Comprehensive and integrated entry system for work data archives

Abstract

The invention discloses a comprehensive and integrated entry system for work data archives, and relates to the technical field of work data archive entry. A work data file image acquisition module; the feature splitting module is used for performing content splitting and fusion on the work data archive image according to a content splitting algorithm to obtain a plurality of fusion feature sets; the matching calculation module is used for calculating the content matching degree of each fusion feature set according to a matching degree algorithm; the screening module is used for screening out a fusion feature set corresponding to the maximum value of the content matching degree, and taking the fusion feature set as an image region division feature group; and the region division module is used for performing content region division on the work data archive image based on the work data archive image region corresponding to each fusion feature in the image region division feature group. High-precision matching between content recognition and features of the work data archive image is ensured, and the precision of content recognition of the work data archive image is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of work data file entry, and in particular to a comprehensive integrated work data file entry system. Background Art

[0002] The digitization of work archives has far-reaching and extensive significance. It is not only an innovation in traditional archive management methods, but also an important means to improve work efficiency, ensure information security, promote resource sharing and promote scientific decision-making. Digital archives can be quickly processed, retrieved and transmitted through computer systems, greatly reducing the time for manual browsing, searching and sorting paper archives. This not only improves the work efficiency of archive managers, but also facilitates users to quickly obtain the required information when needed. Traditional work data archive entry systems usually scan archives into pictures for digital preservation. However, digital archives generated in this way are difficult to adjust and change conveniently, and it is difficult to meet the efficient operation and management of archives. Therefore, converting physical archives into editable digital content is an important technical issue in this field.

[0003] In the existing work data archive comprehensive integrated entry system, the analysis depth of the work data archive image features is insufficient, it is difficult to identify the intrinsic correlation of multiple features in the work data archive image, and it is difficult to fully understand the overall meaning of the work data archive image. When generating editable digital content, it is easy to produce incorrect recognition and matching, resulting in work data archive entry errors. Summary of the invention

[0004] In order to solve the above-mentioned technical problems, a work data archive comprehensive integrated entry system is provided. The technical solution solves the problem that in the above-mentioned existing work data archive comprehensive integrated entry system, the analysis depth of the work data archive image features is insufficient, it is difficult to identify the intrinsic correlation of multiple features in the work data archive image, it is difficult to fully understand the overall meaning of the work data archive image, and it is easy to produce erroneous recognition and matching when generating editable digital content, resulting in errors in the entry of work data archives.

[0005] In order to achieve the above purpose, the technical solution adopted by the present invention is:

[0006] A comprehensive integrated work data file entry system, characterized by comprising:

[0007] A storage module, wherein the storage module is used to store a content database;

[0008] A working data archive image acquisition module, wherein the working data archive image acquisition module is used to acquire working data archive images;

[0009] A feature splitting module, the feature splitting module is electrically connected to the working data archive image acquisition module, and the feature splitting module is used to perform content splitting and fusion on the working data archive image according to the content splitting algorithm to obtain a plurality of fusion feature sets;

[0010] A matching calculation module, the matching calculation module is electrically connected to the storage module and the feature separation module, and the matching calculation module is used to calculate the content matching degree of each fused feature set according to a matching degree algorithm;

[0011] A screening module, the screening module is electrically connected to the matching calculation module, and the screening module is used to screen out a fusion feature set corresponding to a maximum content matching degree, and use the fusion feature set as an image region division feature group;

[0012] A region division module is electrically connected to the screening module, and is used to divide the working data archive image into content regions based on the working data archive image region corresponding to each fusion feature in the image region division feature group.

[0013] Preferably, the feature splitting module includes:

[0014] A feature extraction unit, the feature extraction unit is used to obtain a total number M of features with the same content in the working data archive image;

[0015] A fusion condition unit, wherein the splitting condition unit is used to generate feature fusion restriction conditions;

[0016] The feature fusion unit is used to generate a plurality of feature fusion sets under the restriction of feature fusion restriction conditions, and extract a plurality of content features with the same boundaries from the working data archive image according to the numbers in the feature fusion array, and combine them into a fusion feature set.

[0017] Preferably, the expression of the feature fusion restriction condition is:

[0018]

[0019] In the formula, z i is the number of identical content features included in the i-th fusion feature, and N+ is a positive integer set.

[0020] Preferably, the step of obtaining the total number M of identical content features in the working data archive image specifically includes:

[0021] The original archive image is gray-scaled by using the weighted average method to obtain a gray-scale image;

[0022] A grayscale similarity deviation threshold is set, and continuous intervals whose grayscale deviation in the working data grayscale image is less than the grayscale similarity deviation threshold are divided into the same content features;

[0023] Count the total number of features with the same content in the grayscale image of the working data, denoted as M;

[0024] The weighted average formula is:

[0025] Gray(i,j)=α1R(i,j)+α2G(i,j)+α3B(i,j),

[0026] Where Gray(i,j) is the grayscale of the pixel in the i-th row and j-th column in the working data archive image, α1, α2, and α3 are the red, green, and blue sensitivity weighting coefficients, respectively, and R(i,j), G(i,j), and B(i,j) are the red, green, and blue components of the pixel in the i-th row and j-th column, respectively.

[0027] Preferably, the extracting of a plurality of identical content features having the same boundaries from the working data archive image according to the numbers in the feature fusion array and combining them into a fusion feature set specifically includes:

[0028] S1, select an element from the feature fusion set, denoted as Z;

[0029] S2, selecting a same content feature from the working data archive image as a reference content feature;

[0030] S3, set a fusion index j, and add an initial value of 1 to j;

[0031] S4, determine whether j is equal to Z, if so, use the reference content feature as the fusion feature and proceed to step S6, if not, add one to the value of j and proceed to step S5;

[0032] S5, selecting the same content feature that has the same boundary as the reference content feature, fusing it with the reference content feature to form a new reference content feature, and returning to step S4;

[0033] S6, determine whether to traverse all elements in the feature fusion set, if yes, proceed to S7, if no, remove Z from the feature fusion set and return to step S1;

[0034] S7, combining all fused features into a fused feature set;

[0035] S8. Repeat steps S1-S7 until a new fused feature set cannot be generated under the feature fusion set.

[0036] Preferably, the matching calculation module includes:

[0037] A similarity fitting unit, wherein the similarity fitting unit is used to calculate the similarity between each fusion feature in the fusion feature set and each content feature in the content database, and select the content feature with the greatest similarity as the target feature of the fusion feature;

[0038] An association calculation unit, the association calculation unit is used to calculate the feature association degree based on the target feature of each fused feature in the fused feature set, and obtain the feature association degree of the fused feature set;

[0039] The matching calculation unit is used to calculate the content matching degree of the fused feature set through a matching degree calculation formula based on the similarity between each fused feature in the fused feature set and its target feature and the feature association degree of the fused feature set.

[0040] Preferably, the matching degree calculation formula is specifically:

[0041]

[0042] In the formula, P is the content matching degree of the fusion feature set, C hi is the similarity between the i-th fusion feature in the fusion feature set and its target feature, G l is the feature correlation of the fused feature set.

[0043] Preferably, the calculation of feature correlation based on the target feature of each fused feature in the fused feature set to obtain the feature correlation of the fused feature set specifically includes:

[0044] Determine the content feature class corresponding to the target feature of each fused feature;

[0045] The fusion features in the fusion feature set are combined into several fusion feature groups in pairs, and the fusion feature groups composed of fusion features with the same boundary are recorded as adjacent fusion feature groups, and the fusion feature groups composed of fusion features with different boundaries are recorded as interval fusion feature groups;

[0046] Based on the set association logic, determine the association degree between the content feature classes corresponding to the target features of the two fused features of each fused feature group, and record it as the association degree of the fused feature group;

[0047] Based on the correlation of all fused feature groups, the feature correlation of the fused feature set is calculated using the feature correlation comprehensive formula;

[0048] The feature correlation comprehensive formula is specifically:

[0049]

[0050] In the formula, g kis the correlation degree of the kth adjacent fusion feature group, g u is the correlation degree of the u-th interval fusion feature group, K is the total number of adjacent fusion feature groups, and U is the total number of interval fusion feature groups.

[0051] Compared with the prior art, the present invention has the following beneficial effects:

[0052] The present invention proposes a comprehensive integrated input scheme for working data archives. By carefully splitting the working data archive image into a plurality of fused feature sets, and combining the content feature fitting of the fused feature set with the content database and the dual recognition and judgment of the correlation between the fused features in the fused feature set, the feature analysis depth of the working data archive image is greatly improved, and the overall matching of the working data archive image when the feature area is divided can be effectively guaranteed, thereby ensuring the high-precision matching between the content recognition and the features of the working data archive image, and improving the accuracy of content recognition of the working data archive image. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 This is a structural block diagram of the integrated integrated input system for work data archives proposed by the present invention; Figure 2 It is a method flow chart of the content splitting algorithm in the present invention; Figure 3 A flow chart of a method for obtaining the total number M of identical content features in a working data archive image in the present invention; Figure 4 A flow chart of a method for extracting a plurality of identical content features having the same boundaries from a working data archive image according to numbers in a feature fusion array and combining them into a fusion feature set in the present invention; Figure 5 is a method flow chart of the matching algorithm in the present invention; Figure 6 It is a flow chart of the method for obtaining the feature correlation of the fused feature set in the present invention; Figure 7 This is a schematic diagram of the structure of the electronic device proposed by the present invention; Figure 8 This is a schematic diagram of the computer-readable storage medium structure proposed by the present invention. DETAILED DESCRIPTION

[0059] The following description is used to disclose the present invention so that those skilled in the art can implement the present invention. The preferred embodiments described below are only examples, and those skilled in the art may think of other obvious variations.

[0060] Reference Figure 1 As shown, a comprehensive integrated input system for work data archives includes:

[0061] A storage module, the storage module is used to store a content database;

[0062] A working data archive image acquisition module, which is used to acquire working data archive images;

[0063] A feature splitting module, the feature splitting module is electrically connected to the working data archive image acquisition module, and the feature splitting module is used to perform content splitting and fusion on the working data archive image according to the content splitting algorithm to obtain a plurality of fusion feature sets;

[0064] A matching calculation module, the matching calculation module is electrically connected to the storage module and the feature splitting module, and the matching calculation module is used to calculate the content matching degree of each fused feature set according to the matching degree algorithm;

[0065] A screening module, the screening module is electrically connected to the matching calculation module, and the screening module is used to screen out a fusion feature set corresponding to a maximum content matching degree, and use the fusion feature set as an image region division feature group;

[0066] The region division module is electrically connected to the screening module, and is used to divide the content regions in the working data archive image based on the working data archive image region corresponding to each fusion feature in the image region division feature group.

[0067] The feature splitting module includes:

[0068] A feature extraction unit, the feature extraction unit is used to obtain a total number M of features with the same content in the working data archive image;

[0069] The fusion condition unit and the split condition unit are used to generate feature fusion restriction conditions;

[0070] The feature fusion unit is used to generate a plurality of feature fusion sets under the feature fusion restriction condition, and extract a plurality of same content features with the same boundaries from the working data archive image according to the numbers in the feature fusion array, and aggregate them into a fusion feature set;

[0071] The matching calculation module includes:

[0072] The similarity fitting unit is used to calculate the similarity between each fusion feature in the fusion feature set and each content feature in the content database, and select the content feature with the greatest similarity as the target feature of the fusion feature;

[0073] The association calculation unit is used to calculate the feature association degree based on the target feature of each fused feature in the fused feature set to obtain the feature association degree of the fused feature set;

[0074] The matching calculation unit is used to calculate the content matching degree of the fused feature set through a matching degree calculation formula based on the similarity between each fused feature in the fused feature set and its target feature and the feature correlation degree of the fused feature set.

[0075] The working process of the above system is:

[0076] Step 1: The work data archive image acquisition module acquires the work data archive image;

[0077] Step 2: The feature extraction unit obtains the total number M of features with the same content in the working data archive image;

[0078] Step 3: Split the conditional units to generate feature fusion constraints;

[0079] Step 4: The feature fusion unit generates several feature fusion sets under the feature fusion restriction conditions, and extracts several features of the same content with the same boundaries from the working data archive image according to the numbers in the feature fusion array, and aggregates them into a fusion feature set;

[0080] Step 5: The similarity fitting unit calculates the similarity between each fusion feature in the fusion feature set and each content feature in the content database, and selects the content feature with the greatest similarity as the target feature of the fusion feature;

[0081] Step 6: The correlation calculation unit calculates the feature correlation based on the target feature of each fused feature in the fused feature set to obtain the feature correlation of the fused feature set;

[0082] Step 7: The matching calculation unit calculates the content matching degree of the fused feature set through a matching degree calculation formula based on the similarity between each fused feature in the fused feature set and its target feature and the feature correlation degree of the fused feature set;

[0083] Step 8: The screening module screens out the fused feature set corresponding to the maximum content matching degree, and uses the fused feature set as the image region division feature group;

[0084] Step nine: The region division module divides the working data archive image into content regions based on the working data archive image region corresponding to each fusion feature in the image region division feature group.

[0085] Reference Figure 2 As shown, the content splitting algorithm specifically includes:

[0086] Obtain the total number M of identical content features in the working data archive image;

[0087] Sequentially traverse the positive integers in the range of 0-M as the feature fusion number n;

[0088] Based on the total number M of features with the same content in the working data archive image and the number n of feature fusions, a feature fusion restriction condition is generated;

[0089] Under the constraints of feature fusion constraints, several feature fusion sets are generated;

[0090] According to the numbers in the feature fusion array, a number of features with the same content and the same boundaries are extracted from the working data archive image, and the features are grouped into a fusion feature set;

[0091] Among them, the expression of feature fusion restriction condition is:

[0092]

[0093] In the formula, z i is the number of identical content features included in the i-th fusion feature, and N+ is a positive integer set.

[0094] Reference Figure 3 As shown, obtaining the total number M of identical content features in the working data archive image specifically includes:

[0095] The original archive image is gray-scaled by using the weighted average method to obtain a gray-scale image;

[0096] A grayscale similarity deviation threshold is set, and continuous intervals in the grayscale image of the working data whose grayscale deviation is less than the grayscale similarity deviation threshold are divided into the same content features;

[0097] Count the total number of features with the same content in the grayscale image of the working data, denoted as M;

[0098] The weighted average formula is:

[0099] Gray(i,j)=α1R(i,j)+α2G(i,j)+α3B(i,j),

[0100] Where Gray(i,j) is the grayscale of the pixel in the i-th row and j-th column in the working data archive image, α1, α2, and α3 are the red, green, and blue sensitivity weighting coefficients, respectively, and R(i,j), G(i,j), and B(i,j) are the red, green, and blue components of the pixel in the i-th row and j-th column, respectively.

[0101] Reference Figure 4 As shown, according to the numbers in the feature fusion array, several features with the same content and the same boundaries are extracted from the working data archive image, and are combined into a fusion feature set, which specifically includes:

[0102] S1, select an element from the feature fusion set, denoted as Z;

[0103] S2, selecting a same content feature from the working data archive image as a reference content feature;

[0104] S3, set a fusion index j, and add an initial value of 1 to j;

[0105] S4, determine whether j is equal to Z, if so, use the reference content feature as the fusion feature and proceed to step S6, if not, add one to the value of j and proceed to step S5;

[0106] S5, selecting the same content feature that has the same boundary as the reference content feature, fusing it with the reference content feature to form a new reference content feature, and returning to step S4;

[0107] S6, determine whether to traverse all elements in the feature fusion set, if yes, proceed to S7, if no, remove Z from the feature fusion set and return to step S1;

[0108] S7, combining all fused features into a fused feature set;

[0109] S8. Repeat steps S1-S7 until a new fused feature set cannot be generated under the feature fusion set.

[0110] The implementation process of the above content splitting algorithm is explained with a specific embodiment;

[0111] In some examples, the number of identical features in the working data archive image obtained by performing identical content features on the working data archive image is 3, and the three identical features all have identical boundaries with each other, and the three identical features are respectively recorded as A, B, and C;

[0112] When determining the feature fusion number n, the values ​​of n are 1, 2, and 3, which respectively represent extracting 1 fusion feature, 2 fusion features, and 3 fusion features from the working data archive image;

[0113] Taking the extraction of two fused features as an example, based on the above-mentioned feature fusion restriction conditions, the generated feature fusion set is {1,2}, which means that in the generated fused feature set, one fused feature is composed of one identical feature, and one fused feature is composed of two identical features. Based on the fact that the three identical features in the example have the same boundaries with each other, the fused feature sets corresponding to the feature fusion set {1,2} are {A,BC}, {B,AC}, and {C,AB}.

[0114] Reference Figure 5 As shown, the matching algorithm is specifically as follows:

[0115] Calculate the similarity between each fusion feature in the fusion feature set and each content feature in the content database, and select the content feature with the greatest similarity as the target feature of the fusion feature;

[0116] Based on the target feature of each fused feature in the fused feature set, feature correlation is calculated to obtain the feature correlation of the fused feature set;

[0117] Based on the similarity between each fused feature in the fused feature set and its target feature and the feature correlation of the fused feature set, the content matching degree of the fused feature set is calculated by the matching degree calculation formula;

[0118] The specific formula for calculating the matching degree is:

[0119]

[0120] In the formula, P is the content matching degree of the fusion feature set, C hi is the similarity between the i-th fusion feature in the fusion feature set and its target feature, G l is the feature correlation of the fused feature set.

[0121] Reference Figure 6 As shown, based on the target feature of each fused feature in the fused feature set, the feature correlation degree is calculated, and the feature correlation degree of the fused feature set is obtained, which specifically includes:

[0122] Determine the content feature class corresponding to the target feature of each fused feature;

[0123] The fusion features in the fusion feature set are combined into several fusion feature groups in pairs, and the fusion feature groups composed of fusion features with the same boundary are recorded as adjacent fusion feature groups, and the fusion feature groups composed of fusion features with different boundaries are recorded as interval fusion feature groups;

[0124] Based on the set association logic, determine the association degree between the content feature classes corresponding to the target features of the two fused features of each fused feature group, and record it as the association degree of the fused feature group;

[0125] Based on the correlation of all fused feature groups, the feature correlation of the fused feature set is calculated using the feature correlation comprehensive formula;

[0126] Among them, the comprehensive formula of feature correlation is as follows:

[0127]

[0128] In the formula, g k is the correlation degree of the kth adjacent fusion feature group, g uis the correlation degree of the u-th interval fusion feature group, K is the total number of adjacent fusion feature groups, and U is the total number of interval fusion feature groups.

[0129] The distance between the fused features of adjacent fused feature groups is close, while the distance between the fused features of interval fused feature groups is far. Therefore, when calculating the feature correlation, this scheme assigns a weight of 2 to the adjacent fused feature group and a weight of 1 to the interval fused feature group.

[0130] Among them, the association logic can be set in advance by the staff based on the mutual correlation between each content feature class. For example, if the probability of A and B appearing in the same scene at the same time is low, then the mutual correlation between feature class A and feature class B is low. If the probability of A and C appearing in the same scene at the same time is high, then the mutual correlation between feature class A and feature class C is high. In this way, the mutual correlation between all content feature classes is set in advance according to the probability.

[0131] The feature fusion content recognition of the working data archive image is performed through this scheme, which not only takes into account the similarity between the fused features and the actual content features, but also comprehensively considers the rationality of each fused feature content appearing in the same working file. Through comprehensive consideration, the feature analysis depth of the working data archive image is greatly improved, which can effectively ensure the overall matching of the working data archive image when performing feature area division, thereby ensuring high-precision matching between content recognition and features of the working data archive image.

[0132] In summary, the advantages of the present invention are: ensuring high-precision matching between content recognition and features of working data archive images, and improving the accuracy of content recognition of working data archive images.

[0133] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited by the above examples. The above examples and descriptions only describe the principles of the present invention. The present invention may be subject to various changes and improvements without departing from the spirit and scope of the present invention. The requirements of these changes and improvements should be within the scope of protection of the present invention. The scope of protection claimed by the present invention is defined by the attached claims and their equivalents.

Claims

1. A comprehensive and integrated work data file entry system, characterized in that: include: A storage module, wherein the storage module is used to store a content database; A working data archive image acquisition module, wherein the working data archive image acquisition module is used to acquire working data archive images; A feature splitting module, the feature splitting module is electrically connected to the working data archive image acquisition module, and the feature splitting module is used to perform content splitting and fusion on the working data archive image according to the content splitting algorithm to obtain a plurality of fusion feature sets; A matching calculation module, the matching calculation module is electrically connected to the storage module and the feature separation module, and the matching calculation module is used to calculate the content matching degree of each fused feature set according to a matching degree algorithm; A screening module, the screening module is electrically connected to the matching calculation module, and the screening module is used to screen out a fusion feature set corresponding to a maximum content matching degree, and use the fusion feature set as an image region division feature group; A region division module is electrically connected to the screening module, and is used to divide the working data archive image into content regions based on the working data archive image region corresponding to each fusion feature in the image region division feature group.

2. A work data archive integrated input system according to claim 1, characterized in that: The feature splitting module comprises: A feature extraction unit, the feature extraction unit is used to obtain a total number M of features with the same content in the working data archive image; A fusion condition unit, wherein the splitting condition unit is used to generate feature fusion restriction conditions; The feature fusion unit is used to generate a plurality of feature fusion sets under the restriction of feature fusion restriction conditions, and extract a plurality of content features with the same boundaries from the working data archive image according to the numbers in the feature fusion array, and combine them into a fusion feature set.

3. A comprehensive integrated work data file entry system according to claim 2, characterized in that: The expression of the feature fusion restriction condition is: In the formula, z i is the number of identical content features included in the i-th fusion feature, and N+ is a positive integer set.

4. A comprehensive integrated work data file entry system according to claim 3, characterized in that: The method of obtaining the total number M of identical content features in the working data archive image specifically includes: The original archive image is gray-scaled using the weighted average method to obtain a gray-scale image; A grayscale similarity deviation threshold is set, and continuous intervals in the grayscale image of the working data whose grayscale deviation is less than the grayscale similarity deviation threshold are divided into the same content features; Count the total number of features with the same content in the grayscale image of the working data, denoted as M; The weighted average formula is: Gray(i,j)=α1R(i,j)+α2G(i,j)+α3B(i,j), Where Gray(i,j) is the grayscale of the pixel in the i-th row and j-th column in the working data archive image, α1, α2, and α3 are the red, green, and blue sensitivity weighting coefficients, respectively, and R(i,j), G(i,j), and B(i,j) are the red, green, and blue components of the pixel in the i-th row and j-th column, respectively.

5. A comprehensive integrated work data file entry system according to claim 4, characterized in that: The extracting of a plurality of features with the same content and the same boundaries from the working data archive image according to the numbers in the feature fusion array and combining them into a fusion feature set specifically includes: S1, select an element from the feature fusion set, denoted as Z; S2, selecting a same content feature from the working data archive image as a reference content feature; S3, set a fusion index j, and add an initial value of 1 to j; S4, determine whether j is equal to Z, if so, use the reference content feature as the fusion feature and proceed to step S6, if not, add one to the value of j and proceed to step S5; S5, selecting the same content feature that has the same boundary as the reference content feature, fusing it with the reference content feature to form a new reference content feature, and returning to step S4; S6, determine whether to traverse all elements in the feature fusion set, if yes, proceed to S7, if no, remove Z from the feature fusion set and return to step S1; S7, combining all fused features into a fused feature set; S8. Repeat steps S1-S7 until a new fused feature set cannot be generated under the feature fusion set.

6. A work data archive integrated input system according to claim 5, characterized in that: The matching calculation module includes: A similarity fitting unit, wherein the similarity fitting unit is used to calculate the similarity between each fusion feature in the fusion feature set and each content feature in the content database, and select the content feature with the greatest similarity as the target feature of the fusion feature; An association calculation unit, the association calculation unit is used to calculate the feature association degree based on the target feature of each fused feature in the fused feature set, and obtain the feature association degree of the fused feature set; The matching calculation unit is used to calculate the content matching degree of the fused feature set through a matching degree calculation formula based on the similarity between each fused feature in the fused feature set and its target feature and the feature association degree of the fused feature set.

7. A work data archive integrated input system according to claim 6, characterized in that: The matching degree calculation formula is specifically: In the formula, P is the content matching degree of the fusion feature set, C hi is the similarity between the i-th fusion feature in the fusion feature set and its target feature, G l is the feature correlation of the fused feature set.

8. A comprehensive integrated work data file entry system according to claim 7, characterized in that: The calculating of the feature correlation based on the target feature of each fused feature in the fused feature set to obtain the feature correlation of the fused feature set specifically includes: Determine the content feature class corresponding to the target feature of each fused feature; The fusion features in the fusion feature set are combined into several fusion feature groups in pairs, and the fusion feature groups composed of fusion features with the same boundary are recorded as adjacent fusion feature groups, and the fusion feature groups composed of fusion features with different boundaries are recorded as interval fusion feature groups; Based on the set association logic, determine the association degree between the content feature classes corresponding to the target features of the two fused features of each fused feature group, and record it as the association degree of the fused feature group; Based on the correlation of all fused feature groups, the feature correlation of the fused feature set is calculated using the feature correlation comprehensive formula; The feature correlation comprehensive formula is specifically: In the formula, g k is the correlation degree of the kth adjacent fusion feature group, g u is the correlation degree of the u-th interval fusion feature group, K is the total number of adjacent fusion feature groups, and U is the total number of interval fusion feature groups.

Images