, I4, I5), wherein I1 is a first preset resolution, I2 is a second preset resolution, I3 is a third preset resolution, I4 is a fourth preset resolution, I5 is a fifth preset resolution, and the numerical values of the preset resolutions are gradually increased in sequence;

when the system identifies the certificate to be identified, the image collector pre-scans the certificate to be identified by using I1 resolution, and conveys the pre-scanned image which is pre-scanned to the resolution adjusting module, the resolution adjusting module can identify the pre-scanned image and extract character information from the pre-scanned image, after the extraction is completed, the resolution adjusting module sequentially measures the size of each character and selects the minimum character size D from the pre-scanned image, and compares the D with each numerical value in a D0 matrix:

When D > D1, the resolution adjustment module adjusts the scan resolution of the image recognition module to I1;

when D1 is more than or equal to D and more than D2, the resolution adjusting module adjusts the scanning resolution of the image identification module to I2;

when D2 is more than or equal to D and more than D3, the resolution adjusting module adjusts the scanning resolution of the image identification module to I3;

when D3 is more than or equal to D and more than D4, the resolution adjusting module adjusts the scanning resolution of the image identification module to I4;

when D4 is more than or equal to D and more than D5, the resolution adjusting module adjusts the scanning resolution of the image identification module to I5;

when the resolution adjusting module finishes adjusting the scanning resolution of the image recognition module, the image recognition module scans the certificate to be recognized and transmits the scanned image file to the OCR recognition module when the scanning is finished.

Specifically, the storage module is further provided with preset language text matrix groups L0 and L0(L1, L2, L3, and L4), where L1 is a first preset language text matrix, L2 is a second preset language text matrix, L3 is a third preset language text matrix, and L4 is a fourth preset language text matrix; when the text recognition module recognizes the text information transmitted by the OCR recognition module, the characters in the text information can be extracted, and the characters are sequentially compared with the preset characters in each preset language text matrix:

When the similarity between the characters in the text information and the characters in the L1 matrix is more than or equal to 80%, the language of the characters in the text information is judged as a first language by the text identification module;

when the similarity between the characters in the text information and the characters in the L2 matrix is more than or equal to 80%, the language of the characters in the text information is judged as a second language by the text identification module;

when the similarity between the characters in the text information and the characters in the L3 matrix is more than or equal to 80%, the language of the characters in the text information is judged as a third language by the text identification module;

when the similarity between the characters in the text information and the characters in the L4 matrix is more than or equal to 80%, the language of the characters in the text information is judged to be a fourth language by the text identification module;

when the similarity between the characters in the text information and the characters in the text matrixes of the preset languages is more than or equal to 80%, the text recognition module judges the language to which the text matrix of the preset language with the highest similarity to the characters in the text information belongs as the language to which the characters in the text information belong;

and after the judgment is finished, the text recognition module transmits the text information and the judgment result to the text analysis module together.

Specifically, the storage module is further provided with preset judgment feature matrix groups R0 and R0(R1, R2, R3 and R4), wherein R1 is a first language judgment feature matrix group, R2 is a second language judgment feature matrix group, R3 is a third language judgment feature matrix group, and R4 is a fourth language judgment feature matrix group;

When the text recognition module finishes the judgment of the language of the text information, the certificate analysis module selects a corresponding language judgment feature matrix Ri from an R0 matrix group prestored in the storage module according to the judgment result of the text recognition module to judge the type of the certificate to which the text information belongs, wherein i is 1, 2, 3 and 4.

Specifically, for the ith language judgment feature matrix group Ri, Ri (Ri1, Ri2, Ri3, Ri4), where Ri1 is the ith language first-class certificate judgment feature matrix, Ri2 is the ith language second-class certificate judgment feature matrix, Ri3 is the ith language third-class certificate judgment feature matrix, and Ri4 is the ith language fourth-class certificate judgment feature matrix; for j-th certificate determination feature matrixes Rij and Rij (Rij1, Rij2, Rij3.. Rijn) of the ith language, wherein Rij1 is a first determination feature of j-th certificate of the ith language, Rij2 is a second determination feature of j-th certificate of the ith language, Rij3 is a third determination feature of j-th certificate of the ith language, and Rijn is an nth determination feature of j-th certificate of the ith language;

when the certificate analysis module analyzes the text information, the text information is subjected to full-text investigation, and in the investigation process, the certificate analysis module establishes a feature statistical matrix N (N1, N2, N3 and N4), wherein N1 is the frequency of occurrence of the first type of certificate judgment features of the ith language in the investigation process, N2 is the frequency of occurrence of the second type of certificate judgment features of the ith language in the investigation process, N3 is the frequency of occurrence of the third type of certificate judgment features of the ith language in the investigation process, and N4 is the frequency of occurrence of the fourth type of certificate judgment features of the ith language in the investigation process; before investigation, N1 ═ N2 ═ N3 ═ N4 ═ 0; when the first certificate judgment feature appears in the checking process, N1 is 0+1 is 1; when the second type of certificate judgment features appear in the checking process, N2 is 0+1 is 1; when the third type of certificate judgment features appear in the checking process, N3 is 0+1 is 1; when the fourth certificate judgment feature appears in the checking process, N4 is 0+1 is 1; when the investigation is finished, the certificate analysis module counts the values of N1, N2, N3 and N4 in turn:

When the numerical value of N1 is maximum, the certificate analysis module judges that the certificate to which the text information belongs to a first type of certificate;

when the numerical value of N2 is maximum, the certificate analysis module judges that the certificate to which the text information belongs to a second type of certificate;

when the numerical value of N3 is maximum, the certificate analysis module judges that the certificate to which the text information belongs to a third type certificate;

when the numerical value of N4 is maximum, the certificate analysis module judges that the certificate to which the text information belongs to a fourth type certificate;

after the judgment is finished, the certificate analysis module marks the searched characteristics on the text information in the troubleshooting process and transmits the judgment result and the text information with the marks to the information output module.

Specifically, each nth judgment feature Rijn of the jth certificate of the ith language comprises a single character feature, a single word feature, a single sentence feature and a single number feature.

Specifically, when the OCR recognition module recognizes the text in the image file in the step 3, the method includes the following steps:

step 3-1: the OCR recognition module preprocesses the received image file and extracts character information in the image file;

step 3-2: the OCR recognition module segments and divides the extracted character information into lines according to the positions of characters in the initial image file;

Step 3-3: the OCR recognition module cuts the characters to reduce the influence of character adhesion and broken strokes in the character information on the character information recognition precision;

step 3-4: the OCR recognition module recognizes the extracted character information;

step 3-5: after the recognition is finished, the OCR recognition module typesets the recognized characters according to the original image file, and generates text information with the same layout as the original image file after the typesetting is finished.

Specifically, when the OCR recognition module preprocesses the received image file, the foreground information and the background information are distinguished by using binarization, noise removal is performed on the foreground information, and then the image information is subjected to tilt correction according to the tilt angle of the image file so as to complete preprocessing of the five files.

In particular, the manner in which the image collector collects the document to be identified includes scanning and imaging.

So far, the technical solutions of the present invention have been described in connection with the preferred embodiments shown in the drawings, but it is easily understood by those skilled in the art that the scope of the present invention is obviously not limited to these specific embodiments. Equivalent changes or substitutions of related technical features can be made by those skilled in the art without departing from the principle of the invention, and the technical scheme after the changes or substitutions can fall into the protection scope of the invention.

The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention; various modifications and alterations to this invention will become apparent to those skilled in the art. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.