Pet information aggregation method based on pet noseprint and related device
By acquiring pet nose print images and performing feature segmentation and clustering, the accuracy problem in pet record management was solved, achieving efficient and accurate pet record clustering.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- NEW RUIPENG PET HEALTHCARE GRP CO LTD
- Filing Date
- 2022-05-23
- Publication Date
- 2026-07-24
AI Technical Summary
Current technologies present challenges in pet identification and management, especially in pet clinics, veterinary hospitals, and pet grooming salons. Improving the accuracy of pet record management is a pressing issue that needs to be addressed.
By acquiring images of a pet's nose print, extracting nose print features, and segmenting them into multiple nose print sub-features, clustering algorithms are used to match and cluster in the subspace to determine the target nose print features, thereby accurately clustering the pet's file into the target file.
It improves the accuracy of pet record retrieval, enhances the accuracy of pet information aggregation, and simplifies pet identity management.
Smart Images
Figure CN115185881B_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of document processing technology, and in particular to a method and related equipment for pet information aggregation based on pet nose prints. Background Technology
[0002] In recent years, more and more people have chosen to keep pets. With the increasing number of pets, pet clinics, veterinary hospitals, and pet grooming salons are facing difficulties in managing pet identification. To effectively solve this problem, some researchers have attempted to transfer methods of human identification to pet identification, thereby creating files for each pet and implementing a one-file-per-pet management model. File aggregation (archiving) is a particularly important part of record information management, and how to improve the accuracy of file aggregation is a pressing issue that pet record management needs to address. Summary of the Invention
[0003] To address the aforementioned issues, this application provides a method and related equipment for pet information aggregation based on pet nose prints, which helps improve the accuracy of pet information aggregation.
[0004] To achieve the above objectives, the first aspect of this application provides a method for pet information archiving based on pet nose prints, the method comprising:
[0005] Obtain the pet's data to be aggregated, which includes the pet's nose print image;
[0006] Based on the nose print image, obtain the nose print characteristics of the pet;
[0007] The nasal print features are segmented to obtain m nasal print sub-features, where m is an integer greater than 1;
[0008] Cluster the m nasal pattern features separately to determine the target nasal pattern feature from the multiple clustered nasal pattern features in the clustering space;
[0009] The files associated with the target nasal print features are identified as target files, and the data to be aggregated is aggregated into the target files.
[0010] As can be seen, this embodiment of the application obtains pet data to be aggregated, including the pet's nose print image; based on the nose print image, it obtains the pet's nose print features; it segments the nose print features to obtain m nose print sub-features, where m is an integer greater than 1; it clusters the m nose print sub-features to determine the target nose print feature from multiple clustered nose print features in the clustering space; it determines the file associated with the target nose print feature as the target file, and aggregates the data to be aggregated into the target file. In this way, based on the nose print sub-features, it determines the corresponding clustered nose print features (i.e., the target nose print features) of the pet, and determines the file associated with the clustered nose print features as the target file of the pet, improving the accuracy of finding the pet's target file, thereby improving the accuracy of pet information aggregation.
[0011] In conjunction with the first aspect, in one possible implementation, the clustering space includes m subspaces, and each of the multiple clustered nasal print features includes m clustered nasal print sub-features. The m clustered nasal print sub-features are respectively clustered in the m subspaces. Clustering is performed on the m nasal print sub-features to determine the target nasal print feature from the multiple clustered nasal print features in the clustering space, including:
[0012] Cluster the m nasal texture features in the corresponding subspaces of the m subspaces to obtain the cluster to which each of the m nasal texture features belongs in the corresponding subspace;
[0013] Obtain the target clustered nasal striae features that match each nasal striae feature in the cluster;
[0014] If m target clustered nasal pattern features belong to the same clustered nasal pattern feature A among multiple clustered nasal pattern features, then clustered nasal pattern feature A is identified as the target nasal pattern feature.
[0015] In this embodiment, by segmenting the nasal print features and matching them in the corresponding subspaces, the target clustered nasal print features with the highest similarity in the corresponding subspaces are obtained. This fully utilizes regional textures for matching. Only when m target clustered nasal print features matched from multiple regional textures belong to the same clustered nasal print feature are the clustered nasal print features used as the target nasal print features that best match the pet's nasal print features, which helps improve the accuracy of nasal print feature matching.
[0016] In conjunction with the first aspect, in one possible implementation, the nasal print feature includes roughness, each nasal print sub-feature includes a sub-roughness of roughness, and obtaining the target clustered nasal print sub-feature that matches each nasal print sub-feature in the cluster includes:
[0017] Calculate the average value of the sub-roughness and the average value of the roughness;
[0018] The first contribution coefficient of each nose texture sub-feature is obtained based on the average value of the sub-roughness and the average value of the roughness.
[0019] Map the first contribution coefficient to the first weight coefficient;
[0020] Based on the corresponding subspace of each nasal texture feature, obtain the distance between each nasal texture feature and multiple clustered nasal texture features in the cluster.
[0021] Based on the first weighting coefficient and the distance, the similarity between each nasal pattern feature and multiple clustered nasal pattern features is obtained;
[0022] The clustered nasal pattern feature with the highest similarity among multiple clustered nasal pattern features is identified as the target clustered nasal pattern feature.
[0023] In this embodiment, the ratio of the average value of the sub-roughness to the average value of the roughness is used as the first contribution coefficient. The first contribution coefficient is used to characterize the degree of contribution of the sub-roughness to the matching of the target nasal texture feature. The first contribution coefficient is mapped to a first weight coefficient using a preset formula. Then, the first weight coefficient is multiplied by the calculated distance to map the distance between each nasal texture sub-feature and multiple clustered nasal texture sub-features in the cluster to the similarity category. This facilitates the electronic device to directly select the target clustered nasal texture sub-feature based on similarity.
[0024] In conjunction with the first aspect, in one possible implementation, the nasal print features include roughness, contrast, orientation, line density, regularity, and coarseness. Each nasal print sub-feature includes sub-roughness of roughness, sub-contrast of contrast, sub-orientation of orientation, sub-line density of line density, sub-regularity of regularity, and sub-coarseness of coarseness. The target clustered nasal print sub-features that match each nasal print sub-feature in the cluster are obtained, including:
[0025] The second contribution coefficient of each nasal texture sub-feature is calculated based on roughness, contrast, orientation, line resolution, regularity, coarseness, sub-roughness, sub-contrast, sub-orientation, sub-line resolution, sub-regularity, and sub-coarseness.
[0026] Map the second contribution coefficient to the second weighting coefficient;
[0027] Based on the corresponding subspace of each nasal texture feature, obtain the distance between each nasal texture feature and multiple clustered nasal texture features in the cluster.
[0028] Based on the second weighting coefficient and the distance, the similarity between each nasal pattern feature and multiple clustered nasal pattern features is obtained;
[0029] The clustered nasal pattern feature with the highest similarity among multiple clustered nasal pattern features is identified as the target clustered nasal pattern feature.
[0030] In conjunction with the first aspect, in one possible implementation, a second contribution coefficient for each nasal texture sub-feature is calculated based on roughness, contrast, orientation, line resolution, regularity, coarseness, sub-roughness, sub-contrast, sub-orientation, sub-line resolution, sub-regularity, and sub-coarseness, including:
[0031] The first ratio is obtained based on the average roughness and the average sub-roughness;
[0032] The second ratio is obtained based on the average of the contrast ratio and the average of the sub-contrast ratios;
[0033] The third ratio is obtained based on the average value of the orientation and the average value of the sub-orientations;
[0034] The fourth ratio is obtained based on the average value of the line resolution and the average value of the sub-line resolution;
[0035] The fifth ratio is obtained based on the average of the regularity and the average of the sub-regularities; and,
[0036] The sixth ratio is obtained based on the average of the coarsness and the average of the sub-coarsness;
[0037] The second contribution coefficient is obtained based on the first ratio, the second ratio, the third ratio, the fourth ratio, the fifth ratio, and the sixth ratio.
[0038] In this embodiment, a second contribution coefficient is established based on a first ratio, a second ratio, a third ratio, a fourth ratio, a fifth ratio, and a sixth ratio. This second contribution coefficient characterizes the contribution of each nasal texture sub-feature's sub-roughness, sub-contrast, sub-direction, sub-line resolution, sub-regularity, and sub-coarseness to the matching of the target nasal texture feature. A preset formula maps the second contribution coefficient to a second weighting coefficient. Then, the second weighting coefficient is multiplied by the calculated distance to map the distance between each nasal texture sub-feature and multiple clustered nasal texture sub-features in the cluster to a similarity category. This facilitates the electronic device directly selecting the target clustered nasal texture sub-feature based on similarity.
[0039] In conjunction with the first aspect, in one possible implementation, the method further includes:
[0040] If m target clustered nasal pattern features do not belong to clustered nasal pattern feature A, repeat the operation of clustering the m nasal pattern features in the corresponding subspaces of the m subspaces until the m target clustered nasal pattern features belong to the same clustered nasal pattern feature among multiple clustered nasal pattern features.
[0041] Record the number of clusters n and the s clusters to which each nasal texture feature belongs in the nth cluster, where s is less than or equal to n.
[0042] In conjunction with the first aspect, in one possible implementation, the method further includes:
[0043] For any nasal feature B among m nasal features, count the frequency of each cluster in n clusters among the s clusters of nasal feature B.
[0044] The clustering performance of the clustering algorithm for the corresponding subspace of the nasal pattern feature B is evaluated based on the frequency of each cluster in n clusterings.
[0045] In this implementation, by evaluating the clustering effect of the clustering algorithm for the corresponding subspace of the nasal print features, clustering algorithms that meet the requirements and those that do not can be obtained. For clustering algorithms that do not meet the requirements in the m subspaces, clustering algorithms that meet the requirements can be used to replace them, so as to improve the clustering effect of the nasal print features, thereby shortening the time to determine the target nasal print features and improving the clustering efficiency.
[0046] A second aspect of this application provides a pet information aggregation device based on pet nose prints, the device including an acquisition unit and a processing unit;
[0047] The acquisition unit is used to acquire the pet's data to be aggregated, which includes the pet's nose print image.
[0048] The processing unit is used to obtain the nose print features of the pet based on the nose print image;
[0049] The processing unit is also used to segment the nasal print features to obtain m nasal print sub-features, where m is an integer greater than 1;
[0050] The processing unit is also used to cluster the m nasal print features respectively, so as to determine the target nasal print feature from the multiple clustered nasal print features in the clustering space.
[0051] The processing unit is also used to identify the files associated with the target nasal print features as target files and to aggregate the data to be aggregated into the target files.
[0052] A third aspect of this application provides an electronic device including an input device and an output device, further including a processor adapted to implement one or more instructions; and a memory storing one or more computer programs adapted to be loaded by the processor and executed as steps in the method described in the first aspect.
[0053] A fourth aspect of this application provides a computer storage medium storing one or more instructions adapted for loading by a processor and executing the steps of the method as described in the first aspect. Attached Figure Description
[0054] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0055] Figure 1 A schematic diagram illustrating an application environment provided in an embodiment of this application;
[0056] Figure 2 A flowchart illustrating a pet information aggregation method based on pet nose prints provided in this application embodiment;
[0057] Figure 3 This is a schematic diagram illustrating the extraction of nasal texture features according to an embodiment of this application.
[0058] Figure 4A This is a schematic diagram illustrating the segmentation of nasal print features according to an embodiment of this application.
[0059] Figure 4B This is a schematic diagram illustrating another method for segmenting nasal vein features according to an embodiment of this application.
[0060] Figure 5 A schematic diagram illustrating the correspondence between nasal pattern sub-features and subspaces provided in an embodiment of this application;
[0061] Figure 6 A schematic diagram of a pet information aggregation device based on pet nose prints provided in this application embodiment;
[0062] Figure 7 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation
[0063] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort should fall within the scope of protection of the present application.
[0064] The terms "comprising" and "having," and any variations thereof, appearing in this specification, claims, and drawings, are intended to cover a non-exclusive inclusion. For example, a process, method, system, product, or apparatus that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to such process, method, product, or apparatus. Furthermore, the terms "first," "second," and "third," etc., are used to distinguish different objects and are not used to describe a specific order.
[0065] Please see Figure 1 , Figure 1 This is a schematic diagram of an application environment provided in an embodiment of this application, such as... Figure 1 As shown, the application environment includes a nose print image acquisition device 101, an electronic device 102, and a file database 103. These devices can transmit data over a network. Specifically, the nose print image acquisition device 101 acquires nose print images of pets and sends them to the electronic device 102. The electronic device 102 is equipped with a small object detection model, which extracts features from the nose print image to obtain the pet's nose print features. Based on these features, it searches for files associated with the pet and then aggregates the pet's files into this database. Specifically, the file database 103 stores all pet files and associates each pet's file with its nose print features. The database also provides a user interface for managing pet files. For example, the file database 103 can be deployed in the cloud or locally. For example, the electronic device 102 can be a computer, server, cloud device, self-service terminal, smart filing cabinet, etc. In some implementations, the nose print image acquisition device 101 can exist independently of the electronic device 102, or it can exist as a part of the electronic device 102.
[0066] Please see Figure 2 , Figure 2 This application provides a flowchart illustrating a method for pet information aggregation based on pet nose prints, which can be applied to electronic devices, such as... Figure 2 As shown, steps 201-204 are included:
[0067] 201: Obtain the pet's pending data, which includes the pet's nose print image.
[0068] In this embodiment, the data to be aggregated also includes basic information about the pet (e.g., breed, gender, size, etc.) and information about the pet owner. In different scenarios, the data to be aggregated may also include different information. For example, in a pet veterinary setting, the data may include the pet's condition, the veterinarian, medications used, and veterinary costs; in a pet grooming setting, the data may include grooming services, groomers, and grooming fees. Specifically, the nose print image can be captured by a nose print image acquisition device and then sent to an electronic device, or it can be sent by the pet owner through a mobile terminal connected to the electronic device.
[0069] 202: Obtain the pet's nose print features based on the nose print image.
[0070] In this embodiment of the application, a small target detection model can be used to detect targets in the nasal print image, and then features can be extracted from the pet's nasal region to obtain nasal print features. For example... Figure 3 As shown, a trained Faster R-CNN object detection framework is used to detect objects in the nasal print image, obtaining the pet's nasal region image. Then, the nasal region image is divided into multiple nasal region sub-images, and features are extracted from each of these sub-images to obtain the features of each sub-image. The features of each sub-image are then concatenated to form a feature map, resulting in the pet's nasal print features. For example, these nasal print features can be texture features. Texture features, as statistical features, are rotation-invariant and highly resistant to noise, making them suitable for retrieving texture images with significant differences in coarseness, density, etc. For example, these texture features can be one or more of Tamura texture features. Tamura texture features typically include six basic features: coarseness, contrast, directionality, linelikeness, regularity, and roughness. Their extraction methods can be found in existing descriptions. Taking roughness as an example, we divide the nose region image into 16 nose region sub-images of the same size in a 4*4 manner. For each nose region sub-image, we calculate its roughness and obtain 16 roughness values. According to the position of each nose region sub-image in the nose region image, we stitch the 16 roughness values into a 4*4 feature map to obtain the pet's nose print features.
[0071] 203: Segment the nasal print features to obtain m nasal print sub-features, where m is an integer greater than 1.
[0072] In this embodiment, the nasal print feature can be divided into the same number of nasal print sub-features according to the number of subspaces in the clustering space. For example, if the clustering space includes 4 subspaces, the nasal print feature can be divided into 4 nasal print sub-features. The specific division method can be found in [reference needed]. Figure 4A or Figure 4B .
[0073] 204: Cluster the m nasal pattern features separately to determine the target nasal pattern feature from the multiple clustered nasal pattern features in the clustering space.
[0074] In the embodiments of this application, such as Figure 5 As shown, m nasal print features are defined as nasal print feature 1, nasal print feature 2, ..., nasal print feature m. The subspaces of the clustering space are defined as subspace 1, subspace 2, ..., subspace m. The subspace corresponding to nasal print feature 1 is subspace 1, the subspace corresponding to nasal print feature 2 is subspace 2, ..., the subspace corresponding to nasal print feature m is subspace m. Different clustering algorithms are used to cluster the nasal print features in each of the m subspaces. For example, subspace 1 uses K-means clustering, subspace 2 uses K-center clustering, and so on. It should be understood that multiple clustered nasal print features are those that have already been clustered and are properly grouped within the clustering space. Multiple clustered nasal print features may include nasal print features from the same pet. Each clustered nasal print feature is also divided into m clustered nasal print sub-features during the clustering process, and is also assigned to the corresponding subspace for clustering. That is, there are multiple clusters in each subspace of the clustering space.
[0075] For example, clustering is performed on m nasal print features to determine the target nasal print feature from multiple clustered nasal print features in the clustering space, including:
[0076] Cluster the m nasal texture features in the corresponding subspaces of the m subspaces to obtain the cluster to which each of the m nasal texture features belongs in the corresponding subspace;
[0077] Obtain the target clustered nasal striae features that match each nasal striae feature in the cluster;
[0078] If m target clustered nasal pattern features belong to the same clustered nasal pattern feature A among multiple clustered nasal pattern features, then clustered nasal pattern feature A is identified as the target nasal pattern feature.
[0079] Specifically, the clustering algorithm corresponding to subspace 1 is used to cluster 1 of nasal texture feature 1, resulting in cluster 1 to which nasal texture feature 1 belongs. The clustering algorithm corresponding to subspace 2 is used to cluster 2 of nasal texture feature 2, resulting in cluster 2 to which nasal texture feature 2 belongs, and so on, thus obtaining the cluster to which each nasal texture feature belongs. For each nasal texture feature, its similarity is calculated with the clustered nasal texture features in its respective cluster. The clustered nasal texture feature with the highest similarity is taken as the target clustered nasal texture feature. For example, the target clustered nasal texture feature corresponding to nasal texture feature 1 is the clustered nasal texture feature 1 from cluster 1, the target clustered nasal texture feature corresponding to nasal texture feature 2 is the clustered nasal texture feature 2 from cluster 2, and so on, the target clustered nasal texture feature corresponding to nasal texture feature m is the clustered nasal texture feature m from cluster m. If clustered nasal print features 1, 2, ..., m belong to the same clustered nasal print feature A, then clustered nasal print feature A is determined as the target nasal print feature that matches the pet's nasal print features. In this embodiment, by segmenting the nasal print features and matching them in corresponding subspaces, the target clustered nasal print feature with the highest similarity in the corresponding subspace is obtained. This fully utilizes regional textures for matching. Only when m target clustered nasal print features matched from multiple regional textures belong to the same clustered nasal print feature is that clustered nasal print feature taken as the target nasal print feature that best matches the pet's nasal print features, which helps improve the accuracy of nasal print feature matching.
[0080] For example, nasal print features may only include roughness; in this case, nasal print sub-features are the sub-roughness obtained by segmenting that roughness. The target clustered nasal print sub-features that match each nasal print sub-feature in the cluster are obtained, including:
[0081] Calculate the average value of the sub-roughness and the average value of the roughness;
[0082] The first contribution coefficient of each nose texture sub-feature is obtained based on the average value of the sub-roughness and the average value of the roughness.
[0083] Map the first contribution coefficient to the first weight coefficient;
[0084] Based on the corresponding subspace of each nasal texture feature, obtain the distance between each nasal texture feature and multiple clustered nasal texture features in the cluster.
[0085] Based on the first weighting coefficient and the distance, the similarity between each nasal pattern feature and multiple clustered nasal pattern features is obtained;
[0086] The clustered nasal pattern feature with the highest similarity among multiple clustered nasal pattern features is identified as the target clustered nasal pattern feature.
[0087] Specifically, such as Figure 4A The nose texture feature is the sub-roughness, which includes four values. The average of these four values is then calculated to obtain the average sub-roughness. Similarly, Figure 4A The nose texture feature is roughness, which includes 16 values. The average of these 16 values is calculated to obtain the average roughness. The ratio of the average of the sub-roughness to the average of the roughness is used as the first contribution coefficient x1, where x1 ranges from [a1, a2]. A first preset formula is used to map x1 to a first weighting coefficient y1:
[0088]
[0089] a = a2 - a1;
[0090] Where e represents the natural constant, a represents the formula parameter, 0≤a1≤1, 0≤a2≤1.
[0091] Here, the m subspaces can be Euclidean, Manhattan, Chebyshev, etc., and can be the same, different, or partially the same. If subspace 1 is Euclidean, the Euclidean distance between nasal texture feature 1 and multiple clustered nasal texture features in its cluster is calculated. If subspace 2 is Manhattan, the Manhattan distance between nasal texture feature 2 and multiple clustered nasal texture features in its cluster is calculated. The distance d between each nasal texture feature and multiple clustered nasal texture features in its cluster is multiplied by y1 to obtain the similarity dy1 between each nasal texture feature and multiple clustered nasal texture features. Generally, the smaller the value of d, the greater the similarity between the nasal texture feature and the corresponding clustered nasal texture features. However, y1 is a negative number. Using dy1 as the similarity, the larger the value of dy1, the greater the similarity. Electronic devices can directly use the clustered nasal texture feature with the largest dy1 value as the target clustered nasal texture feature. In this embodiment, the ratio of the average value of the sub-roughness to the average value of the roughness is used as the first contribution coefficient x1. The first contribution coefficient x1 is used to characterize the degree of contribution of the sub-roughness to the matching of the target nasal texture feature. The first contribution coefficient x1 is mapped to the first weight coefficient y1 using a preset formula. Then, the first weight coefficient y1 is multiplied by the calculated distance d to map the distance between each nasal texture sub-feature and multiple clustered nasal texture sub-features in the cluster to the similarity category. This facilitates the electronic device to directly select the target clustered nasal texture sub-feature based on the value of dy1.
[0092] For example, nasal print features include roughness, contrast, orientation, line density, regularity, and coarseness. Each nasal print sub-feature includes sub-roughness of roughness, sub-contrast of contrast, sub-orientation of orientation, sub-line density of line density, sub-regularity of regularity, and sub-coarseness of coarseness. The target clustered nasal print sub-features that match each nasal print sub-feature in the cluster are obtained, including:
[0093] The second contribution coefficient of each nasal texture sub-feature is calculated based on roughness, contrast, orientation, line resolution, regularity, coarseness, sub-roughness, sub-contrast, sub-orientation, sub-line resolution, sub-regularity, and sub-coarseness.
[0094] Map the second contribution coefficient to the second weighting coefficient;
[0095] Based on the corresponding subspace of each nasal texture feature, obtain the distance between each nasal texture feature and multiple clustered nasal texture features in the cluster.
[0096] Based on the second weighting coefficient and the distance, the similarity between each nasal pattern feature and multiple clustered nasal pattern features is obtained;
[0097] The clustered nasal pattern feature with the highest similarity among multiple clustered nasal pattern features is identified as the target clustered nasal pattern feature.
[0098] Specifically, nasal texture features include roughness, contrast, directionality, linearity, regularity, and coarseness, which can be understood as... Figure 4A The nasal texture features at each location include six values: roughness, contrast, orientation, line density, regularity, and coarseness. Figure 4AThe feature map shown has 6 channels. Similarly, for each sub-roughness in the nasal texture sub-feature, a first ratio is calculated between its average value and the average roughness in the nasal texture feature; for each sub-contrast in the nasal texture sub-feature, a second ratio is calculated between its average value and the average contrast in the nasal texture feature; for each sub-directivity in the nasal texture sub-feature, a third ratio is calculated between its average value and the average directionality in the nasal texture feature; for each sub-line resolution in the nasal texture sub-feature, a fourth ratio is calculated between its average value and the average line resolution in the nasal texture feature; for each sub-regularity in the nasal texture sub-feature, a fifth ratio is calculated between its average value and the average regularity in the nasal texture feature; and for each sub-coarseness in the nasal texture sub-feature, a sixth ratio is calculated between its average value and the average coarseness in the nasal texture feature. Based on the first, second, third, fourth, fifth, and sixth ratios, a second contribution coefficient x2 is obtained. For example, x2 can be the average of the first, second, third, fourth, fifth, and sixth ratios, or the median of the first, second, third, fourth, fifth, and sixth ratios. The value range of x2 is [a1, a2]. The second preset formula is used to map x2 to a second weighting coefficient y2:
[0099]
[0100] The electronic device calculates the distance *d* between each nasal texture feature and multiple clustered nasal texture features in its respective cluster, based on the corresponding subspace of each nasal texture feature. For example, if the corresponding subspace is Euclidean space, the Euclidean distance between the nasal texture feature and multiple clustered nasal texture features in its respective cluster is calculated. The distance *d* between each nasal texture feature and multiple clustered nasal texture features in its respective cluster is multiplied by *y*² to obtain the similarity *dy*² between each nasal texture feature and multiple clustered nasal texture features. The clustered nasal texture feature with the largest *dy*² value is then selected as the target clustered nasal texture feature. In this embodiment, a second contribution coefficient *x* is used based on a first ratio, a second ratio, a third ratio, a fourth ratio, a fifth ratio, and a sixth ratio. The second contribution coefficient *x* represents the contribution of sub-roughness, sub-contrast, sub-direction, sub-line resolution, sub-regularity, and sub-coarseness to the matching of the target nasal texture feature. The second contribution coefficient x2 is mapped to the second weight coefficient y2 using a preset formula. Then, the second weight coefficient y2 is multiplied by the calculated distance d to map the distance between each nasal texture feature and multiple clustered nasal texture features in the cluster to the similarity category. This makes it easier for electronic devices to directly select the target clustered nasal texture feature based on the value of dy2.
[0101] 205: Identify the files associated with the target nasal print features as the target files, and aggregate the data to be aggregated into the target files.
[0102] In this embodiment, multiple clustered nasal print features have associated files. For example, if pet C previously visited a clinic, pet C's file already exists in the clinic's database and is associated with its clustered nasal print features. When pet C visits the clinic again, the electronic device can determine pet C's clustered nasal print features (i.e., target nasal print features) through steps 201-204, and then aggregate the pet's medical condition, attending physician, medication, and medical expenses into the target file.
[0103] For example, the method further includes:
[0104] If m target clustered nasal pattern features do not belong to clustered nasal pattern feature A, the operation of clustering the m nasal pattern features in their corresponding subspaces within the m subspaces is repeated until the m target clustered nasal pattern features belong to the same clustered nasal pattern feature among multiple clustered nasal pattern features. The number of clustering steps n and the s clusters to which each nasal pattern feature belongs in n clustering steps are recorded, where s is less than or equal to n. For example, when n = 3, the cluster to which nasal pattern feature 1 belongs in the first clustering step is cluster a, the cluster to which nasal pattern feature 1 belongs in the second clustering step is cluster b, and the cluster to which nasal pattern feature 1 belongs in the third clustering step is cluster a. Then, cluster a and cluster b are the s clusters to which nasal pattern feature 1 belongs.
[0105] For example, the method further includes:
[0106] For any nasal feature B among m nasal features, count the frequency of each cluster in n clusters among the s clusters of nasal feature B.
[0107] The clustering performance of the clustering algorithm for the corresponding subspace of the nasal pattern feature B is evaluated based on the frequency of each cluster in n clusterings.
[0108] In this embodiment, if a target cluster with a frequency greater than or equal to n / 2 exists among the s clusters of nasal print feature B, and nasal print feature B is clustered into the target cluster in the nth clustering, then the clustering algorithm for the corresponding subspace of nasal print feature B is considered an effective clustering algorithm, meaning its clustering effect meets the requirements. Conversely, if the target cluster is not clustered, the clustering algorithm for the corresponding subspace of nasal print feature B is considered an invalid clustering algorithm, meaning its clustering effect does not meet the requirements. Furthermore, for invalid clustering algorithms in the m subspaces, effective clustering algorithms can be used to replace them to improve the clustering effect of nasal print features, thereby shortening the time to determine the target nasal print feature and improving clustering efficiency.
[0109] As can be seen, this embodiment of the application obtains pet data to be aggregated, including the pet's nose print image; based on the nose print image, it obtains the pet's nose print features; it segments the nose print features to obtain m nose print sub-features, where m is an integer greater than 1; it clusters the m nose print sub-features to determine the target nose print feature from multiple clustered nose print features in the clustering space; it determines the file associated with the target nose print feature as the target file, and aggregates the data to be aggregated into the target file. In this way, based on the nose print sub-features, it determines the corresponding clustered nose print features (i.e., the target nose print features) of the pet, and determines the file associated with the clustered nose print features as the target file of the pet, improving the accuracy of finding the pet's target file, thereby improving the accuracy of pet information aggregation.
[0110] Based on the description of the above embodiment of the pet information aggregation method based on pet nose prints, please refer to... Figure 6 , Figure 6 A schematic diagram of a pet information aggregation device based on pet nose prints provided in this application embodiment is shown below. Figure 6 As shown, the device includes an acquisition unit 601 and a processing unit 602; wherein:
[0111] The acquisition unit 601 is used to acquire the pet's data to be aggregated, which includes the pet's nose print image.
[0112] The processing unit 602 is used to obtain the nose print features of a pet based on the nose print image;
[0113] The processing unit 602 is also used to segment the nasal print features to obtain m nasal print sub-features, where m is an integer greater than 1;
[0114] The processing unit 602 is also used to cluster the m nasal print features respectively, so as to determine the target nasal print feature from the multiple clustered nasal print features in the clustering space.
[0115] The processing unit 602 is also used to identify the file associated with the target nasal print feature as the target file, and to aggregate the data to be aggregated into the target file.
[0116] It can be seen that, Figure 6The pet information aggregation device based on pet nose prints, as shown, acquires pet data to be aggregated, including the pet's nose print image. Based on the nose print image, the pet's nose print features are obtained. These features are then segmented into m sub-features, where m is an integer greater than 1. Each of the m sub-features is clustered to identify the target nose print feature from multiple clustered features in the clustering space. The files associated with the target nose print feature are designated as target files, and the data to be aggregated is aggregated into the target files. This method, by identifying the corresponding clustered nose print features (i.e., the target nose print feature) based on the nose print sub-features and defining the files associated with these clustered features as the pet's target files, improves the accuracy of finding the pet's target files, thereby enhancing the accuracy of pet information aggregation.
[0117] In one possible implementation, the clustering space includes m subspaces, and each of the multiple clustered nasal print features includes m clustered nasal print sub-features. The m clustered nasal print sub-features are respectively clustered in the m subspaces. In determining the target nasal print feature from the multiple clustered nasal print features in the clustering space by clustering the m nasal print sub-features respectively, the processing unit 602 is specifically used for:
[0118] Cluster the m nasal texture features in the corresponding subspaces of the m subspaces to obtain the cluster to which each of the m nasal texture features belongs in the corresponding subspace;
[0119] Obtain the target clustered nasal striae features that match each nasal striae feature in the cluster;
[0120] If m target clustered nasal pattern features belong to the same clustered nasal pattern feature A among multiple clustered nasal pattern features, then clustered nasal pattern feature A is identified as the target nasal pattern feature.
[0121] In one possible implementation, the nasal print features include roughness, and each nasal print sub-feature includes a sub-roughness of roughness. In acquiring the target clustered nasal print sub-features that match each nasal print sub-feature in the cluster, the processing unit 602 is specifically configured to:
[0122] Calculate the average value of the sub-roughness and the average value of the roughness;
[0123] The first contribution coefficient of each nose texture sub-feature is obtained based on the average value of the sub-roughness and the average value of the roughness.
[0124] Map the first contribution coefficient to the first weight coefficient;
[0125] Based on the corresponding subspace of each nasal texture feature, obtain the distance between each nasal texture feature and multiple clustered nasal texture features in the cluster.
[0126] Based on the first weighting coefficient and the distance, the similarity between each nasal pattern feature and multiple clustered nasal pattern features is obtained;
[0127] The clustered nasal pattern feature with the highest similarity among multiple clustered nasal pattern features is identified as the target clustered nasal pattern feature.
[0128] In one possible implementation, the nasal print features include roughness, contrast, orientation, line density, regularity, and coarseness. Each nasal print sub-feature includes sub-roughness of roughness, sub-contrast of contrast, sub-orientation of orientation, sub-line density of line density, sub-regularity of regularity, and sub-coarseness of coarseness. In acquiring the target clustered nasal print sub-features that match each nasal print sub-feature in the cluster, the processing unit 602 is specifically configured to:
[0129] The second contribution coefficient of each nasal texture sub-feature is calculated based on roughness, contrast, orientation, line resolution, regularity, coarseness, sub-roughness, sub-contrast, sub-orientation, sub-line resolution, sub-regularity, and sub-coarseness.
[0130] Map the second contribution coefficient to the second weighting coefficient;
[0131] Based on the corresponding subspace of each nasal texture feature, obtain the distance between each nasal texture feature and multiple clustered nasal texture features in the cluster.
[0132] Based on the second weighting coefficient and the distance, the similarity between each nasal pattern feature and multiple clustered nasal pattern features is obtained;
[0133] The clustered nasal pattern feature with the highest similarity among multiple clustered nasal pattern features is identified as the target clustered nasal pattern feature.
[0134] In one possible implementation, in calculating the second contribution coefficient of each nasal texture sub-feature based on roughness, contrast, orientation, line resolution, regularity, coarseness, sub-roughness, sub-contrast, sub-orientation, sub-line resolution, sub-regularity, and sub-coarseness, the processing unit 602 is specifically configured to:
[0135] The first ratio is obtained based on the average roughness and the average sub-roughness;
[0136] The second ratio is obtained based on the average of the contrast ratio and the average of the sub-contrast ratios;
[0137] The third ratio is obtained based on the average value of the orientation and the average value of the sub-orientations;
[0138] The fourth ratio is obtained based on the average value of the line resolution and the average value of the sub-line resolution;
[0139] The fifth ratio is obtained based on the average of the regularity and the average of the sub-regularities; and,
[0140] The sixth ratio is obtained based on the average of the coarsness and the average of the sub-coarsness;
[0141] The second contribution coefficient is obtained based on the first ratio, the second ratio, the third ratio, the fourth ratio, the fifth ratio, and the sixth ratio.
[0142] In one possible implementation, the processing unit 602 is further configured to:
[0143] If m target clustered nasal pattern features do not belong to clustered nasal pattern feature A, repeat the operation of clustering the m nasal pattern features in the corresponding subspaces of the m subspaces until the m target clustered nasal pattern features belong to the same clustered nasal pattern feature among multiple clustered nasal pattern features.
[0144] Record the number of clusters n and the s clusters to which each nasal texture feature belongs in the nth cluster, where s is less than or equal to n.
[0145] In one possible implementation, the processing unit 602 is further configured to:
[0146] For any nasal feature B among m nasal features, count the frequency of each cluster in n clusters among the s clusters of nasal feature B.
[0147] The clustering performance of the clustering algorithm for the corresponding subspace of the nasal pattern feature B is evaluated based on the frequency of each cluster in n clusterings.
[0148] Based on the description of the method and apparatus embodiments above, this application also provides an electronic device. Please refer to... Figure 7 The electronic device includes at least a processor 701, an input device 702, an output device 703, and a memory 704. The processor 701, input device 702, output device 703, and memory 704 within the electronic device can be connected via a bus or other means.
[0149] The memory 704 can be stored in the memory of the electronic device. The memory 704 is used to store computer programs, which include program instructions. The processor 701 is used to execute the program instructions stored in the memory 704. The processor 701 (or CPU (Central Processing Unit)) is the computing and control core of the electronic device. It is suitable for implementing one or more instructions, specifically for loading and executing one or more instructions to achieve corresponding method flows or corresponding functions.
[0150] In one embodiment, the processor 701 of the electronic device provided in this application can be used to perform a series of pet information aggregation methods based on pet nose prints:
[0151] Obtain the pet's data to be aggregated, which includes the pet's nose print image;
[0152] Based on the nose print image, obtain the nose print characteristics of the pet;
[0153] The nasal print features are segmented to obtain m nasal print sub-features, where m is an integer greater than 1;
[0154] Cluster the m nasal pattern features separately to determine the target nasal pattern feature from the multiple clustered nasal pattern features in the clustering space;
[0155] The files associated with the target nasal print features are identified as target files, and the data to be aggregated is aggregated into the target files.
[0156] In another embodiment, the clustering space includes m subspaces, and each of the multiple clustered nasal print features includes m clustered nasal print sub-features. The m clustered nasal print sub-features are respectively clustered in the m subspaces. The processor 701 performs clustering on the m nasal print sub-features respectively to determine the target nasal print feature from the multiple clustered nasal print features in the clustering space, including:
[0157] Cluster the m nasal texture features in the corresponding subspaces of the m subspaces to obtain the cluster to which each of the m nasal texture features belongs in the corresponding subspace;
[0158] Obtain the target clustered nasal striae features that match each nasal striae feature in the cluster;
[0159] If m target clustered nasal pattern features belong to the same clustered nasal pattern feature A among multiple clustered nasal pattern features, then clustered nasal pattern feature A is identified as the target nasal pattern feature.
[0160] In another embodiment, the nasal print feature includes roughness, and each nasal print sub-feature includes a sub-roughness of roughness. The processor 701 performs the process of acquiring target clustered nasal print sub-features that match each nasal print sub-feature in a cluster, including:
[0161] Calculate the average value of the sub-roughness and the average value of the roughness;
[0162] The first contribution coefficient of each nose texture sub-feature is obtained based on the average value of the sub-roughness and the average value of the roughness.
[0163] Map the first contribution coefficient to the first weight coefficient;
[0164] Based on the corresponding subspace of each nasal texture feature, obtain the distance between each nasal texture feature and multiple clustered nasal texture features in the cluster.
[0165] Based on the first weighting coefficient and the distance, the similarity between each nasal pattern feature and multiple clustered nasal pattern features is obtained;
[0166] The clustered nasal pattern feature with the highest similarity among multiple clustered nasal pattern features is identified as the target clustered nasal pattern feature.
[0167] In another embodiment, the nasal print features include roughness, contrast, orientation, line resolution, regularity, and coarseness. Each nasal print sub-feature includes sub-roughness of roughness, sub-contrast of contrast, sub-orientation of orientation, sub-line resolution of line resolution, sub-regularity of regularity, and sub-coarseness of coarseness. The processor 701 performs the following steps: obtaining the target clustered nasal print sub-features that match each nasal print sub-feature in the cluster, including:
[0168] The second contribution coefficient of each nasal texture sub-feature is calculated based on roughness, contrast, orientation, line resolution, regularity, coarseness, sub-roughness, sub-contrast, sub-orientation, sub-line resolution, sub-regularity, and sub-coarseness.
[0169] Map the second contribution coefficient to the second weighting coefficient;
[0170] Based on the corresponding subspace of each nasal texture feature, obtain the distance between each nasal texture feature and multiple clustered nasal texture features in the cluster.
[0171] Based on the second weighting coefficient and the distance, the similarity between each nasal pattern feature and multiple clustered nasal pattern features is obtained;
[0172] The clustered nasal pattern feature with the highest similarity among multiple clustered nasal pattern features is identified as the target clustered nasal pattern feature.
[0173] In another embodiment, processor 701 performs calculations based on roughness, contrast, orientation, line resolution, regularity, coarseness, sub-roughness, sub-contrast, sub-orientation, sub-line resolution, sub-regularity, and sub-coarseness to obtain a second contribution coefficient for each nasal texture sub-feature, including:
[0174] The first ratio is obtained based on the average roughness and the average sub-roughness;
[0175] The second ratio is obtained based on the average of the contrast ratio and the average of the sub-contrast ratios;
[0176] The third ratio is obtained based on the average value of the orientation and the average value of the sub-orientations;
[0177] The fourth ratio is obtained based on the average value of the line resolution and the average value of the sub-line resolution;
[0178] The fifth ratio is obtained based on the average of the regularity and the average of the sub-regularities; and,
[0179] The sixth ratio is obtained based on the average of the coarsness and the average of the sub-coarsness;
[0180] The second contribution coefficient is obtained based on the first ratio, the second ratio, the third ratio, the fourth ratio, the fifth ratio, and the sixth ratio.
[0181] In another embodiment, the processor 701 is further configured to:
[0182] If m target clustered nasal pattern features do not belong to clustered nasal pattern feature A, repeat the operation of clustering the m nasal pattern features in the corresponding subspaces of the m subspaces until the m target clustered nasal pattern features belong to the same clustered nasal pattern feature among multiple clustered nasal pattern features.
[0183] Record the number of clusters n and the s clusters to which each nasal texture feature belongs in the nth cluster, where s is less than or equal to n.
[0184] In another embodiment, the processor 701 is further configured to:
[0185] For any nasal feature B among m nasal features, count the frequency of each cluster in n clusters among the s clusters of nasal feature B.
[0186] The clustering performance of the clustering algorithm for the corresponding subspace of the nasal pattern feature B is evaluated based on the frequency of each cluster in n clusterings.
[0187] For example, the electronic device includes, but is not limited to, a processor 701, an input device 702, an output device 703, and a memory 704. It may also include memory, a power supply, an application client module, etc. The input device 702 may be a keyboard, a touchscreen, an RF receiver, etc., and the output device 703 may be a speaker, a display, an RF transmitter, etc. Those skilled in the art will understand that the schematic diagram is merely an example of an electronic device and does not constitute a limitation on the electronic device; it may include more or fewer components than illustrated, or combine certain components, or use different components.
[0188] It should be noted that since the processor 701 of the electronic device executes the computer program to implement the steps in the above-described pet information filing method based on pet nose prints, all embodiments of the above-described pet information filing method based on pet nose prints are applicable to the electronic device and can achieve the same or similar beneficial effects.
[0189] This application embodiment also provides a computer storage medium (memory), which is a memory device in an electronic device used to store programs and data. It is understood that the computer storage medium here can include the built-in storage medium in a terminal, or it can include an extended storage medium supported by the terminal. The computer storage medium provides storage space that stores the terminal's operating system. Furthermore, this storage space also stores one or more instructions suitable for loading and execution by the processor 701. These instructions can be one or more computer programs (including program code). It should be noted that the computer storage medium here can be a high-speed RAM memory, or a non-volatile memory, such as at least one disk storage device; optionally, it can also be at least one computer storage medium located remotely from the aforementioned processor 701. In one embodiment, the processor 701 can load and execute one or more instructions stored in the computer storage medium to implement the corresponding steps of the above-described pet information aggregation method based on pet nose prints.
[0190] For example, a computer program on a computer storage medium includes computer program code, which may be in the form of source code, object code, executable file, or some intermediate form. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording media, USB flash drive, portable hard drive, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc.
[0191] It should be noted that since the computer program on the computer storage medium is executed by the processor to implement the steps in the above-described pet information archiving method based on pet nose prints, all embodiments of the above-described pet information archiving method based on pet nose prints are applicable to the computer storage medium and can achieve the same or similar beneficial effects.
[0192] The embodiments of this application have been described in detail above. Specific examples have been used to illustrate the principles and implementation methods of this application. The description of the above embodiments is only for the purpose of helping to understand the method and core ideas of this application. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this application. Therefore, the content of this specification should not be construed as a limitation of this application.
Claims
1. A method for pet information archiving based on pet nose prints, characterized in that, The method includes: Obtain the pet's data to be aggregated, including the pet's nose print image; Based on the nose print image, the nose print characteristics of the pet are obtained; The nasal print features are segmented to obtain m nasal print sub-features, where m is an integer greater than 1; Cluster the m nasal pattern features respectively to determine the target nasal pattern feature from the multiple clustered nasal pattern features in the clustering space; The files associated with the target nasal print features are identified as target files, and the files to be aggregated are aggregated into the target files; The clustering space includes m subspaces. Each clustered nasal print feature includes m clustered nasal print sub-features. The m clustered nasal print sub-features are respectively clustered in the m subspaces. The m subspaces include Euclidean space, Manhattan space, and Chebyshev space. The m subspaces are different. The step of clustering the m nasal print sub-features to determine the target nasal print feature from the multiple clustered nasal print features in the clustering space includes: The m nasal pattern features are clustered in the corresponding subspaces of the m subspaces respectively to obtain the cluster of each nasal pattern feature in the corresponding subspace; Obtain the target clustered nasal pattern feature that matches each nasal pattern feature in the cluster; If m target clustered nasal print features belong to the same clustered nasal print feature A among the plurality of clustered nasal print features, then the clustered nasal print feature A is determined as the target nasal print feature. The nasal texture feature includes roughness, and each nasal texture sub-feature includes a sub-roughness of the roughness. The step of obtaining the target clustered nasal texture sub-feature that matches each nasal texture sub-feature in the cluster includes: Calculate the average value of the sub-roughness and the average value of the roughness; Based on the average value of the sub-roughness and the average value of the roughness, a first contribution coefficient for each nasal texture sub-feature is obtained; wherein the value range of the first contribution coefficient is... ; The first contribution coefficient is mapped to the first weighting coefficient using the following formula: ； ； in, This represents the first weighting coefficient. Represents the natural constant. Indicates formula parameters, , ; Based on the corresponding subspace of each nasal pattern feature, the distance between each nasal pattern feature and multiple clustered nasal pattern features in the cluster is obtained; Based on the first weighting coefficient and the distance, the similarity between each nasal pattern feature and the plurality of clustered nasal pattern features is obtained; The clustered nasal pattern feature with the highest similarity among the multiple clustered nasal pattern features is determined as the target clustered nasal pattern feature.
2. The method according to claim 1, characterized in that, The nasal texture features include roughness, contrast, orientation, line density, regularity, and coarseness. Each nasal texture sub-feature includes sub-roughness of the roughness, sub-contrast of the contrast, sub-orientation of the orientation, sub-line density of the line density, sub-regularity of the regularity, and sub-coarseness of the coarseness. Obtaining the target clustered nasal texture sub-features that match each nasal texture sub-feature in the cluster includes: The second contribution coefficient of each nasal texture sub-feature is calculated based on the roughness, contrast, orientation, line resolution, regularity, coarseness, sub-roughness, sub-contrast, sub-orientation, sub-line resolution, sub-regularity, and sub-coarseness. Map the second contribution coefficient to the second weighting coefficient; Based on the corresponding subspace of each nasal pattern feature, the distance between each nasal pattern feature and multiple clustered nasal pattern features in the cluster is obtained; Based on the second weighting coefficient and the distance, the similarity between each nasal pattern feature and the plurality of clustered nasal pattern features is obtained; The clustered nasal pattern feature with the highest similarity among the multiple clustered nasal pattern features is determined as the target clustered nasal pattern feature.
3. The method according to claim 2, characterized in that, The calculation of the second contribution coefficient for each nasal texture sub-feature based on the roughness, contrast, orientation, line resolution, regularity, coarseness, sub-roughness, sub-contrast, sub-orientation, sub-line resolution, sub-regularity, and sub-coarseness includes: A first ratio is obtained based on the average value of the roughness and the average value of the sub-roughness; A second ratio is obtained based on the average value of the contrast ratio and the average value of the sub-contrast ratio; A third ratio is obtained based on the average value of the orientation and the average value of the sub-orientations; A fourth ratio is obtained based on the average value of the line resolution and the average value of the sub-line resolution; A fifth ratio is obtained based on the average of the regularity and the average of the sub-regularity; and, The sixth ratio is obtained based on the average value of the coarsness and the average value of the sub-coarsness; The second contribution coefficient is obtained based on the first ratio, the second ratio, the third ratio, the fourth ratio, the fifth ratio, and the sixth ratio.
4. The method according to any one of claims 1-3, characterized in that, The method further includes: If m target clustered nasal pattern features do not belong to the clustered nasal pattern feature A, the operation of clustering the m nasal pattern features in the corresponding subspaces of the m subspaces is repeated until the m target clustered nasal pattern features belong to the same clustered nasal pattern feature among the multiple clustered nasal pattern features. Record the number of clustering n and the s clusters to which each nasal texture feature belongs in the nth clustering, where s is less than or equal to n.
5. The method according to claim 4, characterized in that, The method further includes: For any one of the m nasal pattern features B, count the frequency of each cluster in the nth clustering of the s clusters of the nasal pattern feature B. The clustering effect of the clustering algorithm for the corresponding subspace of the nasal texture feature B is evaluated based on the frequency of occurrence of each cluster in the nth clustering.
6. A pet information aggregation device based on pet nose prints, characterized in that, The device includes an acquisition unit and a processing unit; The acquisition unit is used to acquire the pet's data to be collected, the data to be collected including the pet's nose print image; The processing unit is used to obtain the nose print features of the pet based on the nose print image; The processing unit is also used to segment the nasal print features to obtain m nasal print sub-features, where m is an integer greater than 1; The processing unit is further configured to cluster the m nasal print features respectively, so as to determine the target nasal print feature from the multiple clustered nasal print features in the clustering space. The processing unit is further configured to identify the file associated with the target nasal print feature as the target file, and to aggregate the file data to be aggregated into the target file; The clustering space includes m subspaces. Each clustered nasal print feature includes m clustered nasal print sub-features. The m clustered nasal print sub-features are respectively clustered in the m subspaces. The m subspaces include Euclidean space, Manhattan space, and Chebyshev space. The m subspaces are different. In terms of clustering the m nasal print sub-features to determine the target nasal print feature from the multiple clustered nasal print features in the clustering space, the processing unit is specifically used for: The m nasal pattern features are clustered in the corresponding subspaces of the m subspaces respectively to obtain the cluster of each nasal pattern feature in the corresponding subspace; Obtain the target clustered nasal pattern feature that matches each nasal pattern feature in the cluster; If m target clustered nasal print features belong to the same clustered nasal print feature A among the plurality of clustered nasal print features, then the clustered nasal print feature A is determined as the target nasal print feature. The nasal texture feature includes roughness, and each nasal texture sub-feature includes a sub-roughness of the roughness. In acquiring target clustered nasal texture sub-features that match each nasal texture sub-feature in the cluster, the processing unit is specifically used for: Calculate the average value of the sub-roughness and the average value of the roughness; Based on the average value of the sub-roughness and the average value of the roughness, a first contribution coefficient for each nasal texture sub-feature is obtained; wherein the value range of the first contribution coefficient is... ; The first contribution coefficient is mapped to the first weighting coefficient using the following formula: ； ； in, This represents the first weighting coefficient. Represents the natural constant. Indicates formula parameters, , ; Based on the corresponding subspace of each nasal pattern feature, the distance between each nasal pattern feature and multiple clustered nasal pattern features in the cluster is obtained; Based on the first weighting coefficient and the distance, the similarity between each nasal pattern feature and the plurality of clustered nasal pattern features is obtained; The clustered nasal pattern feature with the highest similarity among the multiple clustered nasal pattern features is determined as the target clustered nasal pattern feature.
7. An electronic device comprising an input device and an output device, characterized in that, Also includes: A processor, suitable for implementing one or more computer programs; as well as, A memory storing one or more computer programs adapted to be loaded by the processor and executed as described in any one of claims 1-5.
8. A computer storage medium, characterized in that, The computer storage medium stores one or more instructions, which are adapted to be loaded by a processor and executed as described in any one of claims 1-5.