A method and system for identifying an individual based on heart sound signals

Abstract

The present invention relates to the technical field of biometric information identification, and discloses an identity identification method based on heart sound wavelet analysis, comprising: obtaining verification heart sounds of a target individual, the verification heart sounds comprising at least a first heart sound signal and / or a second heart sound signal; preprocessing the verification heart sounds to obtain a heart sound verification signal; performing wavelet decomposition transformation on the heart sound verification signal according to a preset wavelet transformer to obtain a plurality of target frequency sub-bands; using a feature extractor to extract verification features from the plurality of target frequency sub-bands to obtain a plurality of heart sound verification features; comparing the plurality of heart sound verification features with a plurality of target features pre-stored in the target individual to calculate the similarity of the heart sound verification signal; judging whether the similarity of the heart sound verification signal is greater than a preset verification threshold; if so, judging that the verification heart sound has passed the verification; if not, judging that the verification heart sound has not passed the verification.

Description

Technical Field

[0001] The present invention relates to the field of biological information recognition, and in particular to an identity recognition method and system based on heart sound wavelet analysis. Background Art

[0002] With the popularization of the Internet of Things and big data systems, the security of identity recognition systems has received much attention. The security of traditional identity recognition modes such as face and fingerprint has begun to be questioned, and people have begun to look for more types of biometric signals as identity authentication methods to improve the reliability of identity recognition.

[0003] Heart sound signals are important biological signals of the human body. They are sound signals generated when blood flow causes the heart valves to open or close. At present, electronic stethoscopes and heart sound signal data acquisition equipment are becoming more and more mature. Through digital analysis, the pathological characteristics of heart sound signals can be extracted to assist doctors in making diagnostic decisions.

[0004] Currently, the analysis of heart sound signals is mainly focused on pathological research in the medical field, and few people have been involved in its application in the field of personal identity recognition, failing to fully tap the application value of heart sound signals in the field of the Internet of Things. Summary of the invention

[0005] The purpose of the present invention is to provide an identity recognition method and system based on heart sound signals, aiming to apply heart sound signals to identity verification and identification, and improve the reliability of the identity recognition system.

[0006] A first aspect of the present invention provides an identity recognition method based on heart sound wavelet analysis, comprising:

[0007] Acquire verification heart sounds of the target individual, wherein the verification heart sounds include at least a first heart sound signal and / or a second heart sound signal;

[0008] Preprocessing the verification heart sound to obtain a heart sound verification signal;

[0009] Performing wavelet decomposition transformation on the heart sound verification signal according to a preset wavelet transformer to obtain multiple target frequency sub-bands;

[0010] Using a feature extractor to extract verification features from the multiple target frequency sub-bands to obtain multiple heart sound verification features;

[0011] Comparing the plurality of heart sound verification features with a plurality of target features pre-stored in the target individual, and calculating the similarity of the heart sound verification signals;

[0012] Determining whether the similarity of the heart sound verification signal is greater than a preset verification threshold;

[0013] If yes, it is determined that the verification heart sound has passed the verification;

[0014] If not, it is determined that the verification heart sound has not passed the verification.

[0015] Preferably, the verified heart sounds include at least a first heart sound and a second heart sound.

[0016] Preferably, the comparing of the multiple heart sound verification features with the multiple target features pre-stored in the target individual to calculate the similarity of the heart sound verification signal includes: comparing the multiple heart sound verification features corresponding to the first heart sound with the multiple first feature values ​​pre-stored in the target individual to obtain a first similarity value; comparing the multiple heart sound verification features corresponding to the second heart sound with the multiple second feature values ​​pre-stored in the target individual to obtain a second similarity value; the heart sound verification signal similarity includes the weighted sum of the first similarity value and the second similarity value, and the sum of the weights of the first similarity value and the second similarity value in the heart sound verification signal similarity is greater than 70%.

[0017] Preferably, the verification heart sound is the heart sound signal of the target individual in the first posture; after the method, the method further includes: obtaining re-verification heart sounds, which are the heart sound signals of the target individual in the second posture; pre-processing the re-verification heart sounds to obtain heart sound re-verification signals; using a preset wavelet transformer to perform wavelet decomposition transform on the heart sound re-verification signal to obtain multiple target frequency sub-bands; using a feature extractor to perform re-verification feature extraction on the multiple target frequency sub-bands to obtain multiple heart sound re-verification features; comparing the multiple heart sound verification features with the multiple target features of the pre-stored target individual to calculate the similarity of the heart sound re-verification signal; if the similarity of the heart sound re-verification signal is greater than a preset re-verification threshold, it is judged that the heart sound sampling signal of the target individual has passed the verification.

[0018] Preferably, the weight proportion of the first similarity value in the similarity of the heart sound verification signal is greater than the weight proportion of the second similarity value in the similarity of the heart sound verification signal.

[0019] Preferably, the weight ratio of the first similarity value and the corresponding second similarity value in the similarity of the heart sound verification signal is 3:2.

[0020] Preferably, before performing wavelet decomposition transformation on the heart sound verification signal according to a preset wavelet transformer, the method further comprises: obtaining an individual secret key corresponding to the target individual; and determining the waveform and frequency of a mother wavelet of a preset wavelet analyzer according to the individual secret key.

[0021] Preferably, the heart sound verification feature includes an energy peak on the target frequency sub-band.

[0022] Preferably, the heart sound verification feature also includes energy entropy of the target frequency sub-band.

[0023] The second aspect of the present invention provides an identity recognition system based on heart sound wavelet analysis, the system specifically comprising: an acquisition module, used to acquire the verification heart sound of the target individual, the verification heart sound comprising at least a first heart sound signal and / or a second heart sound signal; a preprocessing module, used to preprocess the verification heart sound to obtain a heart sound verification signal; a decomposition and transformation module, used to perform a wavelet decomposition and transformation on the heart sound verification signal according to a preset wavelet transformer to obtain a plurality of target frequency sub-bands; a feature extraction module, used to perform a verification feature extraction on the plurality of target frequency sub-bands using a feature extractor to obtain a plurality of heart sound verification features; a calculation module, used to compare the plurality of heart sound verification features with a plurality of target features pre-stored in the target individual to calculate the similarity of the heart sound verification signal; a judgment module, used to judge whether the similarity of the heart sound verification signal is greater than a preset verification threshold, and to determine that the verification heart sound has passed the verification if the judgment is yes, and to confirm that the verification heart sound has not passed the verification if the judgment is no.

[0024] Compared with the prior art, the present invention adopts the wavelet analysis method to expand the application of heart sound signals to identity recognition, which significantly improves the accuracy and reliability of identity recognition results. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 It is a schematic diagram of a flow chart of an embodiment provided by an embodiment of the present invention;

[0026] Figure 2 It is a schematic diagram of a flow chart of an embodiment provided by an embodiment of the present invention;

[0027] Figure 3 It is a schematic diagram of a flow chart of an embodiment provided by an embodiment of the present invention;

[0028] Figure 4 It is a schematic diagram of an embodiment provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0029] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0030] The same or similar numbers in the drawings of this embodiment correspond to the same or similar parts; in the description of the present invention, it should be understood that if the terms "upper", "lower", "left", "right" and the like indicate directions or positional relationships based on the directions or positional relationships shown in the drawings, it is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific direction, be constructed and operated in a specific direction. Therefore, the terms describing the positional relationship in the drawings are only used for illustrative purposes and cannot be understood as limitations on this patent. For ordinary technicians in this field, the specific meanings of the above terms can be understood according to specific circumstances.

[0031] The implementation of the present invention is described in detail below in conjunction with specific embodiments. For ease of expression, the present invention refers to the user who needs to undergo identity authentication as the target individual, and it is assumed that the target individual has correctly pre-stored the characteristic information of his or her heart sound signal in the identity authentication system in advance. It should be noted that the first heart sound signal and the second heart sound signal in the present invention specifically refer to the first heart sound and the second heart sound divided according to the cardiac cycle in cardiac physiology, that is, the first heart sound is the mechanical wave corresponding to the beginning of the cardiac contraction period in the cardiac cycle, and the second heart sound is the mechanical wave corresponding to the contraction and relaxation of the myocardium in the cardiac cycle.

[0032] Reference Figure 1 The figure shows a basic embodiment of the present invention. The method for identity recognition based on heart sound wavelet analysis of the present invention comprises the following steps:

[0033] 101. Acquire verification heart sounds of a target individual, where the verification heart sounds include at least a first heart sound signal and / or a second heart sound signal;

[0034] When the identity of the target individual needs to be verified, a segment of the target individual's heart sound is collected as the verification heart sound. Generally speaking, the verification heart sound will include the heart sound corresponding to a complete cardiac cycle. However, depending on the actual situation of the acquisition device, the device may not be able to accurately collect the first heart sound and the second heart sound at the same time. Therefore, the verification heart sound can include only a first heart sound signal, or only a second heart sound signal, or a combination of the two.

[0035] 102. Preprocess the verification heart sound to obtain a heart sound verification signal;

[0036] The verification heart sounds may contain clutter, so the verification heart sounds are preprocessed first to facilitate subsequent wavelet decomposition and analysis. The preprocessing may include operations such as filtering, denoising, and signal enhancement.

[0037] 103. Perform wavelet decomposition transformation on the heart sound verification signal according to a preset wavelet transformer to obtain multiple target frequency sub-bands;

[0038] The heart sound verification signal is subjected to wavelet decomposition transformation using a preset wavelet transformer to obtain multiple target frequency sub-bands corresponding to the heart sound verification signal. Since the heart sound is a nonlinear and non-stationary signal, it is suitable to use wavelet decomposition to decompose and transform it into multiple target frequency sub-bands for analysis. The selection of the mother wave of the wavelet decomposition transformation has an important influence on the result of the decomposition transformation. In the basic scheme of this embodiment, a mother wave with a fixed waveform and frequency can be used for all users based on experience to obtain the best decomposition transformation result.

[0039] 104. Using a feature extractor to extract verification features from multiple target frequency sub-bands to obtain multiple heart sound verification features;

[0040] By using a feature extractor formed by a neural network model, verification features are extracted for each target frequency sub-band obtained in step 103, so as to obtain a plurality of heart sound verification features corresponding to the verification heart sound. The heart sound verification feature corresponding to each target frequency sub-band may include one or more feature values ​​corresponding to the target frequency sub-band, such as energy peak value, energy distribution shape, information entropy, etc.

[0041] 105. Compare the multiple heart sound verification features with the multiple target features pre-stored in the target individual, and calculate the similarity of the heart sound verification signal;

[0042] The heart sound verification features corresponding to the verified heart sounds are compared with multiple target features pre-stored in the target individual stored in the verification system. The specific comparison method can be to compare the extracted heart sound verification features with the multiple target features pre-stored in the target individual one by one according to the corresponding target frequency sub-bands, and obtain a comprehensive value as the similarity of the heart sound verification signal.

[0043] 106. Determine whether the similarity of the heart sound verification signal is greater than a preset verification threshold;

[0044] The heart sound verification signal similarity quantified in step 105 is compared with a preset verification threshold to confirm whether the heart sound verification signal similarity meets the required level. The verification threshold may be a fixed value derived from multiple experiences.

[0045] 107. If yes, then it is determined that the verification of the heart sound has passed the verification;

[0046] If the similarity of the heart sound verification signal is greater than the preset verification threshold, it can be confirmed that the obtained verification heart sound matches multiple target features pre-stored in the target individual, and the two are most likely from the same person.

[0047] 108. If not, it is determined that the verification heart sound has not passed the verification.

[0048] If the similarity of the heart sound verification signal is less than or equal to the preset verification threshold, it is considered that the verification heart sound used for identity recognition does not match the target features pre-stored in the target individual to an insufficient degree, and it cannot be confirmed that the two are from the same person.

[0049] In order to better demonstrate the principles and effects of the present invention, in another embodiment, the present invention places the focus of verification on the utilization of the first heart sound and the second heart sound to verify the heart sound. Taking into account that the heart sound collection is difficult and there are more noises, and the first heart sound and the second heart sound have relatively obvious waveforms in a heart sound cycle, thus ensuring that the verified heart sound contains both the first heart sound and the second heart sound will have a better recognition effect. As a preferred implementation, the verified heart sound includes a complete heart sound cycle, or at least includes the first heart sound and the second heart sound in the same heart sound cycle. In order to illustrate the specific utilization of the first heart sound and the second heart sound, based on Figure 1 Step 105 of the embodiment shown is described in detail. Step 105 may specifically include:

[0050] 1051. Compare a plurality of heart sound verification features corresponding to the first heart sound with a plurality of first feature values ​​pre-stored in the target individual to obtain a first similarity value;

[0051] 1052. Compare the multiple heart sound verification features corresponding to the second heart sound with the multiple second feature values ​​pre-stored in the target individual to obtain a second similarity value;

[0052] In this embodiment, the similarity of the heart sound verification signal is to extract the heart sound verification feature of the entire verification heart sound, compare multiple target features pre-stored in the target individual in each target frequency sub-band, and finally perform weighted summation. Since the first heart sound and the second heart sound features are more obvious, the first heart sound and the second heart sound will occupy a relatively important position in the verification of the heart sound. A recommended parameter is that the weight of the heart sound verification feature corresponding to the first heart sound and the heart sound verification feature corresponding to the second heart sound should account for at least 70% of the weight of the entire verification of the heart sound.

[0053] Considering that the first heart sound has a greater intensity and is easier to collect, the waveform of the first heart sound may be more valuable for reference than the waveform of the second heart sound. Therefore, as a recommended implementation method, when calculating the heart sound verification features of the verification heart sound, the first heart sound and the second heart sound account for about 3:2, and together account for more than 70%.

[0054] Based on any of the above embodiments, in another embodiment of the present invention, the security of heart sound verification is improved. Since heart sound is a biological signal, it has the risk of being stolen and copied just like human face and fingerprint. Therefore, the present invention adds an encryption measure in the wavelet decomposition transformation process of heart sound. Figure 2 , the embodiment comprises:

[0055] 201. Acquire verification heart sounds of a target individual, where the verification heart sounds include at least a first heart sound signal and / or a second heart sound signal;

[0056] 202. Preprocess the verification heart sound to obtain a heart sound verification signal;

[0057] Steps 201 and 202 of this embodiment are Figure 1 Steps 101 and 102 of the embodiment are similar and will not be described again here.

[0058] 203. Obtain the individual secret key corresponding to the target individual;

[0059] Obtain the secret key corresponding to the target individual, which can be a physical device with physical meaning, such as a USB flash drive, IC card or digital password card, or a non-physical password, number, etc. In some specific implementations, the individual secret key can be one of fingerprints, irises or voice features.

[0060] 204. Determine the waveform and frequency of a mother wavelet of a preset wavelet analyzer according to the individual secret key;

[0061] According to the individual secret key, a waveform function is confirmed as the mother wavelet of the preset wavelet analyzer. The subsequent work of the wavelet analyzer will be based on this mother wavelet to decompose and transform the signal.

[0062] 205. Perform wavelet decomposition transformation on the heart sound verification signal according to a preset wavelet transformer to obtain multiple target frequency sub-bands;

[0063] 206. Using a feature extractor to extract verification features from multiple target frequency sub-bands to obtain multiple heart sound verification features;

[0064] 207. Compare the multiple heart sound verification features with the multiple target features pre-stored in the target individual, and calculate the similarity of the heart sound verification signal;

[0065] 208. Determine whether the similarity of the heart sound verification signal is greater than a preset verification threshold;

[0066] 209. If yes, then it is determined that the verification of the heart sound has passed the verification;

[0067] 210. If not, it is determined that the verification heart sound has not passed the verification.

[0068] Steps 205 to 210 of this embodiment are Figure 1Steps 103 to 108 of the illustrated embodiment are similar and will not be described again here. In this embodiment, the identity recognition system only stores multiple target features of the target individual obtained under a specific mother wavelet decomposition transform, but does not store complete heart sounds, thereby enhancing the protection of the target individual's personal privacy. During identity authentication, the target individual or other third party is also required to provide an individual secret key in order to correctly decompose and transform the target individual's verification heart sounds, which greatly improves the security of information transmission in identity authentication.

[0069] The collection of heart sound signals is often accompanied by a lot of noise and is easily affected by the individual's posture. Therefore, in another embodiment, a re-test step is added when the heart sound verification fails. In this embodiment, the verified heart sound is the heart sound signal of the target individual in the first posture. For details, please refer to Figure 3 , the specific steps of this embodiment are as follows:

[0070] 301. Acquire verification heart sounds of a target individual, where the verification heart sounds include at least a first heart sound signal and / or a second heart sound signal;

[0071] 302. Preprocess the verification heart sound to obtain a heart sound verification signal;

[0072] 303. Perform wavelet decomposition transformation on the heart sound verification signal according to a preset wavelet transformer to obtain multiple target frequency sub-bands;

[0073] 304. Using a feature extractor, extract verification features from multiple target frequency sub-bands to obtain multiple heart sound verification features;

[0074] 305. Compare the multiple heart sound verification features with the multiple target features pre-stored in the target individual, and calculate the similarity of the heart sound verification signal;

[0075] 306. Determine whether the similarity of the heart sound verification signal is greater than a preset verification threshold;

[0076] 307. If yes, then it is determined that the verification of the heart sound has passed the verification;

[0077] 308. If no, it is determined that the verification heart sound has not passed the verification;

[0078] Steps 301 to 308 of this embodiment are Figure 1 Steps 103 to 108 of the embodiment are similar and will not be described again here.

[0079] 309. Obtain retest heart sounds;

[0080] If the verification heart sound of the target individual is not sufficient to prove the identity of the target individual, then a re-verification heart sound that is different from the verification heart sound can be obtained to confirm the identity of the target individual again.

[0081] 310. Preprocess the retest heart sound to obtain a heart sound retest signal;

[0082] The repeated heart sounds may also contain clutter, so the repeated heart sounds are preprocessed for subsequent wavelet decomposition and analysis. The preprocessing may include operations such as filtering, denoising and signal enhancement.

[0083] 311. Use a preset wavelet transformer to perform wavelet decomposition transformation on the heart sound verification signal to obtain multiple target frequency sub-bands;

[0084] The heart sound recheck signal is subjected to wavelet decomposition transformation using a preset wavelet transformer to obtain a plurality of target frequency sub-bands corresponding to the heart sound recheck signal. The recheck signal and the aforementioned verification signal usually use the same wavelet transformer.

[0085] 312. Use a feature extractor to extract re-test features of multiple target frequency sub-bands to obtain multiple heart sound re-test features;

[0086] By using a feature extractor formed by a neural network model, multiple heart sound re-verification features corresponding to the verified heart sounds can be obtained by performing re-verification feature extraction on each target frequency sub-band obtained in step 312. The heart sound re-verification features corresponding to each target frequency sub-band may include one or more feature values ​​corresponding to the target frequency sub-band, such as energy peak value, energy distribution shape, information entropy, etc.

[0087] 313. Compare the multiple heart sound re-test features with the multiple target features of the pre-stored target individual, and calculate the similarity of the heart sound re-test signals;

[0088] The heart sound verification features corresponding to the repeated heart sounds are compared with a plurality of target features pre-stored in the system for the target individuals. The comparison method can refer to the steps of the heart sound verification features described above.

[0089] 314. Determine whether the similarity of the heart sound re-test signal is greater than a preset re-test threshold;

[0090] The heart sound retest signal similarity quantified in step 313 is compared with a preset retest threshold to confirm whether the heart sound retest signal similarity meets the required level. The retest threshold may be a fixed value derived from multiple experiences.

[0091] 315. If yes, then the retested heart sounds are judged to have passed the verification.

[0092] If the heart sound re-test signal is greater than the preset verification threshold, it can be confirmed that the obtained re-test heart sound matches multiple target features pre-stored in the target individual, and the two are most likely from the same person.

[0093] The retesting of the heart sound in this embodiment is to collect the heart sound signal of the target individual in a second posture different from the first posture.

[0094] In some embodiments, the first posture is a normal standing posture for the convenience of the user. The second posture is to raise the right arm to enhance the strength of the first and second heart sounds and reduce clutter. In other embodiments, the reason for the failure of the verification of the heart sound may be that the first heart sound and the second heart sound correspond to different optimal collection points on the human body, so the user needs to collect them twice respectively. In this case, the first posture corresponds to the heart sound signal collected at one position of the target individual, and the second posture is the heart sound signal collected at another position of the target individual.

[0095] In order to implement the above method, the present invention also provides an identity recognition system based on heart sound wavelet analysis. Figure 4 The system comprises:

[0096] An acquisition module 401 is used to acquire verification heart sounds of a target individual, where the verification heart sounds include at least a first heart sound signal and / or a second heart sound signal;

[0097] A preprocessing module 402 is used to preprocess the verification heart sound to obtain a heart sound verification signal;

[0098] A decomposition and transformation module 403 is used to perform wavelet decomposition and transformation on the heart sound verification signal according to a preset wavelet transformer to obtain multiple target frequency sub-bands;

[0099] A feature extraction module 404 is used to extract verification features from multiple target frequency sub-bands using a feature extractor to obtain multiple heart sound verification features;

[0100] A calculation module 405 is used to compare the multiple heart sound verification features with the multiple target features pre-stored in the target individual to calculate the similarity of the heart sound verification signal;

[0101] The judgment module 406 is used to judge whether the similarity of the heart sound verification signal is greater than a preset verification threshold, and determine that the verification heart sound has passed the verification if the judgment is yes, and confirm that the verification heart sound has not passed the verification if the judgment is no.

[0102] Those skilled in the art can clearly understand that for the convenience and brevity of description, Figure 3 The specific working process of the identity recognition system based on heart sound wavelet analysis shown can refer to the corresponding process in the aforementioned embodiment of the identity recognition method based on heart sound wavelet analysis, which will not be repeated here.

[0103] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. An identity recognition method based on heart sound wavelet analysis, characterized in that: include: Obtain verification heart sounds of the target individual; Preprocessing the verification heart sound to obtain a heart sound verification signal; The verification heart sound includes at least a first heart sound and a second heart sound; Performing wavelet decomposition transformation on the heart sound verification signal according to a preset wavelet transformer to obtain a plurality of target frequency sub-bands; Using a feature extractor to extract verification features from the multiple target frequency sub-bands to obtain multiple heart sound verification features; Comparing the plurality of heart sound verification features with a plurality of target features pre-stored in the target individual, and calculating the similarity of the heart sound verification signals; Determining whether the similarity of the heart sound verification signal is greater than a preset verification threshold; If yes, it is determined that the verification heart sound has passed the verification; If not, it is determined that the verification heart sound has not passed the verification; The step of comparing the plurality of heart sound verification features with a plurality of target features pre-stored in the target individual to calculate the similarity of the heart sound verification signals includes: Comparing a plurality of heart sound verification features corresponding to the first heart sound with a plurality of first feature values ​​pre-stored in the target individual to obtain a first similarity value; Comparing the plurality of heart sound verification features corresponding to the second heart sound with the plurality of second feature values ​​pre-stored in the target individual to obtain a second similarity value; The heart sound verification signal similarity includes a weighted sum of the first similarity value and the second similarity value, and a sum of weight proportions of the first similarity value and the second similarity value in the heart sound verification signal similarity is greater than 70%.

2. The identity recognition method based on heart sound wavelet analysis according to claim 1 is characterized in that: The verification heart sound is a heart sound signal of the target individual in the first posture; After the method, the method further comprises: Acquire a retest heart sound, where the retest heart sound is a heart sound signal of the target individual in the second posture; Preprocessing the repeated heart sounds to obtain a heart sound repeated signal; Using a preset wavelet transformer, performing wavelet decomposition transformation on the heart sound verification signal to obtain a plurality of target frequency sub-bands; Using a feature extractor to perform re-test feature extraction on the multiple target frequency sub-bands to obtain multiple heart sound re-test features; Comparing the plurality of heart sound verification features with a plurality of pre-stored target features of a target individual, and calculating the similarity of the heart sound verification signal; If the similarity of the heart sound re-verification signal is greater than a preset re-verification threshold, it is determined that the heart sound sampling signal of the target individual has passed the verification.

3. The identity recognition method based on heart sound wavelet analysis according to claim 1 is characterized in that: The weight proportion of the first similarity value in the similarity of the heart sound verification signal is greater than the weight proportion of the second similarity value in the similarity of the heart sound verification signal.

4. The identity recognition method based on heart sound wavelet analysis according to claim 2 is characterized in that: The weight ratio of the first similarity value to the second similarity value in the similarity of the heart sound verification signal is 3:

2.

5. The identity recognition method based on heart sound wavelet analysis according to claim 1 is characterized in that: Before performing wavelet decomposition transformation on the heart sound verification signal according to a preset wavelet transformer, the method further includes: Obtain the individual key corresponding to the target individual; According to the individual secret key, the waveform and frequency of the mother wavelet of the preset wavelet analyzer are determined.

6. The identity recognition method based on heart sound wavelet analysis according to claim 1 is characterized in that: The heart sound verification feature includes an energy peak on the target frequency sub-band.

7. The identity recognition method based on heart sound wavelet analysis according to claim 6 is characterized in that: The heart sound verification feature also includes energy entropy of the target frequency sub-band.

8. An identity recognition system based on heart sound wavelet analysis, characterized in that: include: An acquisition module, used to acquire verification heart sounds of a target individual; A preprocessing module, used for preprocessing the verification heart sound to obtain a heart sound verification signal; The verification heart sound includes at least a first heart sound and a second heart sound; A decomposition and transformation module, used for performing wavelet decomposition and transformation on the heart sound verification signal according to a preset wavelet transformer to obtain a plurality of target frequency sub-bands; A feature extraction module, configured to extract verification features from the plurality of target frequency sub-bands using a feature extractor to obtain a plurality of heart sound verification features; A calculation module, used for comparing the plurality of heart sound verification features with a plurality of target features pre-stored in the target individual, and calculating the similarity of the heart sound verification signal; The step of comparing the plurality of heart sound verification features with a plurality of target features pre-stored in the target individual to calculate the similarity of the heart sound verification signals includes: Comparing a plurality of heart sound verification features corresponding to the first heart sound with a plurality of first feature values ​​pre-stored in the target individual to obtain a first similarity value; Comparing the plurality of heart sound verification features corresponding to the second heart sound with the plurality of second feature values ​​pre-stored in the target individual to obtain a second similarity value; The heart sound verification signal similarity includes a weighted sum of the first similarity value and the second similarity value, and the sum of the weights of the first similarity value and the second similarity value in the heart sound verification signal similarity is greater than 70%; The judgment module is used to judge whether the similarity of the heart sound verification signal is greater than a preset verification threshold, and determine that the verification heart sound has passed the verification if the judgment is yes, and confirm that the verification heart sound has not passed the verification if the judgment is no.

Images