CN114443875A - File merging method and device, electronic equipment and storage medium - Google Patents

Abstract

The invention discloses a method and a device for merging files, electronic equipment and a storage medium, wherein the method for merging the files comprises the following steps: carrying out similarity calculation on any two archived data in the archived data sets to obtain a similarity set, wherein the archived data set is an archived data set consisting of a plurality of archived data of the same person; determining a density value set among a plurality of archived data according to the similarity set; calculating a distance value set of a plurality of archived data according to the density value set and the similarity set; determining a plurality of archived data corresponding to the density values and the distance values meeting preset conditions as a plurality of clustering center points; and taking the filing data corresponding to the plurality of cluster center points as cover images. The method avoids the generated cover image from being high in randomness, and improves the accuracy and the representativeness of the cover image.

Description

File merging method, device, electronic device and storage medium

technical field

The present invention relates to the technical field of data processing, in particular to a file merging method, device, electronic device and storage medium.

Background technique

At present, the combination of surveillance cameras and face recognition technology monitors personnel in various public places, which increases the security of the city. While monitoring personnel in public places, a personnel file can be established for each person to facilitate the monitoring of personnel. management. At present, the quality of the snapshot data is mainly stratified, and the pictures with better snapshot quality are compared with the existing archive covers. If they are not similar, they are used as the cover of the new archive. However, the generated archive covers are more random, which leads to accuracy. and less representative.

SUMMARY OF THE INVENTION

In the first aspect, the main purpose of the present invention is to provide a file merging method, including:

Perform similarity calculation on any two archived data in the archived data set to obtain a similarity set, the archived data set is an archived data set formed by multiple archived data of the same person;

According to the similarity set, determine a set of density values between a plurality of the archived data;

Calculate a plurality of distance value sets of the archived data according to the density value set and the similarity set;

Determining multiple archived data corresponding to density values and distance values that satisfy preset conditions as multiple cluster center points;

The archived data corresponding to the plurality of cluster center points is used as a cover image.

Optionally, the determining a set of density values between multiple archived data according to the set of similarity includes:

determining the similarity threshold of each of the archived data in the similarity set;

Screening the similarity set according to the similarity threshold of each archived data;

The filtered similarity set is calculated according to a predetermined algorithm to obtain the density value set.

Optionally, calculating the filtered similarity set according to a predetermined algorithm to obtain the density value set includes:

Determine the corresponding similarity according to the filtered similarity set;

Summing the similarity to obtain a plurality of density values corresponding to each archived data;

Sorting the density values corresponding to a plurality of the archived data to obtain the density value set.

Optionally, the calculating a plurality of distance value sets of the archived data according to the density value set and the similarity set includes:

Sort the archived data set according to the density value set of each archived data, and filter the sorted archived data set;

Perform distance calculation on the archived data set after screening to obtain the distance value set.

Optionally, sorting the archived data set according to the density value set of each archived data, and screening the sorted archived data set includes:

Sorting the archived data sets from large to small according to the density value of each archived data;

For each of the archived data,

determining a plurality of reference archived data whose density value is greater than the density value of the archived data itself;

The target archived data with the smallest similarity with the archived data is determined from the plurality of the reference archived data, so as to perform distance calculation according to the target archived data and the archived data.

Optionally, the method further includes:

According to the plurality of cluster center points, the archived data sets are merged to obtain a cluster set; wherein, each of the clusters includes a corresponding cover image;

Determine the cover image of each cluster according to the cluster set;

When a plurality of the clusters and the cover image satisfy a preset relationship, the plurality of clusters are merged.

Optionally, when the cluster set and the cover image satisfy a preset relationship, merging a plurality of the clusters includes:

Judging whether the number of the plurality of clusters is greater than the number of the cover images;

When the number of the plurality of clusters is greater than the number of the cover images, determining the similarity between the cluster center points of the two clusters;

In the case that the similarity of the two-by-two cluster center points is greater than the preset similarity, the corresponding two-by-two clusters are merged.

In a second aspect, an embodiment of the present invention provides a file merging device, including:

a first calculation module, configured to perform similarity calculation on any two archived data in the archived data set to obtain a similarity set, where the archived data set is an archived data set composed of multiple archived data of the same person;

a first determining module, configured to determine a set of density values between a plurality of the archived data according to the set of similarity;

a second calculation module, configured to calculate a plurality of distance value sets of the archived data according to the density value set and the similarity set;

a second determination module, configured to determine a plurality of archived data corresponding to the density value and the distance value that meet the preset conditions as a plurality of cluster center points;

The third determining module is configured to use the archived data corresponding to the plurality of cluster center points as the cover image.

In a third aspect, an embodiment of the present invention provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, when the processor executes the computer program Implement the steps of the file merging method as described above.

In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium, where the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, implements the steps of the above method for merging files.

The above-mentioned scheme of the present invention at least includes the following beneficial effects:

The archive merging method provided by the present invention firstly performs similarity calculation on any two archived data in the archived data set to obtain a similarity set, and the archived data set is an archived data set composed of multiple archived data of the same person; Set, determine the set of density values between multiple archived data; calculate the set of distance values of multiple archived data according to the set of density values and the set of similarity; It is determined as multiple cluster center points; the archived data corresponding to multiple cluster center points is used as the cover image. This avoids the generated cover image from being very random, and improves the accuracy and representativeness of the cover image.

Description of drawings

In order to explain the embodiments of the present invention or the technical solutions in the prior art more clearly, the following briefly introduces the accompanying drawings that need to be used in the description of the embodiments or the prior art. Obviously, the accompanying drawings in the following description are only These are some embodiments of the present invention, and for those of ordinary skill in the art, other drawings can also be obtained according to the structures shown in these drawings without creative efforts.

1 is a schematic overall flow diagram of a file merging method provided by an embodiment of the present invention;

FIG. 2 is a schematic flowchart of step S20 provided by an embodiment of the present invention;

FIG. 3 is a schematic flowchart of a specific flow of step S23 provided by an embodiment of the present invention;

FIG. 4 is a schematic flowchart of a specific flow of step S30 provided by an embodiment of the present invention;

FIG. 5 is another schematic flowchart of step S32 provided by an embodiment of the present invention;

6 is another schematic flowchart of a file merging method provided by an embodiment of the present invention;

FIG. 7 is another schematic flowchart of step 32 provided by an embodiment of the present invention;

8 is a structural block diagram of a file merging apparatus provided by an embodiment of the present invention;

FIG. 9 is a structural block diagram of an electronic device provided by an embodiment of the present invention.

The realization, functional characteristics and advantages of the present invention will be further described with reference to the accompanying drawings in conjunction with the embodiments.

Detailed ways

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

The terms "first", "second" and "third" in the description and claims of the present invention and the above drawings are used to distinguish different objects, rather than to describe a specific order. Furthermore, the term "comprising" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device comprising a series of steps or units is not limited to the listed steps or units, but optionally also includes unlisted steps or units, or optionally also includes For other steps or units inherent to these processes, methods, products or devices.

First, the solutions of the embodiments of the present application are introduced by way of example with reference to the relevant drawings.

As shown in FIG. 1, a specific embodiment of the present invention provides a file merging method, including:

S10. Perform similarity calculation on any two archived data in the archived data set to obtain a similarity set; the archived data set is an archived data set composed of multiple archived data of the same person.

In this embodiment, an archive data set is stored in the database, the archive data set includes an archive data set, and the archive data set may include multiple image archives. Of course, the archive data may also be archive data of other attributes; the archive data may be in Image files collected in public places, such as city streets, high-speed railway stations, bus stations, airports and other places, the database can be the database of the police platform. It can be more convenient to manage various groups of people.

In addition, when the similarity calculation is performed on the archived data set, the cosine theorem can be used to calculate the similarity between the two archived data. The cosine theorem is also called the cosine similarity. A measure of the magnitude of the difference between two individuals, the closer the cosine value is to 1, the closer the included angle is to 0 degrees, that is, the more similar the two vectors are; that is, the closer the cosine value between the two archived data is. When it is close to 1, it means that the similarity between the two archived data is greater.

The archive data set may be aidN{aid1, aid2, aid3...aidN}, and the archive data set in each archive data set aidN may _be an {a ₁ , a ₂ , a ₃ , .. a _n } , the similarity between the two archived data is calculated by performing n× _n on the archived data set an in the archived data set aidN. Therefore, the calculated similarity set can be sim _ij {sim ₁₂ ,sim ₁₃ ... .sim _ij }, i, j represent data a _i , a _j ; for example, the archived data set includes picture a, picture b, picture c, and picture d. , after the similarity of picture d, the obtained similarity set is sim _ab , sim _ac , sim _ad , sim _bc , sim _bd , and sim _cd ; thus, the picture with the largest similarity in the archived data set can be determined.

S20. Determine, according to the similarity set, a set of density values between multiple archived data.

In this embodiment, the density value is expressed as the similarity density of the archived data set, and the density value set contains multiple density values; each archived data may represent corresponding multiple pieces of image data, and each piece of image data may be used as a data point , by calculating the density value between each data point corresponding to each cluster, the density value set can be determined; generally, the similarity between each archived data is relatively small, but the The similarity between data points is relatively large.

For example, after archiving the image file of A, the image file of A contains image a, image b, image c, and image d, and after archiving the image file of B, the image file of B contains image e, image f. Picture g and picture h, after the clustering is merged, the image file of A can be used as a cluster, and the image file of B can be used as a cluster, so the archived data corresponding to A can be obtained through picture a, picture b, picture c, picture The similarity of d determines multiple density values, and the archived data corresponding to B can also determine multiple density values through picture e, picture f, picture g, and picture h.

As shown in Figure 2, the above-mentioned determination of the density value set between multiple archived data according to the similarity set includes:

S21, determine the similarity threshold of each archived data in the similarity set;

S22, screening the similarity set according to the similarity threshold of each archived data;

S23. Calculate the filtered similarity set according to a predetermined algorithm to obtain a density value set.

In this embodiment, the similarity threshold may be preset, and through the similarity threshold of each archived data, multiple similarities of each archived data may be screened, and after screening, each archived data The multiple similarities are calculated, thereby determining the density value set of each archived data.

通过相似度阈值

对多个相似度进行筛选，筛选条件可以是

也就是说，相似度小于等于

的可以滤除，而大于

的相似度可以保留，由此，可以得到筛选后的相似度集合以计算密度值集合。Among them, in the similarity set sim _ij {sim ₁₂ ,sim ₁₃ ....sim _ij } of the file data aidN obtained by the above calculation, the similarity threshold can be determined from the similarity set, and the similarity threshold is set for

pass similarity threshold

Filter multiple similarities, the filter conditions can be

That is, the similarity is less than or equal to

can be filtered out, while greater than

The similarity can be retained, and thus, the filtered similarity set can be obtained to calculate the density value set.

As shown in Figure 3, the above-mentioned calculation of the filtered similarity set according to the predetermined algorithm to obtain the density value set includes:

S231. Determine the corresponding similarity according to the filtered similarity set;

S232, the similarity is summed to obtain a plurality of density values corresponding to each archived data;

S233. Sort the density values corresponding to the multiple archived data to obtain a density value set.

In this embodiment, the determined similarity is summed, and the summation result is used as the density value of the archived data, and after the density value of each archived data is determined, the multiple density values are sorted from large to small. Sorting is performed to obtain a set of density values, which can be expressed as β(sim _ij )∈{β ₁ ,β ₂ ,...β _n }.

Specifically, the predetermined algorithm can be calculated using the following formula:

为最小相似度，x表示相似度数量，因此，上述公式表示在

时，则x＝1，在小于0时，则x＝0，通过将确定的相似度进行求和，由此可以确定出每个归档数据对应的密度值集合。Among them, β _i is the density value, sim _ij is the similarity,

is the minimum similarity, x represents the number of similarity, therefore, the above formula expresses in

When it is less than 0, then x=1, and when it is less than 0, then x=0. By summing the determined similarities, the set of density values corresponding to each archived data can be determined.

For example, the archived data set includes the image files of three persons, A, B, and C. The image file of A includes picture a, picture b, picture c, and picture d, and the image file of B includes picture e, picture f, Picture g, picture h, picture i, after calculation, the similarity set corresponding to the image file of A is sim _ab , sim _ac , sim _ad , sim _bc , sim _bd , sim _cd ; the similarity set sim corresponding to the image file of B _ef , sim _eg , sim _eh , sim _ei , sim _fg , sim _fh , sim _fi , sim _gh , sim _gi , sim _hi ; when calculating the density value of A, when the similarity of A is summed, the calculated density value is 4; when calculating the density value of B, when the similarity of B is summed up to obtain a density value of 9; after determining the density values corresponding to A and B, sort 9 and 4 from large to small, The resulting set of density values is β(sim _AB )∈{9,4}.

S30. Calculate distance value sets of multiple archived data according to the density value set and the similarity set.

As shown in FIG. 4 , the specific implementation manner of the above step S30 includes:

S31. Sort the archived data set according to the density value set of each archived data, and filter the sorted archived data set;

S32. Perform distance calculation on the filtered archived data set to obtain a distance value set.

In this embodiment, the distance value can be expressed as the distance between archived data, that is, the distance value between each archived data; when sorting the archived data set, it can be sorted by the density value set, and the density value set is Sort from large to small, so the archived data sets can be sorted according to the density value. After the sorting is completed, each archived data can be calculated by distance, and the density values of multiple archived data can be compared recursively, and then the most The distance between the archived data; for example, the set of density values obtained from the four archived data of A, B, C, and D is (600, 500, 400, 300), so B and A can be recursively compared, C performs recursive comparison with A and B, and D performs recursive comparison with A, B, and C, thereby determining the distances between multiple archived data.

As shown in Figure 5, the above-mentioned sorting of the archived data sets according to the density value set of each archived data, and screening of the sorted archived data sets includes:

S311, sorting the archived data sets from large to small according to the density value of each archived data;

S312, for each archived data, determine a plurality of reference archived data whose density value is greater than the density value of the archived data itself;

S313: Determine the target archived data with the smallest similarity with the archived data from the multiple reference archived data, so as to perform distance calculation according to the target archived data and the archived data.

Wherein, when the archived data is recursively compared, the reference archived data whose density value is greater than itself can be recursively compared; optionally, the archived data corresponding to the maximum density value in the density value set can be calculated by calculating the maximum distance from the archived data. The distance value between the target archived data; therefore, after determining the reference archived data, it can be judged whether the similarity between the archived data and its corresponding target archived data is the smallest, if the similarity between the two is the smallest, then It can be determined as the target archived data, and the distance between the two is calculated through the target archived data and the archived data, and then the distance value is determined. After the distance value of all the archived data is calculated, the distance value set can be determined; , the distance value can be calculated using the following formula:

Among them, γ _i represents the distance value. After the density value set and the distance value set are determined, the density value set and the distance value set can be represented by two-dimensional vector coordinates. For example, the x-axis is the density value, and the y-axis is the distance value. Therefore, in the two-dimensional vector coordinates, it can be judged whether the value of each density value and distance value is relatively large, and then the corresponding cluster center point can be determined; that is, the density value and distance value corresponding to the archived data are compared. When it is large, the archived data can be determined as the cluster center point.

For example, the set of density values obtained from the four archived data of A, B, C, and D is (600, 500, 400, 300). Therefore, B can be compared with A, C can be compared with A, B, D Compare with A, B, and C, and the calculated set of distance values is (0.9, 0.4, 0.7, 0.2), that is, the maximum distance from A is 0.9, the distance between B and A is 0.4, and C The distance with the smallest similarity calculated with A and B respectively is 0.7, and the distance with the smallest similarity calculated by D and A, B and C respectively is 0.2.

S40. Determine the multiple archived data corresponding to the density value and the distance value that satisfy the preset condition as multiple cluster center points.

In this embodiment, the preset condition indicates that the density value and the distance value are relatively large, that is to say, the density value of each file data is relatively large, but the distance value between it and other file data is relatively large, so , and the corresponding archived data is determined as the cluster center point. When the archived data is re-clustered and archived, the cluster center point can be used for archiving more accurately to improve the accuracy.

S50. Use the archived data corresponding to the plurality of cluster center points as a cover image.

In this embodiment, the cover image may be a face image, and each archived data may contain multiple image files. After clustering and merging the cluster centers obtained by the above calculation, the image files of each person can be merged. Form a cluster or multiple clusters. Since the image files corresponding to each cluster center point are denser and farther away from other cluster center points, it can be used as a representative cover image. During subsequent filing, the cover image can be used to more accurately cluster and merge different image files to improve accuracy.

The archive merging method provided by the present invention firstly performs similarity calculation on any two archived data in the archived data set to obtain a similarity set, and the archived data set is an archived data set composed of multiple archived data of the same person; Set, determine the set of density values between multiple archived data; calculate the set of distance values of multiple archived data according to the set of density values and the set of similarity; It is determined as multiple cluster center points; the archived data corresponding to multiple cluster center points is used as the cover image. This avoids the generated cover image from being very random, and improves the accuracy and representativeness of the cover image.

As shown in FIG. 6 , the file merging method provided by the embodiment of the present invention further includes:

30. Re-merge the archived data sets according to multiple cluster center points to obtain a cluster cluster set; wherein, each cluster cluster includes a corresponding cover image;

31. Determine the cover image of each cluster according to the cluster set;

32. When the multiple clusters and the cover image satisfy the preset relationship, merge the multiple clusters.

In this embodiment, the preset relationship indicates that the number of clusters is greater than the number of cover images, each cluster may be represented as an image file of a certain person, and an image file of one person may also form multiple clusters, Multiple cover images can be determined in multiple clusters. In order to reduce the number of clusters corresponding to the same file, multiple clusters can be merged, and the corresponding cover images can be re-determined after merging ; For example, the image file of A has 10 clusters, and there are only 5 cover images in the image file of A. By merging the 10 clusters to form 5 clusters, the image file of A forms a cluster Fewer clusters.

As shown in FIG. 7 , when the cluster set and the cover image meet the preset relationship, the multiple clusters are merged to determine the corresponding cover image, including:

321. Determine whether the number of multiple clusters is greater than the number of cover images;

322. When the number of multiple clusters is greater than the number of cover images, determine the similarity between the cluster center points of the two clusters;

323. In the case that the similarity between the center points of the pairwise clusters is greater than the preset similarity, merge the corresponding pairwise clusters.

In this embodiment, the preset similarity may be the similarity preset by the user. When merging multiple clusters, the similarity between two clusters may be judged, and if the similarity is greater than the predetermined similarity When the similarity is set, the two clusters can be merged; it can be understood that when judging whether the similarity between the two clusters is greater than the preset similarity through the above-determined cluster center points, the The similarity calculation can be performed by the cluster center points corresponding to the pairwise clusters, wherein the preset similarity can be set to o _sim , and the similarity of the pairwise cluster center points can be sim _ij . When the similarity between the two clusters is greater than the preset similarity, it is calculated by sim _ij -o _sim > 0 to determine whether the similarity between the two clusters is greater than the preset similarity to determine whether When performing cluster merging, it can be understood that when the number of clusters is less than the number of cover images, the merging operation may not be performed between the clusters.

For example, A's image file includes cluster A, cluster B, cluster C, and cluster D, and the cover image corresponding to A includes face image e and face image f; B's image The file includes cluster G, cluster H, and cluster I, and the cover image corresponding to B includes face picture 1, face picture 2, face picture 3, and face picture 4; therefore, the corresponding cover image of A If the number of clusters is greater than the number of cover images, cluster A, cluster B, cluster C, and cluster D can be merged, and the distance between the cluster center points of each cluster can be calculated by calculating The pairwise similarities of the The number of clusters corresponding to B is less than the number of its cover images. Therefore, the clusters of B may not be merged, and when more clusters are formed subsequently, the clusters can be merged again.

As shown in FIG. 8, an embodiment of the present invention provides a file merging apparatus 10, including:

The first calculation module 11 is used to perform similarity calculation on any two archived data in the archived data set to obtain a similarity set, and the archived data set is an archived data set composed of multiple archived data of the same person;

a first determining module 12, configured to determine a set of density values between a plurality of archived data according to the set of similarity;

The second calculation module 13 is used to calculate the distance value sets of the multiple archived data according to the density value set and the similarity set;

The second determination module 14 is configured to determine a plurality of archived data corresponding to the density value and the distance value satisfying the preset condition as a plurality of cluster center points;

The third determining module 15 is configured to use the archived data corresponding to the plurality of cluster center points as the cover image.

The archive merging device 10 provided by the present invention firstly performs similarity calculation on any two archived data in the archived data set to obtain a similarity set, and the archived data set is an archived data set composed of multiple archived data of the same person; degree set, determine the density value set between multiple archived data; calculate the distance value set of multiple archived data according to the density value set and similarity set; The data is determined as multiple cluster center points; the archived data corresponding to the multiple cluster center points is used as the cover image. This avoids the generated cover image from being very random, and improves the accuracy and representativeness of the cover image.

It should be noted that the file merging apparatus 10 provided by the specific embodiment of the present invention is an apparatus corresponding to the above-mentioned file merging method, and all the above-mentioned file merging methods are applicable to the file merging apparatus 10. The above-mentioned file merging apparatus 10 embodiments There are corresponding modules corresponding to the steps in the above-mentioned file merging method, which can achieve the same or similar beneficial effects. To avoid excessive repetition, each module in the file merging apparatus 2 is not described here.

As shown in FIG. 9 , a specific embodiment of the present invention further provides an electronic device 20 , including a memory 202 , a processor 201 and a computer program stored in the memory 202 and running on the processor 201 , the processor 201 The steps of the above-mentioned file merging method are realized when the computer program is executed.

Specifically, the processor 201 is configured to call the computer program stored in the memory 202, and perform the following steps:

Calculate the similarity of any two archived data in the archived data set to obtain a similarity set, and the archived data set is an archived data set composed of multiple archived data of the same person;

According to the similarity set, determine the density value set between multiple archived data;

Calculate the distance value set of multiple archived data according to the density value set and the similarity set;

Determining multiple archived data corresponding to density values and distance values that satisfy preset conditions as multiple cluster center points;

The archived data corresponding to multiple cluster center points are used as the cover image.

Optionally, the processor 201 executes

Determining the set of density values between multiple archived data according to the set of similarity includes:

Determine the similarity threshold for each archived data in the similarity set;

Filter the similarity set according to the similarity threshold of each archived data;

The filtered similarity set is calculated according to a predetermined algorithm to obtain a density value set.

Optionally, the calculation performed by the processor 201 to calculate the filtered similarity set according to a predetermined algorithm to obtain a density value set includes:

Determine the corresponding similarity according to the filtered similarity set;

Sum the similarity to get multiple density values corresponding to each archived data;

Sort the density values corresponding to multiple archived data to obtain a density value set.

Optionally, calculating the distance value set of multiple archived data according to the density value set and the similarity set executed by the processor 201 includes:

Sort the archived datasets according to the density value set of each archived data, and filter the sorted archived datasets;

Perform distance calculation on the filtered archive data set to obtain a set of distance values.

Optionally, the processing performed by the processor 201 to sort the archived data set according to the density value set of each archived data, and to filter the sorted archived data set includes:

Sort archived datasets from largest to smallest according to the density value of each archived data;

For each archived data, determine a plurality of reference archived data whose density value is greater than the density value of the archived data itself;

The target archived data with the smallest similarity with the archived data is determined among the plurality of reference archived data, so as to perform distance calculation according to the target archived data and the archived data.

Optionally, the method performed by the processor 201 further includes:

According to a plurality of cluster center points, the archived data sets are re-merged to obtain a cluster cluster set; wherein each cluster cluster includes a corresponding cover image;

Determine the cover image of each cluster according to the cluster set;

When the multiple clusters and the cover image satisfy the preset relationship, the multiple clusters are merged.

Optionally, when the cluster set and the cover image meet a preset relationship, the execution of the processor 201 to merge multiple clusters includes:

Determine whether the number of multiple clusters is greater than the number of cover images;

When the number of multiple clusters is greater than the number of cover images, determine the similarity between the cluster center points of the two clusters;

When the similarity of the center points of the pairwise clusters is greater than the preset similarity, the corresponding pairwise clusters are merged.

That is, in the specific embodiment of the present invention, when the processor 201 of the electronic device 20 executes the computer program, it implements the steps of the above-mentioned file merging method, thereby avoiding the generated cover image with high randomness and improving the accuracy of the cover image. and representation.

It should be noted that, since the processor 201 of the electronic device 20 executes the computer program to implement the steps of the above-mentioned file merging method, all embodiments of the above-mentioned file merging method are applicable to the electronic device 20, and can achieve the same or similar beneficial effect.

In the computer-readable storage medium provided in the embodiment of the present invention, a computer program is stored on the computer-readable storage medium, and when the computer program is executed by the processor, each of the file merging method or the application-side file merging method provided by the embodiment of the present invention is implemented. process, and can achieve the same technical effect, in order to avoid repetition, it will not be repeated here.

Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be implemented by instructing relevant hardware through a computer program, and the program can be stored in a computer-readable storage medium. During execution, the processes of the embodiments of the above-mentioned methods may be included. The storage medium may be a magnetic disk, an optical disk, a read-only memory (Read-Only Memory, ROM), or a random access memory (Random Access Memory, RAM for short).

In the description of this specification, description with reference to the terms "one embodiment," "some embodiments," "example," "specific example," or "some examples", etc., mean specific features described in connection with the embodiment or example , structure, material or feature is included in at least one embodiment or example of the present invention. In this specification, schematic representations of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the particular features, structures, materials or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.

The above descriptions are only preferred embodiments of the present invention, and are not intended to limit the scope of the present invention. Under the conception of the present invention, the equivalent structural transformations made by the contents of the description and accompanying drawings of the present invention, or directly/indirectly applied in Other related technical fields are included within the scope of patent protection of the present invention.

Claims (10)

1. A file merging method is characterized by comprising the following steps:

carrying out similarity calculation on any two archived data in the archived data sets to obtain a similarity set, wherein the archived data set is an archived data set consisting of a plurality of archived data of the same person;

determining a density value set among the plurality of archived data according to the similarity set;

calculating a plurality of sets of distance values of the archived data according to the sets of density values and the sets of similarity;

determining a plurality of archived data corresponding to the density values and the distance values meeting preset conditions as a plurality of clustering center points;

and taking the filing data corresponding to the plurality of cluster center points as cover images.

2. The archive merging method according to claim 1, wherein the determining a set of density values among the plurality of archived data according to the set of similarities comprises:

determining a similarity threshold value of each archived data in the similarity set;

screening the similarity set according to the similarity threshold value of each archived data;

and calculating the screened similarity set according to a preset algorithm to obtain the density value set.

3. The archive merging method according to claim 2, wherein the calculating the filtered similarity sets according to a predetermined algorithm to obtain the density value sets comprises:

determining corresponding similarity according to the screened similarity set;

summing the similarity to obtain a plurality of density values corresponding to each archived data;

and sorting the density values corresponding to the plurality of archived data to obtain the density value set.

4. The archive merging method of claim 1, wherein said calculating a set of distance values for a plurality of the archived data from the set of density values and the set of similarities comprises:

sorting the archived data sets according to the density value set of each archived data, and screening the sorted archived data sets;

and performing distance calculation on the screened archival data set to obtain the distance value set.

5. The archive merging method of claim 4, wherein the sorting the archive data sets according to the density value set of each archive data and the screening the sorted archive data sets comprises:

sorting the archived data sets from large to small according to the density value of each archived data;

determining a plurality of reference filing data of which the density values are greater than the density values of the filing data for each filing data;

and determining target archival data with the minimum similarity with the archival data from the plurality of reference archival data, and performing distance calculation according to the target archival data and the archival data.

6. The archive merging method according to claim 1, further comprising:

merging the archived data sets according to the plurality of clustering central points to obtain a clustering set; wherein each cluster comprises a corresponding cover image;

determining a cover image of each cluster according to the cluster set;

and when the plurality of clustering clusters and the cover image meet a preset relation, merging the clustering clusters.

7. The archive merging method according to claim 6, wherein the merging the plurality of clusters when the cluster set and the cover image satisfy a preset relationship comprises:

judging whether the number of the plurality of clustering clusters is larger than that of the cover images or not;

when the number of the plurality of clustering clusters is larger than that of the cover images, determining the similarity between the clustering center points of every two clustering clusters;

and combining the corresponding pairwise clustering clusters under the condition that the similarity of the pairwise clustering central points is greater than the preset similarity.

8. An archive merging apparatus, comprising:

the first calculation module is used for carrying out similarity calculation on any two archival data in the archival data sets to obtain a similarity set, wherein the archival data set is an archival data set consisting of a plurality of archival data of the same person;

a first determining module, configured to determine a density value set among the plurality of archived data according to the similarity set;

the second calculation module is used for calculating a plurality of distance value sets of the archived data according to the density value set and the similarity set;

the second determining module is used for determining a plurality of archived data corresponding to the density values and the distance values meeting the preset conditions as a plurality of clustering center points;

and the third determining module is used for taking the filing data corresponding to the plurality of clustering center points as a cover image.

9. An electronic device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, characterized in that the processor implements the steps of the archive merging method according to any of claims 1 to 7 when executing the computer program.

10. A computer-readable storage medium, in which a computer program is stored which, when being executed by a processor, carries out the steps of the archive merging method according to any one of claims 1 to 7.

Images