A fingerprint spectrum type identification method, an electronic device and a storage medium

Abstract

The present application relates to the technical field of sample analysis, and specifically provides a fingerprint spectrum type identification method, an electronic device and a storage medium, aiming to solve the problem of how to accurately identify the type of fingerprint spectrum. To this end, the method of the present application comprises obtaining a fingerprint spectrum to be identified, inputting the fingerprint spectrum to be identified into a model set composed of at least one effective identification model, and obtaining a fingerprint spectrum type identification result; wherein the at least one effective identification model is obtained by screening a trained identification model according to a preset condition, and the identification model is trained based on corresponding ion peak clusters. Through the above implementation, the overall fingerprint spectrum to be identified can be analyzed, and the limitation of relying on identifying individual characteristic peaks in the traditional method is eliminated, so that the identification result is more accurate. Moreover, the fingerprint spectrum type identification method is not limited by the source of the sample, is simpler and more reliable, and has a wider range of applications.

Description

Technical Field

[0001] This invention relates to the field of sample analysis technology, specifically to a fingerprint pattern type identification method, electronic device, and storage medium. Background Technology

[0002] Traditional fingerprint pattern type identification methods mostly rely on feature peak identification. However, some fingerprint patterns have indistinct or low-abundance feature peaks, making them difficult to detect and thus preventing fingerprint pattern type identification.

[0003] Accordingly, a new technical solution is needed in this field to solve the above problems. Summary of the Invention

[0004] To overcome the above-mentioned deficiencies, the present invention is proposed to provide a fingerprint pattern type identification method, electronic device, and storage medium that solves or at least partially solves the technical problem of how to accurately identify fingerprint pattern types.

[0005] Firstly, a fingerprint pattern type identification method is provided, the method comprising:

[0006] Obtain the fingerprint pattern to be identified;

[0007] The fingerprint spectrum to be identified is input into a model set consisting of at least one effective identification model to obtain a fingerprint spectrum type identification result. The at least one effective identification model is obtained by filtering the trained identification models according to preset conditions. The identification model is trained based on the corresponding ion peak cluster.

[0008] In one technical solution of the above fingerprint pattern type recognition method, the method further includes acquiring a model set, wherein acquiring the model set specifically includes:

[0009] Multiple recognition models are trained to obtain multiple trained recognition models;

[0010] Based on the preset conditions, the multiple trained recognition models are filtered to obtain at least one effective recognition model;

[0011] The at least one effective recognition model is combined to obtain the model set.

[0012] In one technical solution of the above fingerprint pattern type recognition method, training multiple recognition models includes:

[0013] Multiple training sets are obtained, wherein each training set is a corresponding ion peak cluster labeled with a type label;

[0014] Each training set is input into the corresponding recognition model, and the corresponding recognition model is trained accordingly.

[0015] In one technical solution of the above fingerprint pattern type recognition method, obtaining multiple training sets includes:

[0016] Obtain multiple fingerprint maps labeled with type tags;

[0017] The multiple fingerprint spectra are normalized to obtain the set of ion peaks of the multiple fingerprint spectra;

[0018] The set of ion peaks is divided into at least one ion peak cluster.

[0019] In one technical solution of the above fingerprint spectrum type identification method, dividing the ion peak set into at least one ion peak cluster includes:

[0020] Obtain the density of the distribution of multiple ion peaks in the ion peak set;

[0021] The plurality of ion peaks are divided into at least one ion peak cluster according to the density.

[0022] In one technical solution of the aforementioned fingerprint pattern type recognition method,

[0023] The preset conditions include at least a preset number of fingerprint maps and a preset accuracy threshold.

[0024] The step of filtering the multiple trained recognition models according to the preset conditions includes:

[0025] Determine whether the number of fingerprint spectra participating in the training of the multiple trained recognition models meets the preset number of fingerprint spectra.

[0026] Determine whether the accuracy of each of the multiple recognition models meets the preset accuracy threshold.

[0027] The identification model that simultaneously satisfies the preset number of fingerprint spectra and the preset accuracy threshold is selected as the effective identification model.

[0028] In one technical solution of the above fingerprint pattern type recognition method, the step of combining the at least one effective recognition model to obtain the model set includes:

[0029] Multiple effective recognition models are combined in different ways to obtain multiple model sets composed of the multiple effective recognition models.

[0030] In one technical solution of the above fingerprint pattern type recognition method, the method further includes:

[0031] Multiple recognition results are output based on the multiple model sets;

[0032] An error function is calculated for the multiple recognition results;

[0033] Based on the calculation results of the error function, each of the multiple model sets is checked to see if it meets the preset error threshold.

[0034] Select one or more model sets that meet the preset error threshold as trained model sets.

[0035] In a second aspect, an electronic device is provided, comprising a processor and a storage device, the storage device being adapted to store a plurality of program codes, the program codes being adapted to be loaded and executed by the processor to perform the fingerprint pattern type recognition method described in any of the above-described technical solutions.

[0036] In a third aspect, a computer-readable storage medium is provided, wherein a plurality of program codes are stored therein, the program codes being adapted to be loaded and run by a processor to perform the fingerprint pattern type recognition method described in any of the above-described technical solutions.

[0037] The present invention comprises one or more of the following technical solutions:

[0038] Beneficial effects:

[0039] In implementing the technical solution of this invention, a fingerprint spectrum to be identified is first obtained. Then, the fingerprint spectrum to be identified is input into a model set consisting of at least one effective identification model to obtain a fingerprint spectrum type identification result. The at least one effective identification model is selected from trained identification models according to preset conditions, and the identification model is trained based on the corresponding ion peak cluster. Through the above implementation method, the entire fingerprint spectrum to be identified can be analyzed, overcoming the limitation of traditional methods that rely on identifying individual characteristic peaks. Therefore, the identification result is more accurate, and this fingerprint spectrum type identification method is not limited by the sample source, is simpler and more reliable, and has a wider range of applications. Attached Figure Description

[0040] The disclosure of this invention will become more readily understood with reference to the accompanying drawings. It will be readily understood by those skilled in the art that these drawings are for illustrative purposes only and are not intended to limit the scope of protection of this invention. Wherein:

[0041] Figure 1 This is a schematic flowchart of the main steps of a fingerprint pattern type identification method according to an embodiment of the present invention;

[0042] Figure 2 This is a schematic diagram of the main steps in obtaining a model set according to an embodiment of the present invention;

[0043] Figure 3 This is a schematic diagram of the structure of a fingerprint pattern type recognition model set according to an embodiment of the present invention;

[0044] Figure 4 This is a schematic diagram of the main steps for training multiple recognition models according to an embodiment of the present invention.

[0045] Figure 5 This is a schematic diagram of the main steps for obtaining multiple training sets according to an embodiment of the present invention;

[0046] Figure 6 This is a schematic diagram of the ion peak set of multiple fingerprint spectra according to an embodiment of the present invention;

[0047] Figure 7 This is a schematic diagram of the main steps in dividing an ion peak set into multiple ion peak clusters according to an embodiment of the present invention;

[0048] Figure 8 This is a schematic diagram illustrating the division of an ion peak set into multiple ion peak clusters according to an embodiment of the present invention;

[0049] Figure 9 This is a schematic diagram of multiple trained recognition models according to an embodiment of the present invention;

[0050] Figure 10 This is a schematic diagram of the main steps of filtering multiple trained recognition models according to preset conditions according to an embodiment of the present invention.

[0051] Figure 11 This is a schematic diagram illustrating the combination of multiple effective recognition models to obtain a model set according to an embodiment of the present invention;

[0052] Figure 12 This is a schematic flowchart of the main steps of a fingerprint pattern type identification method according to another embodiment of the present invention;

[0053] Figure 13 This is a schematic diagram of the main process of a fingerprint pattern type identification method according to an embodiment of the present invention;

[0054] Figure 14 This is a schematic diagram of the main structure of an electronic device according to an embodiment of the present invention.

[0055] List of reference numerals in the attached diagram:

[0056] 301: Acquisition module; 302: Processing module; 303: Clustering module; 304: Training module; 305: Filtering module; 306: Combination module; 307: Validation module; 1401: Processor; 1402: Storage device. Detailed Implementation

[0057] Some embodiments of the present invention will now be described with reference to the accompanying drawings. Those skilled in the art should understand that these embodiments are merely illustrative of the technical principles of the present invention and are not intended to limit the scope of protection of the present invention.

[0058] In the description of this invention, "module" and "processor" can include hardware, software, or a combination of both. A module can include hardware circuitry, various suitable sensors, communication ports, memory, and may also include software components, such as program code, or a combination of software and hardware. A processor can be a central processing unit, microprocessor, image processor, digital signal processor, or any other suitable processor. The processor has data and / or signal processing capabilities. The processor can be implemented in software, in hardware, or a combination of both. Non-transitory computer-readable storage media includes any suitable medium capable of storing program code, such as magnetic disks, hard disks, optical disks, flash memory, read-only memory, random access memory, etc. The term "A and / or B" means all possible combinations of A and B, such as only A, only B, or A and B. The terms "at least one A or B" or "at least one of A and B" have a similar meaning to "A and / or B" and can include only A, only B, or A and B. The singular terms "a" or "this" can also include plural forms.

[0059] Here we will first explain some of the terms involved in this invention.

[0060] Fingerprint spectroscopy refers to the chromatogram or spectrum obtained by appropriately processing and analyzing certain complex substances, such as traditional Chinese medicine, DNA of an organism, tissue, or cell, or proteins, which reveals their chemical characteristics. Fingerprint spectroscopy is mainly divided into traditional Chinese medicine fingerprint spectroscopy, DNA fingerprint spectroscopy, and peptide fingerprint spectroscopy, etc.

[0061] Ion peak: This refers to the peak produced by the ionization of a molecule by one electron. Generally, the peak with the highest mass number is the molecular ion peak, and its mass is the molecular weight. It is usually located at the end of the mass spectrum with the highest mass-to-charge ratio, and its mass number is the molecular weight of the compound.

[0062] MALDI-TOFMS: matrix-assisted laser desorption / ionization time-of-flight mass spectrometry, is an analytical instrument used in the field of biology.

[0063] SVM: Support Vector Machine, is a type of generalized linear classifier that performs binary classification of data using supervised learning. Its decision boundary is the maximum-margin hyperplane obtained by solving for the learning samples.

[0064] Dualistic classification, also known as binary classification, refers to classification based on two indicators in comparative studies involving two categories. For example, in studying the relationship between students' intelligence and gender, classifying by the two indicators of intelligence and gender is a dualistic classification. For observations based on dualistic classification, it is necessary to consider using a chi-square test or dualistic analysis of variance, depending on the characteristics of the data.

[0065] Peptide mass fingerprinting, also known as protein fingerprinting, is a high-throughput protein analysis method used to identify proteins. Endopeptides first cleave the unknown target protein into smaller peptides. The absolute mass of these peptides can be accurately measured using mass spectrometry, and a list of peptide peaks for the unknown protein can also be obtained.

[0066] As described in the background section, most traditional fingerprint spectrum type identification methods use feature peak identification. However, some fingerprint spectrum feature peaks are not obvious or have low abundance, making them difficult to detect and resulting in the inability to identify the fingerprint spectrum type.

[0067] To address the aforementioned problems, this invention provides a fingerprint pattern type identification method, an electronic device, and a storage medium.

[0068] See appendix Figure 1 , Figure 1 This is a schematic flowchart illustrating the main steps of a fingerprint pattern type identification method according to an embodiment of the present invention. Figure 1 As shown, the fingerprint pattern type identification method in this embodiment of the invention mainly includes the following steps S101 to S102.

[0069] Step S101: Obtain the fingerprint map to be identified.

[0070] Step S102: Input the fingerprint spectrum to be identified into a model set consisting of at least one valid identification model to obtain the fingerprint spectrum type identification result.

[0071] Among them, at least one effective recognition model is selected from the trained recognition models according to preset conditions.

[0072] The identification model is trained based on the corresponding ion peak clusters.

[0073] Based on the method described in steps S101 to S102 above, the entire fingerprint spectrum to be identified can be analyzed, which gets rid of the limitation of traditional methods that rely on identifying individual feature peaks. Therefore, the identification results are more accurate. Moreover, this fingerprint spectrum type identification method is not limited by the sample source, is simpler and more reliable, and has a wider range of applications.

[0074] The following provides a further explanation of steps S101 to S102.

[0075] In some embodiments of step S101 above, the fingerprint spectrum to be identified can be obtained based on a MALDI-TOF MS device.

[0076] Among them, the MALDI-TOFMS instrument, or matrix-assisted laser desorption / ionization time-of-flight mass spectrometry, is an analytical instrument used in the field of biology. It features high sensitivity, high accuracy, and high resolution, providing a powerful analytical testing method for life sciences and other fields. Its functions can be briefly summarized as follows:

[0077] ① Matrix-assisted: The sample is lysed, and the protein is thoroughly mixed with the small molecule matrix solution. After the solvent evaporates, a co-crystallization is formed.

[0078] ②Laser desorption ionization: Under laser radiation, the matrix absorbs energy and transfers charge to sample molecules, forming an ionized sample;

[0079] ③ Flight time: Ionized protein samples fly through the flight tube under the action of an electric field. According to the principle that the mass-to-charge ratio (m / z) is proportional to the flight time of ions, ions of different masses are detected due to the difference in the time it takes for them to reach the detector, thus forming different mass spectra.

[0080] ④ Comparison and identification: Different proteins have different compositions. The fingerprint spectrum of the unknown sample collected by the device is compared with the identification spectrum of the known bacterial species in the database by statistical cluster analysis to obtain the identification results.

[0081] Therefore, fingerprints of samples such as serum, protein, peptide, and nucleic acid can be collected using MALDI-TOFMS equipment, without any limitations here.

[0082] The above is an explanation of step S101. The following is a further explanation of step S102.

[0083] In some embodiments of step S102 above, the fingerprint spectrum to be identified is input into a model set consisting of at least one valid identification model, and the fingerprint spectrum type identification result can be obtained.

[0084] Specifically, after obtaining the fingerprint spectrum, the fingerprint spectrum can be input into the model set to obtain the fingerprint spectrum type identification result, such as type A or type B, where type A and type B are two opposite or opposing types, such as type A being negative and type B being positive.

[0085] Furthermore, in some embodiments, the fingerprint pattern type identification method provided by the present invention further includes acquiring a model set.

[0086] See appendix Figure 2 , Figure 2 This is a schematic diagram illustrating the main steps of obtaining a model set according to an embodiment of the present invention. Figure 2 As shown, it mainly includes the following steps S201 to S203.

[0087] Step S201: Train multiple recognition models to obtain multiple trained recognition models.

[0088] Step S202: Select from multiple trained recognition models according to preset conditions to obtain at least one effective recognition model.

[0089] Step S203: Combine at least one valid recognition model to obtain a model set.

[0090] In some implementations, see Appendix Figure 3 , Figure 3 This is a schematic diagram of the structure of a model set according to an embodiment of the present invention. Figure 3 As shown, the model set may include an acquisition module 301, a processing module 302, a clustering module 303, a training module 304, a filtering module 305, a combination module 306, and a validation module 307.

[0091] The following is combined Figure 3 Further explanation is given for steps S201 to S203 above.

[0092] In some embodiments of step S201 above, see Appendix Figure 4 , Figure 4 This is a schematic diagram illustrating the main steps of training multiple recognition models according to an embodiment of the present invention. Figure 4 As shown, the above step S201 mainly includes steps S2011 to S2012.

[0093] Step S2011: Obtain multiple training sets.

[0094] Each training set consists of multiple ion peak clusters labeled with their respective types.

[0095] Specifically, see the appendix. Figure 5 , Figure 5This is a schematic diagram illustrating the main steps of obtaining multiple training sets according to an embodiment of the present invention. Figure 5 As shown, the above step S2011 mainly includes steps S501 to S503.

[0096] Step S501: Obtain multiple fingerprint maps labeled with type tags.

[0097] In some implementations, it can be based on Figure 3 The acquisition module 301 shown acquires multiple fingerprint maps labeled with type tags.

[0098] Specifically, it mainly includes the following steps:

[0099] (1) Label multiple samples of known types with type labels.

[0100] For example, the two known types of a sample, type A and type B (type A and type B are opposite or complementary types, such as type A being negative and type B being positive), are labeled with type labels '1' and '-1' respectively.

[0101] (2) Preprocess the multiple samples labeled with type tags and input them into the analysis device.

[0102] In some implementations, the methods for preprocessing multiple samples labeled with type tags include direct smearing, extended direct smearing, extraction, precipitation smearing, etc., which are not limited here.

[0103] Among them, the direct smear method involves directly smearing the sample to be tested onto the sample target, and then applying the matrix solution before running the instrument; the extended direct smear method involves covering the target with a 70% formic acid solution before covering the matrix to assist in cell wall disruption and lysis; the extraction method uses an extractant to extract pathogen proteins; the precipitation smear method is more suitable for mass spectrometry identification of trace pathogens, and the spectral quality is comparable to that of the extraction method.

[0104] (3) Based on the analysis equipment, multiple fingerprint maps labeled with type tags are obtained.

[0105] The analytical device can be the aforementioned MALDI-TOFMS device, which will not be elaborated further here. Furthermore, based on the MALDI-TOFMS device, multiple fingerprint profiles labeled with type tags can be obtained, such as the peptide mass fingerprint profile of serum, etc., which are not limited here.

[0106] The above is a further explanation of step S501.

[0107] Step S502: Normalize multiple fingerprint spectra to obtain a set of ion peaks from multiple fingerprint spectra.

[0108] In some implementations, it can be based on Figure 3 The processing module 302 shown performs normalization processing on multiple fingerprint spectra.

[0109] Specifically, data normalization algorithms, such as linear normalization, can be used to process multiple fingerprint spectra. This involves placing the ion peaks of all fingerprint spectra in the same space, normalizing the relative charge intensity and ion mass of the ion currents represented by the multiple fingerprint spectra, and obtaining a set of ion peaks from the multiple fingerprint spectra labeled with their type.

[0110] Specifically, see the appendix. Figure 6 , Figure 6 This is a schematic diagram of the ion peak set of multiple fingerprint spectra according to an embodiment of the present invention.

[0111] Figure 6 This involves superimposing 20 fingerprint spectra labeled with type tags using a data normalization algorithm to obtain a set of ion peaks. The horizontal axis represents the mass-to-charge ratio of ions, and the vertical axis represents ion intensity. Diamond-shaped dots represent positive samples, and circular dots represent negative samples. In practical applications, those skilled in the art can adjust the number of fingerprint spectra according to the specific scenario; no limitation is made here.

[0112] The above is a further explanation of step S502.

[0113] Step S503: Divide the ion peak set into at least one ion peak cluster.

[0114] In some implementations, it can be based on Figure 3 The clustering module 303 shown divides the ion peak set into at least one ion peak cluster.

[0115] Specifically, see the appendix. Figure 7 , Figure 7 This is a schematic flowchart illustrating the main steps of dividing an ion peak set into multiple ion peak clusters according to an embodiment of the present invention. Figure 7 As shown, step S503 mainly includes steps S5031 to S5032.

[0116] Step S5031: Obtain the density of ion peak distribution in the ion peak set.

[0117] Step S5032: Divide multiple ion peaks into at least one ion peak cluster according to their density.

[0118] Specifically, since some ion peaks in the ion peak set are densely distributed within a mass range and have a certain significant mass interval with other ion peaks outside the mass range, the clustering module 303 can divide multiple ion peaks that are clustered together in the ion peak set into an ion peak cluster according to the preset mass interval. Thus, the ion peak set will be divided into at least one ion peak cluster.

[0119] In some implementations, a computer can automatically divide the set of ion peaks into at least one ion peak cluster according to a preset mass interval.

[0120] In other embodiments, the ion peak set can be manually divided into at least one ion peak cluster by those skilled in the art according to a preset mass interval, which is not limited here.

[0121] See appendix Figure 8 , Figure 8 This is a schematic diagram illustrating the division of an ion peak set into multiple ion peak clusters according to an embodiment of the present invention. Figure 8 As shown, those skilled in the art can manually divide the ion peak sets of multiple fingerprint spectra into ion peak cluster 1, ion peak cluster 2, ion peak cluster 3, ion peak cluster 4 and ion peak cluster 5 according to a preset mass interval of 800.

[0122] It should be noted that the above examples of preset mass intervals are merely illustrative of suitability. In practical applications, those skilled in the art can set preset mass intervals according to specific scenarios, and no limitations are imposed here.

[0123] The above is a further explanation of step S2011.

[0124] Step S2012: Input each training set into the corresponding recognition model and train the corresponding recognition model respectively.

[0125] Specifically, multiple fingerprint pattern type recognition models can be trained based on multiple training sets, preset training parameters, and kernel parameters.

[0126] In some implementations, it can be based on Figure 3 The training module 304 shown inputs each training set into the corresponding recognition model, and trains the corresponding recognition model based on each training set, preset training parameters and kernel function to obtain multiple trained recognition models.

[0127] Furthermore, in some implementations, each training set, preset training parameters, and kernel function can be trained based on a binary classifier such as a support vector machine (SVM).

[0128] SVM is a type of generalized linear classifier that performs binary classification of data using supervised learning. Its decision boundary is the maximum-margin hyperplane obtained by solving for the learning samples.

[0129] SVM introduces maximum margin, resulting in high classification accuracy. It can accurately classify data even with a small sample size and has good generalization ability. The kernel function introduced by SVM can easily solve nonlinear problems and can also solve classification and regression problems with high-dimensional features. It can perform well even with data whose feature dimension is greater than the data size.

[0130] The preset training parameters and kernel functions of SVM can be set according to the actual application scenario, and are not limited here.

[0131] Further, see appendix. Figure 9 , Figure 9 This is a schematic diagram of multiple trained recognition models according to an embodiment of the present invention. Figure 9 As shown, recognition models 1, 2, 3, 4, and 5 are obtained by training on training sets 1, 2, 3, 4, and 5, respectively.

[0132] The above is a further explanation of step S201. The following is a further explanation of step S202.

[0133] In some embodiments of step S202 above, the following methods can be used: Figure 3 The filtering module 305 shown filters multiple trained recognition models according to preset conditions.

[0134] The preset conditions include at least the preset number of fingerprint patterns and the preset accuracy threshold.

[0135] See appendix Figure 10 , Figure 10 This is a schematic flowchart illustrating the main steps of filtering multiple trained recognition models according to preset conditions, based on an embodiment of the present invention. Figure 10 As shown, step S202 mainly includes steps S2021 to S2023.

[0136] Step S2021: Determine whether the number of fingerprint spectra participating in the training of the multiple trained recognition models meets the preset number of fingerprint spectra.

[0137] Step S2022: Determine whether the accuracy of each of the multiple recognition models meets the preset accuracy threshold.

[0138] Step S2023: Select the recognition models that simultaneously meet the preset number of fingerprint spectra and the preset accuracy threshold as valid recognition models.

[0139] In some implementations, a preset number of fingerprint spectra and a preset accuracy threshold can be set based on the training results. For example, if ten recognition models are obtained through training, the preset number of fingerprint spectra and the preset accuracy threshold can be set based on the number of fingerprint spectra used in the training of these ten recognition models and their accuracy.

[0140] For example, Figure 9 The five identification models shown were selected according to the above screening criteria, resulting in three effective identification models: effective identification model 1, effective identification model 2, and effective identification model 3.

[0141] In other implementations, if the number of fingerprint spectra involved in the training of multiple recognition models is small and / or the model accuracy is low, the number of samples can be appropriately increased, and the models can be trained again. The models can then be screened again based on the trained recognition models, or the preset number of fingerprint spectra and the preset accuracy threshold can be appropriately changed and screened again to obtain at least one effective recognition model. This is not limited here.

[0142] The above is a further explanation of step S202. The following is a further explanation of step S203.

[0143] In some embodiments of step S203 above, it can be based on Figure 3 The combination module 306 shown combines at least one valid recognition model to obtain a model set.

[0144] Specifically, multiple effective recognition models can be combined in different ways to obtain multiple model sets composed of multiple effective recognition models.

[0145] To improve the accuracy of the recognition model, multiple effective recognition models can be combined in various ways to obtain multiple model sets, where each model set consists of at least one effective model.

[0146] In some implementations, see Appendix Figure 11 , Figure 11 This is a schematic diagram illustrating the combination of multiple effective recognition models to obtain a model set according to an embodiment of the present invention. Figure 11As shown, any two of the three effective models selected above can be combined to form a model set, resulting in three model sets: the first model set composed of effective recognition model 1 and effective recognition model 2, the second model set composed of effective recognition model 1 and effective recognition model 3, and the third model set composed of effective recognition model 2 and effective recognition model 3.

[0147] Taking the first model set as an example, this model set is based on the joint recognition of the input fingerprint spectrum by effective recognition model 1 and effective recognition model 2.

[0148] In some implementations, the three effective models mentioned above can also be combined into a model set, namely, a fourth model set consisting of effective identification model 1, effective identification model 2 and effective identification model 3, which is not limited here.

[0149] In other implementations, if the number of valid identification models obtained through screening is other than six, then m (m≤6) models can be randomly selected from the six different models to form a group. If two models are randomly selected, then according to C... 2 6 = 15, which yields 15 model sets. If we randomly select three, then according to C... 3 6 = 20, which can yield a set of 20 sample type recognition models, without limitation here.

[0150] Of course, those skilled in the art will understand that it is also possible to use a set of models consisting of only one selected effective recognition model for recognition, but the combination of multiple effective models will yield more accurate recognition results than using a single effective recognition model.

[0151] The above is a further explanation of step S203.

[0152] Furthermore, in some embodiments, after performing the above step S203, it is also possible to base on Figure 3 The verification module 307 shown verifies the model set.

[0153] Specifically, see the appendix. Figure 12 , Figure 12 This is a schematic flowchart of the main steps of a fingerprint pattern type identification method according to another embodiment of the present invention. Figure 12 As shown, it mainly includes steps S1201 to S1204.

[0154] Step S1201: Output multiple recognition results based on multiple model sets.

[0155] In some implementations, a portion of the fingerprint samples labeled with type tags can be used as the training set for the training described above, while another portion can be used as the validation set for the validation here. For example, the allocation can be in an 8:2 ratio.

[0156] Furthermore, the validation set is input into multiple model sets, and the recognition results are output separately for each set.

[0157] Step S1202: Calculate the error function for multiple recognition results.

[0158] Taking SVM as an example, the error function includes calculating the classification error and the margin error.

[0159] Step S1203: Based on the calculation results of the error function, check whether multiple model sets meet the preset error threshold.

[0160] The preset error threshold can be set according to the specific use case, and there is no limitation here.

[0161] Step S1204: Select one or more model sets that meet the preset error threshold as trained model sets.

[0162] Specifically, one or more optimal model sets can be selected based on a preset error threshold to execute the aforementioned fingerprint pattern type identification method. For example, based on the preset error threshold... Figure 11 The four model sets shown are filtered to obtain the third model set as the optimal model set. Therefore, the fingerprint pattern type identification method described above can be executed based on the third model set.

[0163] Furthermore, in some embodiments, see Appendix Figure 13 , Figure 13 This is a schematic diagram of the main flow of a fingerprint pattern type identification method according to an embodiment of the present invention. Figure 13 As shown, the main processes include obtaining fingerprint data, model training, and model validation.

[0164] The specific working process and related explanations for fingerprint spectrum acquisition and preprocessing, model training and model verification can be found in the embodiments of the fingerprint spectrum type identification method described above, and will not be repeated here.

[0165] The fingerprint spectrum type identification method provided by this invention can analyze the entire fingerprint spectrum to be identified, thus overcoming the limitation of traditional methods that rely on identifying individual characteristic peaks. Therefore, the identification results are more accurate. Furthermore, this fingerprint spectrum type identification method is not limited by the sample source, is simpler and more reliable, and has a wider range of applications.

[0166] It should be noted that although the steps in the above embodiments are described in a specific order, those skilled in the art will understand that in order to achieve the effects of the present invention, different steps do not necessarily have to be executed in such an order. They can be executed simultaneously (in parallel) or in other orders, and these variations are all within the scope of protection of the present invention.

[0167] Those skilled in the art will understand that all or part of the processes in the method of the above embodiment of the present invention can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable file, or some intermediate form. The computer-readable storage medium can include any entity or device capable of carrying the computer program code, a medium, a USB flash drive, a portable hard drive, a magnetic disk, an optical disk, a computer memory, a read-only memory, a random access memory, an electrical carrier signal, a telecommunication signal, and a software distribution medium, etc.

[0168] Furthermore, the present invention also provides an electronic device. (See appendix.) Figure 14 , Figure 14 This is a schematic diagram of the main structure of an electronic device according to an embodiment of the present invention. Figure 14 As shown, the electronic device in this embodiment of the invention mainly includes a processor 1401 and a storage device 1402. The storage device 1402 can be configured to store a program for executing the fingerprint pattern type recognition method of the above-described method embodiments. The processor 1401 can be configured to execute the program in the storage device 1402, which includes, but is not limited to, a program for executing the fingerprint pattern type recognition method of the above-described method embodiments. For ease of explanation, only the parts related to the embodiments of the present invention are shown. For specific technical details not disclosed, please refer to the method section of the embodiments of the present invention.

[0169] In some possible embodiments of the present invention, the electronic device may include multiple processors 1401 and multiple storage devices 1402. The program executing the fingerprint pattern type recognition method of the above method embodiments can be divided into multiple subroutines. Each subroutine can be loaded and run by a processor 1401 to execute different steps of the fingerprint pattern type recognition method of the above method embodiments. Specifically, each subroutine can be stored in different storage devices 1402, and each processor 1401 can be configured to execute programs in one or more storage devices 1402 to jointly implement the fingerprint pattern type recognition method of the above method embodiments. That is, each processor 1401 executes different steps of the fingerprint pattern type recognition method of the above method embodiments to jointly implement the fingerprint pattern type recognition method of the above method embodiments.

[0170] The aforementioned multiple processors 1401 can be processors deployed on the same device. For example, the aforementioned electronic device can be a high-performance device composed of multiple processors, and the aforementioned multiple processors 1401 can be processors configured on that high-performance device. Alternatively, the aforementioned multiple processors 1401 can also be processors deployed on different devices. For example, the aforementioned electronic device can be a server cluster, and the aforementioned multiple processors 1401 can be processors on different servers within the server cluster; the aforementioned electronic device can be a driving device cluster, and the aforementioned multiple processors 1401 can be processors on different driving devices within the driving device cluster.

[0171] Furthermore, the present invention also provides a computer-readable storage medium. In one embodiment of the computer-readable storage medium according to the present invention, the computer-readable storage medium can be configured to store a program for executing the fingerprint pattern type identification method of the above-described method embodiments. This program can be loaded and run by a processor to implement the fingerprint pattern type identification method described above. For ease of explanation, only the parts related to the embodiments of the present invention are shown; for specific technical details not disclosed, please refer to the method section of the embodiments of the present invention. The computer-readable storage medium can be a storage device device comprising various electronic devices. Optionally, in the embodiments of the present invention, the computer-readable storage medium is a non-transitory computer-readable storage medium.

[0172] The technical solution of the present invention has been described above with reference to one embodiment shown in the accompanying drawings. However, it will be readily understood by those skilled in the art that the scope of protection of the present invention is obviously not limited to these specific embodiments. Without departing from the principles of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions resulting from such changes or substitutions will all fall within the scope of protection of the present invention.

Claims

1. A fingerprint pattern type identification method, characterized in that, The method includes: Obtain the fingerprint pattern to be identified; The fingerprint spectrum to be identified is input into a model set consisting of at least one effective recognition model to obtain a fingerprint spectrum type recognition result. The at least one effective recognition model is obtained by filtering the trained recognition models according to preset conditions. The recognition model is trained based on the corresponding ion peak cluster. The method further includes obtaining a model set, wherein obtaining the model set specifically includes: Multiple recognition models are trained to obtain multiple trained recognition models; Based on the preset conditions, the multiple trained recognition models are filtered to obtain at least one effective recognition model; The at least one effective recognition model is combined to obtain the model set; The training of multiple recognition models includes: Multiple training sets are obtained, wherein each training set is a corresponding ion peak cluster labeled with a type label; Each training set is input into the corresponding recognition model, and the corresponding recognition model is trained accordingly.

2. The method according to claim 1, characterized in that, The acquisition of multiple training sets includes: Obtain multiple fingerprint maps labeled with type tags; The multiple fingerprint spectra are normalized to obtain the set of ion peaks of the multiple fingerprint spectra; The set of ion peaks is divided into at least one ion peak cluster.

3. The method according to claim 2, characterized in that, The step of dividing the ion peak set into at least one ion peak cluster includes: Obtain the density of the distribution of multiple ion peaks in the ion peak set; The plurality of ion peaks are divided into at least one ion peak cluster according to the density.

4. The method according to claim 1, characterized in that, The preset conditions include at least a preset number of fingerprint maps and a preset accuracy threshold. The step of filtering the multiple trained recognition models according to the preset conditions includes: Determine whether the number of fingerprint spectra participating in the training of the multiple trained recognition models meets the preset number of fingerprint spectra. Determine whether the accuracy of each of the multiple recognition models meets the preset accuracy threshold. The identification model that simultaneously satisfies the preset number of fingerprint spectra and the preset accuracy threshold is selected as the effective identification model.

5. The method according to claim 1, characterized in that, The step of combining the at least one effective recognition model to obtain the model set includes: Multiple effective recognition models are combined in different ways to obtain multiple model sets composed of the multiple effective recognition models.

6. The method according to claim 5, characterized in that, The method further includes: Multiple recognition results are output based on the multiple model sets; An error function is calculated for the multiple recognition results; Based on the calculation results of the error function, each of the multiple model sets is checked to see if it meets the preset error threshold. Select one or more model sets that meet the preset error threshold as trained model sets.

7. An electronic device comprising a processor and a storage device, said storage device being adapted to store a plurality of program codes, characterized in that, The program code is adapted to be loaded and run by the processor to perform the fingerprint pattern type identification method according to any one of claims 1 to 6.

8. A computer-readable storage medium storing a plurality of program codes, characterized in that, The program code is adapted to be loaded and run by a processor to perform the fingerprint pattern type identification method according to any one of claims 1 to 6.

Images