Scanning device and image automatic processing method for intelligent archive management system
By establishing the position correlation between the matrix and the rectangular graph, the three-dimensional distance diagram of the archive file is calculated, which solves the problem of inaccurate sorting of archive files in the existing technology, and realizes efficient collection and convenient reference of archive files.
Patent Information
- Application Number
- CN202411083042.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-08
- Publication Date
- 2024-11-15
- Estimated Expiration
- 2044-08-08
AI Technical Summary
In the existing smart archive management system, the scanning and collection methods of archive files rely on a small number of key information, resulting in the inability to accurately identify and collect archive files, resulting in confusing order and inconvenient review.
By obtaining scanned images of multiple archive files, extracting identification information and forming an
It realizes accurate sorting and efficient collection of archive files, simplifies the subsequent review process, and improves the efficiency of archive management.
Smart Images

Figure CN118629064B_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the technical field of archive image processing, and in particular to a scanning device and an automatic image processing method for a smart archive management system. Background Art
[0002] The smart archive management system is an information-based archive management system that applies smart management technology. It is the development and improvement of the digital archive management system. It is a new type of information-based archive management system based on smart management technology. On the basis of the existing functions of the digital archive management system, smart management technology should be fully applied to make its functions have the characteristics of intelligence and intelligence, and innovate the archive business management model and archive information resource development method. The construction of a smart archive management system will help to manage archives more scientifically, develop archive information more accurately, improve business efficiency more comprehensively, and maintain archive security more effectively. The core should be to promote the close integration of information technology and archive management system business, with the coordinated development of archive management system informatization and smart city construction, and the goal of improving the archive management system's ability to serve the economy and society. We should do a good job in top-level design and carry out the functional design and organization of the smart archive management system.
[0003] In the prior art, the smart archive management system is based on the scanning and uploading of archive files, so the ability to scan and collect archive files is the basis for effective management of archive files; the traditional scanning and collection method of archive files is to scan the archive number, name and other related information on the archive file to identify and mark the archive file, and thereby realize the orderly uploading and archiving of the archive file. However, this conventional archival file identification method often causes the archive file to be unable to be accurately identified due to the incompleteness and unclearness of a few key information such as the archive number of the archive file, thereby causing problems such as disordered order of the archive files after they are collected, which causes inconvenience in subsequent reference. Therefore, it is necessary to propose a scanning device and image automation processing method for a smart archive management system to solve the above problems. Summary of the invention
[0004] The present invention provides an automatic image processing method for a smart archive management system, the method comprising the following steps:
[0005] S1, obtaining scanned images of multiple archive files, extracting identification information and corresponding identification data in each of the scanned images, and forming an array of <archive, identification, value>;
[0006] S2, establishing a first matrix according to the <file, identifier, value> array obtained by scanning the preset archive file, and establishing a first rectangular diagram corresponding to the first matrix, wherein the first matrix and the first rectangular diagram are marked with the identifier information of the preset archive file, and are used to lay out the positions of all the identifier information, and store the corresponding values according to the positions of the identifier information;
[0007] S3, establishing a t-th matrix and a t-th rectangular diagram of other archive files except the preset archive file based on the first matrix and the first rectangular diagram;
[0008] S4, respectively calculating the distance D between the t-th matrix and the first matrix, establishing a position association relationship between the first rectangular diagram and the t-th rectangular diagram based on the distance D, and forming a three-dimensional distance diagram; the three-dimensional distance diagram uses the plane where the first rectangular diagram is located as the first and second dimensions, and uses the vertical distance between the t-th rectangular diagram and the first rectangular diagram as the third dimension;
[0009] S5, arranging the order of the plurality of archive files according to the three-dimensional distance map.
[0010] As a preferred implementation, the distance between the t-th matrix and the first matrix is calculated respectively. , specifically including:
[0011] ,
[0012] Among them, m is the vector coefficient, when When m=1; when < When m=-1; represents the value of the i-th marker of the t-th matrix, Represents the first matrix The value of the ith identifier of represents the weight of the ith identifier of the tth matrix, Represents the tth matrix and the first matrix distance.
[0013] As a preferred embodiment, it also includes:
[0014] The first matrix All values and the t-th matrix All values of are normalized once, and then the first matrix The value of the ith identifier and all the tth matrices The value of the ith identifier is subjected to secondary normalization;
[0015] Setting the initial weight of the t-th matrix weight, and obtaining the values corresponding to all the identifiers of the t-th matrix after secondary normalization processing;
[0016] The values of the tth matrix are calibrated according to the values corresponding to all the identifiers of the tth matrix, and the initial weights are adjusted according to the calibration results.
[0017] As a preferred implementation, the adjusting the initial weight according to the calibration result further includes:
[0018] If the error between the value of the i-th identifier of the t-th matrix and the corresponding values of one or more associated identifiers of the t-th matrix is greater than a first threshold, reducing the initial weight;
[0019] If the error between the value of the i-th identifier of the t-th matrix and the corresponding values of one or more associated identifiers of the t-th matrix is less than or equal to a first threshold, increasing the initial weight;
[0020] If the value of the i-th identifier of the t-th matrix and the corresponding values of one or more associated identifiers are mutually exclusive data, the initial weight of the corresponding value of the identifier with a higher error rate among the corresponding values of all associated identifiers of the t-th matrix is set to 0.
[0021] As a preferred implementation, the initial weight is reduced by a weight reduction formula, and the weight reduction formula is:
[0022] ,
[0023] Among them, α and β represent the preset parameters in the weight reduction formula. represents the initial weight reduction multiple of the value corresponding to the identifier to be reduced, and the reduction multiple is positively correlated with the ratio of the error to the first threshold value, It represents the weight after the initial weight of the value corresponding to the i-th identifier of the t-th matrix is reduced. It represents the initial weight of the value corresponding to the mark to be reduced, and e represents a natural constant.
[0024] As a preferred implementation, the initial weight is increased by a weight increasing formula, and the weight increasing formula is:
[0025] ,
[0026] in, represents the sum of weights corresponding to all identifiers of the first matrix before the increase, a, b, c represent the preset parameters in the weight increase formula, represents the initial weight increase multiple of the value corresponding to the identifier that needs to be increased, and the increase multiple is negatively correlated with the ratio of the error to the first threshold value, It represents the weight after the initial weight of the value corresponding to the i-th identifier of the t-th matrix is increased, and e represents a natural constant.
[0027] As a preferred implementation, the initial weight of the corresponding value of the identifier with a higher error rate among the corresponding values of all associated identifiers of the t-th matrix is set to 0, specifically including:
[0028] Obtain the corresponding values of all associated identifiers of the t-th matrix, and perform positive and negative voting according to the attributes of all the values; the corresponding identifier of the attribute corresponding to the value with a lower vote rate is used as the identifier with a higher error rate;
[0029] Alternatively, the corresponding values of all associated identifiers of matrices other than the t-th matrix are obtained, and positive and negative voting is performed according to the attributes of all the values; the corresponding identifiers of the values corresponding to the attributes with lower vote rates are used as identifiers with higher error rates.
[0030] As another embodiment, the present invention provides a scanning device for a smart archive management system, the device comprising the following modules:
[0031] An information extraction module, used for obtaining scanned images of multiple archive files, extracting identification information and corresponding identification data in each of the scanned images, and forming an array of <archive, identification, value>;
[0032] An information storage module, used to establish a first matrix according to the <file, identifier, value> array obtained by scanning a preset archive file, and to establish a first rectangular diagram corresponding to the first matrix, wherein the first matrix and the first rectangular diagram are marked with the identifier information of the preset archive file, and are used to lay out the positions of all the identifier information, and store the corresponding values according to the positions of the identifier information;
[0033] An information layout module, used for establishing a t-th matrix and a t-th rectangular diagram of other archive files except the preset archive file based on the first matrix and the first rectangular diagram;
[0034] An information calculation module, used to calculate the distance D between the t-th matrix and the first matrix respectively, establish the position association relationship between the first rectangular diagram and the t-th rectangular diagram based on the distance D, and form a three-dimensional distance diagram; the three-dimensional distance diagram uses the plane where the first rectangular diagram is located as the first and second dimensions, and uses the distance between the t-th rectangular diagram and the first rectangular diagram as the third dimension;
[0035] An information sorting module is used to arrange the order of the multiple archive files according to the three-dimensional distance map.
[0036] As a preferred implementation, the distance between the t-th matrix and the first matrix is calculated respectively. , specifically including:
[0037] ,
[0038] Among them, m is the vector coefficient, when When m=1; when < When m=-1; represents the value of the i-th marker of the t-th matrix, Represents the first matrix The value of the ith identifier of represents the weight of the ith identifier of the tth matrix, Represents the tth matrix and the first matrix distance.
[0039] As a preferred embodiment, it also includes:
[0040] The first matrix All values and the t-th matrix All values of are normalized once, and then the first matrix The value of the ith identifier and all the tth matrices The value of the ith identifier is subjected to secondary normalization;
[0041] Setting the initial weight of the t-th matrix weight, and obtaining the values corresponding to all the identifiers of the t-th matrix after secondary normalization processing;
[0042] The values of the tth matrix are calibrated according to the values corresponding to all the identifiers of the tth matrix, and the initial weights are adjusted according to the calibration results.
[0043] As a preferred implementation, the adjusting the initial weight according to the calibration result further includes:
[0044] If the error between the value of the i-th identifier of the t-th matrix and the corresponding values of one or more associated identifiers of the t-th matrix is greater than a first threshold, reducing the initial weight;
[0045] If the error between the value of the i-th identifier of the t-th matrix and the corresponding values of one or more associated identifiers of the t-th matrix is less than or equal to a first threshold, increasing the initial weight;
[0046] If the value of the i-th identifier of the t-th matrix and the corresponding values of one or more associated identifiers are mutually exclusive data, the initial weight of the corresponding value of the identifier with a higher error rate among the corresponding values of all associated identifiers of the t-th matrix is set to 0.
[0047] As a preferred implementation, the initial weight is reduced by a weight reduction formula, and the weight reduction formula is:
[0048] ,
[0049] Among them, α and β represent the preset parameters in the weight reduction formula. represents the initial weight reduction multiple of the value corresponding to the identifier to be reduced, and the reduction multiple is positively correlated with the ratio of the error to the first threshold value, It represents the weight after the initial weight of the value corresponding to the i-th identifier of the t-th matrix is reduced. It represents the initial weight of the value corresponding to the mark to be reduced, and e represents a natural constant.
[0050] As a preferred implementation, the initial weight is increased by a weight increasing formula, and the weight increasing formula is:
[0051] ,
[0052] in, represents the sum of weights corresponding to all identifiers of the first matrix before the increase, a, b, c represent the preset parameters in the weight increase formula, represents the initial weight increase multiple of the value corresponding to the identifier that needs to be increased, and the increase multiple is negatively correlated with the ratio of the error to the first threshold value, It represents the weight after the initial weight of the value corresponding to the i-th identifier of the t-th matrix is increased, and e represents a natural constant.
[0053] As a preferred implementation, the initial weight of the corresponding value of the identifier with a higher error rate among the corresponding values of all associated identifiers of the t-th matrix is set to 0, specifically including:
[0054] Obtain the corresponding values of all associated identifiers of the t-th matrix, and perform positive and negative voting according to the attributes of all the values; the corresponding identifier of the attribute corresponding to the value with a lower vote rate is used as the identifier with a higher error rate;
[0055] Alternatively, the corresponding values of all associated identifiers of matrices other than the t-th matrix are obtained, and positive and negative voting is performed according to the attributes of all the values; the corresponding identifiers of the values corresponding to the attributes with lower vote rates are used as identifiers with higher error rates.
[0056] As another embodiment, the present invention provides a scanning device for a smart archive management system, wherein the system executes the automatic image processing method.
[0057] As another embodiment, the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and the computer program executes the automatic image processing method.
[0058] It can be seen that the scanning device and image automatic processing method for a smart archive management system of the present invention establishes a matrix and a rectangular diagram of each archive file according to an array including identification information obtained from scanning the archive file, so as to layout the position of the identification information and its corresponding value, thereby storing multiple key information of the archive file according to the layout position; then, according to the distance between the matrices corresponding to different archive files, a position association relationship between the rectangular diagrams corresponding to different archive files is established, thereby forming a three-dimensional distance diagram of all archive files; and finally, the order of multiple archive files is obtained based on the three-dimensional distance diagram. In this way, the effective sorting of archive files is realized according to the positional relationship between the rectangular diagrams composed of several key information of the archive files, thereby achieving a more accurate sorting result, which is conducive to the efficient collection of archive files and facilitates the subsequent reference to archive files. BRIEF DESCRIPTION OF THE DRAWINGS
[0059] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments and the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.
[0060] Figure 1 It is a schematic diagram of the steps of an automatic image processing method for a smart archive management system of the present invention.
[0061] Figure 2 It is a schematic structural diagram of a three-dimensional distance map of multiple archive files of the present invention. DETAILED DESCRIPTION
[0062] The embodiments of the present invention are further described below with reference to the accompanying drawings.
[0063] Embodiment 1:
[0064] like Figure 1 As shown, the present invention provides an automatic image processing method for a smart archive management system, the method comprising the following steps:
[0065] S1, obtain scanned images of multiple archive files, extract identification information and corresponding identification data in each of the scanned images, and form an array of <archive, identification, value>; it should be noted that, since the information in the archive file is multiple, for example, it may include information such as archive number, name, gender, age, learning experience, work experience, etc., in order to get rid of the problem of inaccurate sorting results caused by sorting archives based only on a few information such as archive number in the prior art, the archive information in the present invention is usually multiple, so as to minimize the error caused by a single information to the sorting result. Among them, extract the identification information and corresponding identification data in each of the scanned images, the identification information is the field name in the archive information, and the identification data is the content of the corresponding field, for example, a certain archive is marked with the field "name: Zhang San", then the corresponding identification information is "name", and the corresponding identification data is "Zhang San"; according to this rule, collect a number of identification information and corresponding identification data in each archive file, so as to form a number of <archive, identification, value> arrays related to each archive, where "archive" can be the scan sequence number of the archive.
[0066] S2, according to the <archive, identifier, value> array obtained by scanning the preset archive file, establish a first matrix, and establish a first rectangular diagram corresponding to the first matrix, the first matrix and the first rectangular diagram are marked with the identifier information of the preset archive file, used to layout the positions of all the identifier information, and store the corresponding values according to the positions of the identifier information; it should be noted that the preset archive file can be a pre-entered archive file sample with complete and accurate answer information, and the preset archive file can also be an archive file sample with relatively complete and accurate information extracted from all scanned archive files, as a reference for other archive files to be scanned. All identifiers and their corresponding values in all <archive, identifier, value> arrays of the preset archive file constitute the first matrix of the preset archive file, and the purpose of establishing the first matrix is to facilitate the calculation of the distance between other matrices. Then, according to the first matrix, a corresponding first rectangular diagram is established, and the first rectangular diagram has a similar data structure to the first matrix, and the main difference is that the first rectangular diagram can intuitively reflect the identifier position and its value obtained by scanning the preset archive file, and can also display and compare the values of the same identifier position in different rectangular diagrams horizontally and intuitively. The first matrix and the first rectangular diagram are marked with identification information of the preset archive file, and are used to layout the positions of all the identification information, such as the identification positions at the nth row and nth column, and store corresponding values according to the positions of the identification information.
[0067] S3, based on the first matrix and the first rectangular diagram, establish the tth matrix and the tth rectangular diagram of other archive files except the preset archive files; it should be noted that, on the basis of establishing the first matrix and the first rectangular diagram, a similar method is adopted to establish the tth matrix and the tth rectangular diagram of the identification information and its values extracted from all archive files to be sorted; in addition, the matrix identifier of the tth matrix is the scanning sequence number of the current archive file, and the elements of each position in the tth matrix are the corresponding values filled in the position where the corresponding identifier is located; thereby, the corresponding tth matrix and the corresponding tth rectangular diagram are established for each archive file.
[0068] S4, respectively calculating the distance D between the t-th matrix and the first matrix, establishing the positional association relationship between the first rectangular diagram and the t-th rectangular diagram based on the distance D, and forming a three-dimensional distance diagram; the three-dimensional distance diagram uses the plane where the first rectangular diagram is located as the first and second dimensions, and the vertical distance between the t-th rectangular diagram and the first rectangular diagram as the third dimension; it should be noted that, when the matrices corresponding to all archive files are established, respectively calculating the distance D between each of the t-th matrix and the first matrix, and using the distance D to identify the distance position relationship between each of the t-th matrix and the first matrix, thereby establishing the positional association relationship between the first rectangular diagram and the t-th rectangular diagram based on the distance D, and forming a three-dimensional distance diagram. Wherein, as Figure 2 As shown, the three-dimensional distance graph takes the plane where the first rectangular graph is located as the first and second dimensions, that is, the plane surrounded by the xy axes, and takes the vertical distance between the t-th rectangular graph and the first rectangular graph as the third dimension, that is, the z-axis; wherein the plane surrounded by the xy axes is perpendicular or nearly perpendicular to the z-axis; thereby realizing a three-dimensional distance graph based on the first rectangular graph that establishes a positional association relationship with several t-th rectangular graphs.
[0069] S5, arranging the order of the plurality of archive files according to the three-dimensional distance map. It should be noted that, since the position association relationship between each rectangular diagram and the first rectangular diagram has been established in the three-dimensional distance map, the distance between each rectangular diagram and the first rectangular diagram is also determined, and this distance is the sorting relationship between each archive file and the preset archive file. Specifically, the closer the distance to the preset file, the closer the sequence number of the archive file is to the preset answer file; for example, if the matrix identification number of the preset archive file is 001, the number of the answer file that is closest to the first matrix corresponding to the preset answer file along the z-axis direction should be 002, and so on, to achieve the sorting relationship of all archive files based on the preset answer file. In addition, if the position relationship between the matrix corresponding to a certain archive file and the first matrix is in the opposite direction along the z-axis direction, the answer file can be numbered as a negative number to achieve the sorting relationship between the answer file and the preset archive file; on this basis, the numbers of all sorting relationships are adjusted to numbers starting from 001. In addition, if the preset archive file is a pre-entered archive file sample with complete and accurate answer information, then in the final sorting process, the number of the archive file will be eliminated to sort other archive files; conversely, if the preset archive file is also an archive file sample with relatively complete and accurate information extracted from all scanned archive files, then in the final sorting process, there is no need to remove the number corresponding to the preset answer file.
[0070] It can be seen that the scanning device and image automatic processing method for a smart archive management system of the present invention establishes a matrix and a rectangular diagram of each archive file according to an array including identification information obtained from scanning the archive file, so as to layout the position of the identification information and its corresponding value, thereby storing multiple key information of the archive file according to the layout position; then, according to the distance between the matrices corresponding to different archive files, a position association relationship between the rectangular diagrams corresponding to different archive files is established, thereby forming a three-dimensional distance diagram of all archive files; and finally, the order of multiple archive files is obtained based on the three-dimensional distance diagram. In this way, the effective sorting of archive files is realized according to the positional relationship between the rectangular diagrams composed of several key information of the archive files, thereby achieving a more accurate sorting result, which is conducive to the efficient collection of archive files and facilitates the subsequent reference to archive files.
[0071] As a preferred implementation, the distance between the t-th matrix and the first matrix is calculated respectively. , specifically including:
[0072] ,
[0073] Among them, m is the vector coefficient, when When m=1, the distance between the current tth matrix and the first matrix is a positive number, that is, the rectangular image corresponding to the current archive file is in the positive direction of the z-axis direction in the three-dimensional distance map; on the contrary, when < When m=-1, the distance between the current t-th matrix and the first matrix is a negative number, that is, the rectangular diagram corresponding to the current archive file is in the opposite direction of the z-axis direction in the three-dimensional distance diagram; represents the value of the i-th marker of the t-th matrix, Represents the first matrix The value of the ith identifier of represents the weight of the ith identifier of the tth matrix, Represents the tth matrix and the first matrix distance.
[0074] As a preferred embodiment, it also includes:
[0075] The first matrix All values and the t-th matrix All values of are normalized once, and then the first matrix The value of the ith identifier and all the tth matrices The value of the ith mark is subjected to a secondary normalization process; it should be noted that, in order to facilitate the comparison of values within the matrix and between different matrices, the present invention performs a secondary normalization process on the matrix data, that is, the first matrix All values and the t-th matrix All values of are normalized once, that is, the normalization process is first performed between the data within the matrix; then the first matrix The value of the ith identifier and all the tth matrices The value of the ith mark is subjected to secondary normalization, that is, the data at the same mark position among all matrices are normalized again. It can be seen that the double data normalization process first performs normalization between the data within the matrix, and then performs normalization between the data at the same mark position among all matrices. It not only takes into account the unified management of data within the matrix, but also takes into account the unified management of data between matrices, which greatly improves the efficiency of data comparison.
[0076] The initial weight of the weight of the t-th matrix is set, and the values corresponding to all the identifiers of the t-th matrix after secondary normalization are obtained; it should be noted that the initial weight of the numerical weight of all the identifier positions of the t-th matrix is set to realize the importance marking of key information in the archive file. Then, the values corresponding to all the identifiers of the t-th matrix after secondary normalization are obtained to establish a corresponding relationship between each of the initial weights and the corresponding values of all the identifiers.
[0077] The values of the t-th matrix are calibrated according to the values corresponding to all the identifiers of the t-th matrix, and the initial weight is adjusted according to the calibration result. It should be noted that since there may be errors or typos in the key information in the archive file during the scanning stage or the input stage, in order to eliminate the influence of such errors, preferably, the present invention performs a matrix internal verification operation on the values corresponding to these key information, that is, all the identifiers, so as to adjust the initial weights of the values corresponding to all the identifiers in the matrix according to the calibration result. Specifically, the numerical calibration corresponding to all the identifiers of the t-th matrix adopts a mutually exclusive calibration method, that is, to select the contradictory values, and then determine which mutually exclusive values are wrong, and reduce the weight of the wrong values or assign the weight to 0; on the contrary, increase the weight of the correct values in the mutually exclusive values, so as to keep the sum of the weights of these mutually exclusive values the same or relatively stable as a whole, thereby realizing the numerical verification and weight adjustment inside the matrix, thereby improving the accuracy of subsequent judgments.
[0078] As a preferred implementation, the adjusting the initial weight according to the calibration result further includes:
[0079] If the error between the value of the i-th identifier of the t-th matrix and the corresponding values of one or more associated identifiers of the t-th matrix is greater than a first threshold, the initial weight is reduced; if the value of the i-th identifier of the t-th matrix and the corresponding values of one or more associated identifiers are mutually exclusive data, the initial weight of the corresponding value of the identifier with a higher error rate among the corresponding values of all associated identifiers of the t-th matrix is set to 0. It should be noted that if the error between the value of the i-th identifier of the t-th matrix and the corresponding value of one or more associated identifiers of the t-th matrix is greater than the first threshold, it indicates that the value of the i-th identifier is very likely to be erroneous. For example, the value of the i-th identifier (gender) is 1, meaning male, and the normalized value is 0.8. Another identifier that may be related to gender (such as educational experience) records "xx Girls' Middle School". After normalization, the numerical information of the identifier and its gender-related attributes obtain a value of 0.2, so the error between the two is m=0.8-0.2=0.6>the first threshold M=0.1. The value of another identifier (such as an ID number) calculated in the same way is normalized, and the error between the value and 0.8 is still greater than the first threshold 0.1. At this time, it indicates that the value of the i-th identifier of the t-th matrix is very likely to be erroneous, so its weight is reduced on the basis of the initial weight to reduce its error or the impact of the error. Furthermore, if the i-th identifier of the t-th matrix only identifies errors or mistakes, its weight is reduced to 0 on the basis of the initial weight; if the i-th identifier of the t-th matrix identifies other attributes in addition to errors or mistakes, for example, "educational experience" also identifies attribute information such as the region and school where the student studied, and "ID number" also identifies attribute information such as place of origin and date of birth, then its weight is appropriately reduced on the basis of the initial weight according to the number of attribute information identified by the identifier; for example, the more attribute information identified by the identifier, the more its weight is reduced on the basis of the initial weight.
[0080] If the error between the numerical value of the ith identifier of the tth matrix and the corresponding numerical values of one or more associated identifiers of the tth matrix is less than or equal to the first threshold, the initial weight is increased; it should be noted that if the error between the numerical value of the ith identifier of the tth matrix and the corresponding numerical values of one or more associated identifiers of the tth matrix is less than or equal to the first threshold, it indicates that the credibility of the numerical information corresponding to these identifiers is relatively high, so its weight is increased on the basis of the initial weight; for example, the error between the numerical value of the ith identifier of the tth matrix and the corresponding numerical values of one or more associated identifiers of the tth matrix is n=0.6-0.55=0.05, <the first threshold M=0.1, then the initial weight is increased.
[0081] As a preferred implementation, the initial weight is reduced by a weight reduction formula, and the weight reduction formula is:
[0082] ,
[0083] Among them, α and β represent the preset parameters in the weight reduction formula, which can be set according to actual needs and are not limited here; represents the reduction multiple of the initial weight of the value corresponding to the identifier to be reduced, and the reduction multiple is positively correlated with the ratio of the error m to the first threshold M. For example, continuing the above embodiment, the initial weight of the value corresponding to the identifier to be reduced is , =m / M=0.6 / 0.1=6, that is, the initial weight of the value corresponding to the mark to be reduced is reduced by 6 times; It represents the weight after the initial weight of the value corresponding to the i-th identifier of the t-th matrix is reduced. Indicates the initial weight of the value corresponding to the identifier to be reduced =0.06, e represents a natural constant.
[0084] As a preferred implementation, the initial weight is increased by a weight increasing formula, and the weight increasing formula is:
[0085] ,
[0086] in, represents the sum of weights corresponding to all identifiers of the first matrix before increase, a, b, c represent preset parameters in the weight increase formula, preferably, a=3, b=-4, c=0.2; represents the initial weight increase multiple of the value corresponding to the identifier to be increased, and the increase multiple n is negatively correlated with the ratio of the error to the first threshold. For example, continuing the above embodiment, the initial weight of the value corresponding to the identifier to be increased is , =1 / (n / M)=M / n=0.1 / 0.05=2, that is, the initial weight increase factor of the value corresponding to the mark to be increased is 2; It represents the weight after the initial weight of the value corresponding to the i-th identifier of the t-th matrix is increased, and e represents a natural constant.
[0087] As a preferred implementation, the initial weight of the corresponding value of the identifier with a higher error rate among the corresponding values of all associated identifiers of the t-th matrix is set to 0, specifically including:
[0088] Obtain the corresponding numerical values of all associated identifiers of the t-th matrix, and perform positive and negative voting according to the attributes of all the numerical values; the corresponding identifier of the numerical value corresponding to the attribute with a lower vote rate is used as the identifier with a higher error rate; it should be noted that in order to accurately identify which one or which of the multiple identifiers with associated attributes are wrong, preferably, the present invention adopts an intra-matrix voting statistics method for identification; that is, obtain the corresponding numerical values of all associated identifiers of the t-th matrix, and perform positive and negative voting according to the attributes of all the numerical values; for example, there are 5 identifiers with gender attributes in the current t-th matrix, of which 3 are male and 2 are female, then the corresponding identifier of the numerical value corresponding to the attribute with a lower vote rate (i.e., female) is used as the identifier with a higher error rate; and based on this, the initial weight of the corresponding numerical value of the identifier with a higher error rate in the corresponding numerical values of all associated identifiers of the t-th matrix is set to 0 to avoid the misleading influence of the corresponding numerical value of the wrong attribute on the analysis result.
[0089] Alternatively, the corresponding values of all associated identifiers of other matrices other than the t-th matrix are obtained, and positive and negative votes are performed according to the attributes of all the numerical values; the corresponding identifier of the corresponding value of the attribute with a lower vote rate is used as an identifier with a higher error rate. It should be noted that in order to accurately identify which or which of the multiple identifiers with associated attributes are wrong, preferably, the present invention adopts a voting statistics method between matrices for identification; that is, the corresponding values of all associated identifiers of other matrices other than the t-th matrix are obtained, and positive and negative votes are performed according to the attributes of all the numerical values. For example, there are 7 identifiers with gender attributes in all matrices other than the t-th matrix, of which 5 are male and 2 are female, then the corresponding identifier of the corresponding value of the attribute with a lower vote rate (i.e., female) is used as an identifier with a higher error rate, and based on this, the initial weight of the corresponding value of the identifier with a higher error rate in the corresponding values of all associated identifiers of the t-th matrix is set to 0 to avoid the misleading influence of the corresponding value of the wrong attribute on the analysis result.
[0090] As another embodiment, the present invention provides a scanning device for a smart archive management system, the device comprising the following modules:
[0091] The information extraction module is used to obtain scanned images of multiple archive files, extract the identification information and corresponding identification data in each of the scanned images, and form an array of <archive, identification, value>; it should be noted that since the information in the archive file is multiple, for example, it can include information such as archive number, name, gender, age, learning experience, work experience, etc., in order to get rid of the problem of inaccurate sorting results caused by sorting archives based only on a few information such as archive number in the prior art, the archive information in the present invention is usually multiple, so as to minimize the error caused by a single information to the sorting result. Among them, the identification information and corresponding identification data in each of the scanned images are extracted, the identification information is the field name in the archive information, and the identification data is the content of the corresponding field. For example, if the field "name: Zhang San" is marked in a certain archive, the corresponding identification information is "name", and the corresponding identification data is "Zhang San"; according to this rule, a number of identification information and corresponding identification data in each archive file are collected, so as to form a number of <archive, identification, value> arrays related to each archive, where "archive" can be the scan sequence number of the archive.
[0092] The information storage module is used to establish a first matrix according to the <archive, identifier, value> array obtained by scanning the preset archive file, and establish a first rectangular diagram corresponding to the first matrix. The first matrix and the first rectangular diagram are marked with the identifier information of the preset archive file, and are used to layout the positions of all the identifier information, and store the corresponding values according to the positions of the identifier information; it should be noted that the preset archive file can be a pre-entered archive file sample with complete and accurate answer information, and the preset archive file can also be an archive file sample with relatively complete and accurate information extracted from all scanned archive files, as a reference for other archive files to be scanned. All identifiers and their corresponding values in all <archive, identifier, value> arrays of the preset archive file constitute the first matrix of the preset archive file, and the purpose of establishing the first matrix is to facilitate the calculation of the distance between other matrices. Then, a corresponding first rectangular diagram is established according to the first matrix, and the first rectangular diagram has a similar data structure to the first matrix, and the main difference is that the first rectangular diagram can intuitively reflect the identifier position and its value obtained by scanning the preset archive file, and can also display and compare the values of the same identifier position in different rectangular diagrams horizontally and intuitively. The first matrix and the first rectangular diagram are marked with identification information of the preset archive file, and are used to layout the positions of all the identification information, such as the identification positions at the nth row and nth column, and store corresponding values according to the positions of the identification information.
[0093] An information layout module is used to establish the tth matrix and the tth rectangular diagram of other archive files except the preset archive files based on the first matrix and the first rectangular diagram; it should be noted that, on the basis of establishing the first matrix and the first rectangular diagram, a similar method is adopted to establish the tth matrix and the tth rectangular diagram of the identification information and its values extracted from all archive files to be sorted; in addition, the matrix identifier of the tth matrix is the scanning serial number of the current archive file, and the elements of each position in the tth matrix are the corresponding values filled in the position where the corresponding identifier is located; thereby, the corresponding tth matrix and the corresponding tth rectangular diagram are established for each archive file.
[0094] The information calculation module is used to calculate the distance D between the t-th matrix and the first matrix respectively, establish the positional association relationship between the first rectangular diagram and the t-th rectangular diagram based on the distance D, and form a three-dimensional distance diagram; the three-dimensional distance diagram uses the plane where the first rectangular diagram is located as the first and second dimensions, and the distance between the t-th rectangular diagram and the first rectangular diagram as the third dimension; it should be noted that, when the matrices corresponding to all archive files are established, the distance D between each t-th matrix and the first matrix is calculated respectively, and the distance position relationship between each t-th matrix and the first matrix is identified by the distance D, so that the positional association relationship between the first rectangular diagram and the t-th rectangular diagram is established based on the distance D, and a three-dimensional distance diagram is formed. Wherein, as Figure 2 As shown, the three-dimensional distance graph takes the plane where the first rectangular graph is located as the first and second dimensions, that is, the plane surrounded by the xy axes, and takes the vertical distance between the t-th rectangular graph and the first rectangular graph as the third dimension, that is, the z-axis; wherein the plane surrounded by the xy axes is perpendicular or nearly perpendicular to the z-axis; thereby realizing a three-dimensional distance graph based on the first rectangular graph that establishes a positional association relationship with several t-th rectangular graphs.
[0095] An information sorting module is used to arrange the order of the multiple archive files according to the three-dimensional distance map. It should be noted that, since the position association relationship between each rectangular diagram in the three-dimensional distance map and the first rectangular diagram has been established, the distance between each rectangular diagram and the first rectangular diagram is also determined, and this distance is the sorting relationship between each archive file and the preset archive file. Specifically, the closer the distance to the preset file, the closer the sequence number of the archive file is to the preset answer file; for example, if the matrix identification number of the preset archive file is 001, the number of the answer file that is closest to the first matrix corresponding to the preset answer file along the z-axis direction should be 002, and so on, to achieve the sorting relationship of all archive files based on the preset answer file. In addition, if the position relationship between the matrix corresponding to a certain archive file and the first matrix is in the opposite direction along the z-axis direction, the answer file can be numbered as a negative number to achieve the sorting relationship between the answer file and the preset archive file; on this basis, the numbers of all sorting relationships are adjusted to numbers starting from 001. In addition, if the preset archive file is a pre-entered archive file sample with complete and accurate answer information, then in the final sorting process, the number of the archive file will be eliminated to sort other archive files; conversely, if the preset archive file is also an archive file sample with relatively complete and accurate information extracted from all scanned archive files, then in the final sorting process, there is no need to remove the number corresponding to the preset answer file.
[0096] It can be seen that the scanning device and image automatic processing method for a smart archive management system of the present invention establishes a matrix and a rectangular diagram of each archive file according to an array including identification information obtained from scanning the archive file, so as to layout the position of the identification information and its corresponding value, thereby storing multiple key information of the archive file according to the layout position; then, according to the distance between the matrices corresponding to different archive files, a position association relationship between the rectangular diagrams corresponding to different archive files is established, thereby forming a three-dimensional distance diagram of all archive files; and finally, the order of multiple archive files is obtained based on the three-dimensional distance diagram. In this way, the effective sorting of archive files is realized according to the positional relationship between the rectangular diagrams composed of several key information of the archive files, thereby achieving a more accurate sorting result, which is conducive to the efficient collection of archive files and facilitates the subsequent reference to archive files.
[0097] Embodiment three:
[0098] As another embodiment, the present invention provides a scanning system for a smart file management system, wherein the system executes the automatic image processing method of the aforementioned embodiment.
[0099] It will be appreciated by those skilled in the art that the present invention includes devices for performing one or more of the operations described in the present application. These devices may be specially designed and manufactured for the desired purpose, or may include known devices in general-purpose computers. These devices have computer programs stored therein, which are selectively activated or reconfigured. Such computer programs may be stored in a device (e.g., computer) readable medium or in any type of medium suitable for storing electronic instructions and coupled to a bus, respectively, the computer readable medium including but not limited to any type of disk (including floppy disk, hard disk, optical disk, CD-ROM, and magneto-optical disk), ROM (Read-Only Memory), RAM (Random Access Memory), EPROM (Erasable Programmable Read-Only Memory), EEPROM (Electrically Erasable Programmable Read-Only Memory), flash memory, magnetic card or optical card. That is, the readable medium includes any medium that stores or transmits information in a readable form by a device (e.g., computer).
[0100] Those skilled in the art will appreciate that each block in these structure diagrams and / or block diagrams and / or flow charts and combinations of blocks in these structure diagrams and / or block diagrams and / or flow charts can be implemented using computer program instructions. Those skilled in the art will appreciate that these computer program instructions can be provided to a general-purpose computer, a professional computer, or a processor of other programmable data processing methods to implement, thereby executing the scheme specified in the block or multiple blocks of the structure diagram and / or block diagram and / or flow chart disclosed in the present invention through the processor of the computer or other programmable data processing method.
[0101] Those skilled in the art will appreciate that the various operations, methods, steps, measures, and schemes discussed in the present invention may be alternated, modified, combined, or deleted. Further, other steps, measures, and schemes in the various operations, methods, and schemes discussed in the present invention may also be alternated, modified, rearranged, decomposed, combined, or deleted. Further, the steps, measures, and schemes in the prior art that are similar to those disclosed in the present invention may also be alternated, modified, rearranged, decomposed, combined, or deleted.
[0102] The above descriptions are merely embodiments of the present invention and are not intended to limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made using the contents of the present invention specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.
Claims
1. An automatic image processing method for a smart archive management system. It is characterized in that The method comprises the following steps: S1, obtaining scanned images of multiple archive files, extracting identification information and corresponding identification data in each of the scanned images, and forming an array of <archive, identification, value>; S2, establishing a first matrix according to the <file, identifier, value> array obtained by scanning the preset archive file, and establishing a first rectangular diagram corresponding to the first matrix, wherein the first matrix and the first rectangular diagram are marked with the identifier information of the preset archive file, and are used to lay out the positions of all the identifier information of the preset archive file, and store the corresponding values according to the positions of the identifier information of the preset archive file; the preset archive file is a pre-entered archive file sample with complete and accurate archive information; S3, establishing a t-th matrix and a t-th rectangular diagram of other archive files except the preset archive file based on the first matrix and the first rectangular diagram; S4, respectively calculating the distance between the t-th matrix and the first matrix , at the distance As a reference, a position association relationship between the first rectangular diagram and the t-th rectangular diagram is established, and a three-dimensional distance diagram is formed; the three-dimensional distance diagram takes the plane where the first rectangular diagram is located as the first and second dimensions, and takes the vertical distance between the t-th rectangular diagram and the first rectangular diagram as the third dimension; S5, arranging the order of the plurality of archive files according to the three-dimensional distance map; Calculate the distance between the tth matrix and the first matrix respectively , including: Among them, m is the vector coefficient, when When m=1; when < When m=-1; represents the value of the i-th marker of the t-th matrix, Represents the first matrix The value of the ith identifier of represents the weight of the ith identifier of the tth matrix, Represents the tth matrix and the first matrix distance.
2. The method for automatic image processing according to claim 1, It is characterized in that Also includes: The first matrix All values and the t-th matrix All values of are normalized once, and then the first matrix The value of the ith identifier and all the tth matrices The value of the ith identifier is subjected to secondary normalization; Setting the initial weight of the t-th matrix weight, and obtaining the values corresponding to all the identifiers of the t-th matrix after secondary normalization processing; An internal matrix calibration operation is performed on the values corresponding to all the identifiers of the t-th matrix after the secondary normalization, and the initial weights are adjusted according to the calibration results.
3. The method for automatic image processing according to claim 2, It is characterized in that The adjusting the initial weight according to the calibration result further includes: If the error between the value of the i-th identifier of the t-th matrix and the corresponding values of one or more associated identifiers of the t-th matrix is greater than a first threshold, reducing the initial weight; If the error between the value of the i-th identifier of the t-th matrix and the corresponding values of one or more associated identifiers of the t-th matrix is less than or equal to a first threshold, increasing the initial weight; If the value of the i-th identifier of the t-th matrix and the corresponding values of one or more associated identifiers are mutually exclusive data, the initial weight of the corresponding value of the identifier with a higher error rate among the corresponding values of all associated identifiers of the t-th matrix is set to 0.
4. The method for automatic image processing according to claim 3, It is characterized in that The initial weight is reduced by a weight reduction formula, and the weight reduction formula is: in, , represents the preset parameter in the weight reduction formula, represents the initial weight reduction multiple of the value corresponding to the identifier to be reduced, and the reduction multiple is positively correlated with the ratio of the error to the first threshold value, It represents the weight after the initial weight of the value corresponding to the i-th identifier of the t-th matrix is reduced. It represents the initial weight of the value corresponding to the mark to be reduced, and e represents a natural constant.
5. The method for automatic image processing according to claim 3, It is characterized in that The initial weight is increased by a weight increasing formula, and the weight increasing formula is: in, represents the sum of weights corresponding to all identifiers of the first matrix before increase, a, b, c represent the preset parameters in the weight increase formula, represents the initial weight increase multiple of the value corresponding to the identifier that needs to be increased, and the increase multiple is negatively correlated with the ratio of the error to the first threshold value, It represents the weight after the initial weight of the value corresponding to the i-th identifier of the t-th matrix is increased, and e represents a natural constant.
6. The method for automatic image processing according to claim 3, It is characterized in that The initial weight of the corresponding value of the identifier with a higher error rate among the corresponding values of all associated identifiers of the t-th matrix is set to 0, specifically including: Obtain the corresponding values of all associated identifiers of the t-th matrix, and perform positive and negative voting according to the attributes of all the values; the corresponding identifier of the attribute corresponding to the value with a lower vote rate is used as the identifier with a higher error rate; Alternatively, the corresponding values of all associated identifiers of matrices other than the t-th matrix are obtained, and positive and negative voting is performed according to the attributes of all the values; the corresponding identifiers of the values corresponding to the attributes with lower vote rates are used as identifiers with higher error rates.
7. A scanning device for a smart archive management system, It is characterized in that To implement the automatic image processing method according to any one of claims 1 to 6, the device comprises the following modules: An information extraction module, used for obtaining scanned images of a plurality of archive files, extracting identification information and corresponding identification data in each of the scanned images, and forming an array of <archive, identification, value>; An information storage module, used to establish a first matrix according to the <archive, identification, value> array obtained by scanning a preset archive file, and to establish a first rectangular diagram corresponding to the first matrix, wherein the first matrix and the first rectangular diagram are marked with identification information of the preset archive file, and are used to lay out the positions of all identification information of the preset archive file, and store corresponding values according to the positions of the identification information of the preset archive file; the preset archive file is a pre-entered archive file sample with complete and accurate archive information; An information layout module, used for establishing a t-th matrix and a t-th rectangular diagram of other archive files except the preset archive file based on the first matrix and the first rectangular diagram; An information calculation module, used to calculate the distance between the t-th matrix and the first matrix respectively , at the distance As a reference, a position association relationship between the first rectangular diagram and the t-th rectangular diagram is established, and a three-dimensional distance diagram is formed; the three-dimensional distance diagram takes the plane where the first rectangular diagram is located as the first and second dimensions, and takes the distance between the t-th rectangular diagram and the first rectangular diagram as the third dimension; An information sorting module, used for arranging the order of the plurality of archive files according to the three-dimensional distance map; Calculate the distance between the tth matrix and the first matrix respectively , specifically including: Among them, m is the vector coefficient, when When m=1; when < When m=-1; represents the value of the i-th marker of the t-th matrix, Represents the first matrix The value of the ith identifier of represents the weight of the ith identifier of the tth matrix, Represents the tth matrix and the first matrix distance.
8. A scanning device for a smart archive management system, It is characterized in that The system executes the automatic image processing method as described in any one of claims 1-6.
9. A computer-readable storage medium storing a computer program. It is characterized in that The computer program executes the automatic image processing method according to any one of claims 1 to 6.
