Earphone anti-mis-touch control method and system based on fingerprint recognition

Abstract

The application discloses a kind of earphone anti-misoperation control method and system based on fingerprint identification, it is related to the field of fingerprint identification, including, user static fingerprint data and dynamic fingerprint data are collected and preprocessed and stored in fingerprint database;Based on the real-time acquisition user dynamic fingerprint data, generate elliptical dynamic fingerprint window;Elliptical dynamic fingerprint window is converted into rectangular dynamic fingerprint window, and fingerprint feature point matching is carried out;User earphone use data is collected, and touch scene analysis model is constructed, and earphone use scene is obtained;According to earphone use scene, dynamically adjust the operation logic of earphone.The application generates rectangular dynamic fingerprint window, accurately captures the dynamic change of user finger in the touch process, can adapt to the change of finger in different operation, realizes the effective tracking and identification of real-time dynamic fingerprint, improves the accuracy of fingerprint matching, reduces misoperation, and enhances the adaptive capacity of system to complex touch operation.

Description

TECHNICAL FIELD

[0001] The present application relates to the fingerprint identification technical field, especially to a kind of earphone anti-misoperation control method and system based on fingerprint identification. BACKGROUND

[0002] With the rapid development of wireless communication technology and portable audio equipment, earphone has become an indispensable device in people's daily life, and early earphone is mainly used for wired connection equipment, and with the popularity of Bluetooth technology, wireless earphone gradually replaces traditional wired earphone, and provides more free use experience, however, with the continuous increase of earphone functions, such as touch operation, voice assistant and noise elimination, the use of earphone becomes more and more complex and intelligent, in this case, the convenience and accuracy of touch operation become one of the important evaluation criteria of user experience, and the existing touch earphone mainly relies on static touch sensor to realize the control of earphone by detecting touch action.

[0003] However, due to the diversity and complexity of user's touch mode, the existing technology has certain deficiencies in the accuracy and sensitivity of fingerprint identification, how to effectively prevent misoperation and improve the accuracy of fingerprint identification and user satisfaction has become a difficult problem to be solved in the current technical field. SUMMARY

[0004] In view of the above problems existing in the prior art earphone anti-misoperation control method based on fingerprint identification, the present application is proposed.

[0005] Therefore, the problem to be solved by the present application is that the identity verification is not accurate enough.

[0006] To solve the above technical problems, the present application provides the following technical scheme: an earphone anti-misoperation control method based on fingerprint identification, which comprises: collecting user static fingerprint data and dynamic fingerprint data, preprocessing and storing in a fingerprint database; generating an elliptical dynamic fingerprint window based on real-time collected user dynamic fingerprint data; converting the elliptical dynamic fingerprint window into a rectangular dynamic fingerprint window, and performing fingerprint data matching; collecting user earphone use data, constructing a touch scene analysis model, and obtaining an earphone use scene; and dynamically adjusting the operation logic of the earphone according to the earphone use scene.

[0007] As a preferred scheme of the earphone anti-misoperation control method based on fingerprint identification, wherein: the collection of user static fingerprint data and dynamic fingerprint data for preprocessing and storage in the fingerprint database uses earphone matching application program to collect user fingerprint image, removes noise in the fingerprint image through filter, enhances the contrast of the fingerprint image using image enhancement technology, and converts the gray image into binary image;

[0008] The ridge line in the binary image is thinned to single-pixel width using a Zhang-Suen thinning algorithm, static feature points including ridge line endpoints and bifurcation points are extracted from the thinned fingerprint image, and description information including position and direction is generated for each static feature point;

[0009] The extracted static feature point information is integrated into a unified format including the position, direction and type of each static feature point, and a static fingerprint data feature vector is generated, the static fingerprint data feature vector including the position coordinates of the static feature points, the direction angle and the number of static feature points;

[0010] A capacitive touch sensor and a pressure sensor are installed on the earphone surface touch area;

[0011] When the user uses the earphone, real-time dynamic fingerprint data is collected, the dynamic fingerprint data including time series data generated during user operation, recording the current fingerprint image, finger position, pressure intensity and touch type at each time point, the touch type including light touch, long press and sliding;

[0012] The fingerprint position coordinates and pressure values in the dynamic fingerprint data are normalized to generate a dynamic fingerprint data feature vector, the dynamic fingerprint data feature vector including dynamic fingerprint position, pressure value and time stamp;

[0013] The static fingerprint data feature vector and the dynamic fingerprint data feature vector are normalized respectively and spliced to form a multi-dimensional fusion data vector, and a fingerprint feature template is generated based on the multi-dimensional fusion data vector and stored in a database.

[0014] As a preferred scheme of the earphone anti-mis-touch control method based on fingerprint recognition, wherein: based on the real-time collected user dynamic fingerprint data, an elliptical dynamic fingerprint window is generated, including,

[0015] The dynamic fingerprint window parameters are calculated, and the current touch point position of the user's finger is set as (x i-1 , y i-1 ), and the next touch point position is (x i , y i );

[0016] Collect the user's earphone historical touch data and collect N touch point data;

[0017] The distance difference between each touch is calculated:

[0018]

[0019] Wherein, (x i ′, y i ′) is the position of the i-th touch of the finger in the user's earphone historical touch data;

[0020] Calculate the mean value of all touch point distance

[0021]

[0022] Using PDR algorithm, predict the next touch point (x i , y i ) position after user finger movement:

[0023]

[0024] Where θ i is the direction angle of user finger movement;

[0025] Calculate the straight line distance c i between two touch points:

[0026]

[0027] Using finger movement distance c i and unit time interval Δt to calculate current speed V i :

[0028]

[0029] Calculate all speed values during finger movement and record the maximum value V max ;

[0030] Calculate the short semi-axis b i :

[0031] b i = (V max -V i ) × Δt + e

[0032] Where b i is the vertical distance of user touch point position change at current speed, V max is the maximum speed, V i is the current speed, and e is the empirical error;

[0033] Calculate the long semi-axis a i :

[0034]

[0035] Where a i is the maximum expansion distance of the elliptical window;

[0036] According to the long semi-axis a i , the short semi-axis b i , and the focal distance c iGenerate an elliptical dynamic fingerprint window, the center of which is (x i y i ).

[0037] As a preferred embodiment of the fingerprint recognition-based headphone anti-mistouch control method of the present invention, wherein: the step of converting the elliptical dynamic fingerprint window into a rectangular dynamic fingerprint window includes,

[0038] The range of the rectangular dynamic fingerprint window is calculated by including the number of grid cells to the left, right, up, and down.

[0039] Number of grid cells to shift left:

[0040]

[0041] Number of grid cells to shift to the right:

[0042]

[0043] Number of grid cells moved up:

[0044]

[0045] Number of grid cells to move down:

[0046]

[0047] Where, N ileft N is the number of grid cells on the left boundary. iright N is the number of grid cells on the right boundary. ibottom N is the number of grid cells at the upper boundary. itop α is the number of grid cells at the lower boundary, α is the rotation angle, and cellsize is the size of the grid.

[0048] Based on the boundaries of the rectangular dynamic fingerprint window, the data of the elliptical dynamic fingerprint window is mapped onto the rectangular dynamic fingerprint window to generate a rectangular dynamic fingerprint window that contains the elliptical dynamic fingerprint window.

[0049] As a preferred embodiment of the fingerprint recognition-based headphone anti-mistouch control method of the present invention, wherein: the fingerprint feature point matching includes,

[0050] Collect real-time dynamic fingerprint data, construct a rectangular dynamic fingerprint window, record all rectangular dynamic fingerprint windows that change with the finger, select the rectangular dynamic fingerprint window whose corner is closest to the center point, and record the coordinates (x, y) of the top left corner. min y min ) and the coordinates of the lower right corner (x max y max Check the coordinates of the extracted feature points, and ensure that x... min <x<xmax and y min <y<y max The feature points of the condition are reserved, and the feature points not meeting the condition are filtered;

[0051] The feature points p of the real-time dynamic fingerprint data and the feature points q of each fingerprint template in the fingerprint database are extracted respectively, the Euclidean distance between each feature point p and the feature point q is calculated, the feature point pair with the minimum Euclidean distance is selected as the final matching feature point pair, and the fingerprint matching degree S is calculated based on the Euclidean distance between the final matching feature point pair:

[0052]

[0053] Wherein M is the number of matching feature point pairs, dist(p f , q f ) is the Euclidean distance between the matching feature point pair (p f , q f ), p f is the feature point of the real-time dynamic fingerprint data in the fth matching feature point pair, and q f is the fingerprint feature template feature point in the fth matching feature point pair.

[0054] The fingerprint matching degree threshold St is set, and compared with the fingerprint matching degree S:

[0055] If S≥St, it is determined that the user identity verification is successful, and the verification result is sent to the mobile terminal application program through Bluetooth, the mobile terminal application program displays a verification success message, and the user is allowed to continue operation;

[0056] If S<St, it is determined that the user identity verification fails, the earphone automatically triggers the anti-mis-touch mechanism, locks the touch operation, restores the touch function after 5 seconds, sends the verification information to the mobile terminal application program through Bluetooth, and issues a warning to the user.

[0057] As a preferred scheme of the earphone anti-mis-touch control method based on fingerprint identification, the user earphone use data is collected, a touch scene analysis model is constructed, and the earphone use scene is obtained, which includes,

[0058] The earphone historical use data of the mobile terminal application program is collected, an earphone historical use data set is generated and divided into a training set and a test set, the earphone historical use data set includes feature parameters and label data, the feature parameters include average acceleration, angular velocity of earphone rotation, ambient light intensity, ambient noise level and heart rate variation, and the label data refers to the earphone use scene corresponding to each touch event, including emergency scene, motion scene and static scene;

[0059] Extract the earphone historical use data set feature parameters to form a high-dimensional feature matrix Q:

[0060]

[0061] Wherein is the average acceleration, w is the angular velocity of earphone rotation, O is the ambient light intensity, U is the ambient noise level, ΔHR is the heart rate variation, m is the number of samples;

[0062] The touch scene analysis model is constructed using a convolutional neural network, including an input layer, a convolutional layer, a pooling layer, a fully connected layer, and an output layer;

[0063] The input content of the input layer is a high-dimensional feature matrix Q;

[0064] The convolutional layer extracts features through multiple convolutional layers, each convolutional layer consisting of multiple convolutional kernels and activation functions, generating a feature map;

[0065] The pooling layer pools the feature map output by the convolutional layer, reduces the dimension of the feature map, and extracts the main features;

[0066] The fully connected layer flattens the output of the pooling layer and performs feature combination and classification through the fully connected layer;

[0067] The output layer uses a Softmax activation function for multi-classification, and outputs the probability distribution of the earphone use scenario;

[0068] The model is trained using the training set data, the cross-entropy loss function is used to calculate the model error, and the Adam optimizer is used to update the parameters;

[0069] The performance of the model is evaluated using the test set, and the model is optimized through hyperparameter tuning and cross-validation;

[0070] The real-time collected earphone use data of the user is input into the trained model to obtain the earphone use scenario.

[0071] As a preferred scheme of the earphone anti-mis-touch control method based on fingerprint recognition, wherein:

[0072] The earphone use scenario includes cycling, driving, walking, running, and quiet scenarios, and the earphone use scenario priority is set;

[0073] The user is in an emergency scenario when cycling and driving, which belongs to the highest priority, the user's attention is focused on the road and traffic conditions, the touch control is disabled and the voice control is enabled, and the earphone ambient sound mode is turned on;

[0074] The user is in a walking state and a running state, which is a sports scene, belongs to a sub-high priority, the user's body and hands are in a sports state, a sports mode is started, music is automatically played when the user starts sports, and the music is automatically paused when the user stops sports;

[0075] The user is in a static state, which is a quiet scene, belongs to a general priority, and the user normally uses the earphone.

[0076] As a preferred scheme of the earphone anti-mis-touch control system based on fingerprint recognition, the system comprises,

[0077] The data acquisition and preprocessing module is configured to acquire static fingerprint data and dynamic fingerprint data of the user and pre-process the fingerprint.

[0078] The window conversion module is configured to generate an elliptical dynamic fingerprint window and convert the elliptical dynamic fingerprint window into a rectangular dynamic fingerprint window according to the acquired fingerprint data of the user.

[0079] The fingerprint matching module is configured to perform fingerprint feature point matching between the real-time acquired dynamic fingerprint data and fingerprint data in the fingerprint database.

[0080] The data analysis module is configured to construct a touch dynamic analysis model and analyze the earphone use scene of the user.

[0081] The dynamic adjustment module is configured to dynamically adjust the operation logic of the earphone according to the earphone use scene.

[0082] A computer device comprises a memory and a processor, the memory stores a computer program, and the processor implements the steps of the earphone anti-mis-touch control method and system based on fingerprint recognition when executing the computer program.

[0083] A computer readable storage medium stores a computer program, and the computer program implements the steps of the earphone anti-mis-touch control method and system based on fingerprint recognition when executed by a processor.

[0084] The present application has the advantages that: the rectangular dynamic fingerprint window is generated, the dynamic change of the user's finger in the touch process is accurately captured, the generated rectangular dynamic fingerprint window can adapt to the change of the finger in different operations, the effective tracking and recognition of the real-time dynamic fingerprint are realized, the accuracy and real-time performance of the fingerprint matching are improved, the possibility of mis-touch is reduced, and the adaptability of the system to complex touch operations is enhanced. BRIEF DESCRIPTION OF DRAWINGS

[0085] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings needed to be used in the embodiments will be briefly introduced as follows. Obviously, the drawings in the following description only constitute some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative effort on the basis of these drawings.

[0086] Figure 1 A flowchart of a method for preventing earphone from being touched by mistake based on fingerprint recognition.

[0087] Figure 2 A flowchart of fingerprint matching.

[0088] Figure 3 A structural diagram of a system for preventing earphone from being touched by mistake based on fingerprint recognition. DETAILED DESCRIPTION

[0089] In order to make the above objectives, features and advantages of the present application more apparent, the specific embodiments of the present application will be described in detail below with reference to the accompanying drawings.

[0090] In the following description, a large number of specific details are set forth in order to facilitate a thorough understanding of the present application, but the present application can also be implemented in other ways different from the description, and those skilled in the art can make similar generalizations without departing from the concept of the present application, so the present application is not limited to the specific embodiments disclosed below.

[0091] Secondly, the "one embodiment" or "embodiment" referred to herein means that the specific features, structures or characteristics can be included in at least one implementation of the present application. The "in one embodiment" appearing in different places in the specification does not mean the same embodiment, nor is it an embodiment that is independent of or selectively excludes other embodiments.

[0092] Embodiment 1

[0093] Reference Figure 1 and Figure 2 The first embodiment of the present application provides a method and system for preventing earphone from being touched by mistake based on fingerprint recognition. The method for preventing earphone from being touched by mistake based on fingerprint recognition comprises the following steps.

[0094] S1, collecting user static fingerprint data and dynamic fingerprint data, preprocessing and storing in a fingerprint database;

[0095] Further, the static fingerprint data and the dynamic fingerprint data of the user are collected, preprocessed and stored in the fingerprint database. When the user uses the earphone for the first time, the earphone application collects the fingerprint image of the user, removes the noise in the fingerprint image through a filter, enhances the contrast of the fingerprint image using image enhancement technology, and converts the gray image into a binary image.

[0096] The ridge line in the binary image is thinned to a single-pixel width using the Zhang-Suen thinning algorithm, and the static feature points including the ridge line endpoints and bifurcation points are extracted from the thinned fingerprint image. Description information is generated for each static feature point, including position and direction.

[0097] The extracted static feature point information is integrated into a unified format, including the position, direction and type of each static feature point, and a static fingerprint data feature vector is generated. The static fingerprint data feature vector includes the position coordinates of the static feature points, the direction angle and the number of static feature points.

[0098] A capacitive touch sensor and a pressure sensor are installed on the surface touch area of the earphone.

[0099] When the user uses the earphone, real-time dynamic fingerprint data is collected, including time series data generated during user operation, current fingerprint image, finger position, pressure intensity and touch type recorded at each time point. The touch type includes tapping, long pressing and sliding.

[0100] The fingerprint position coordinates and pressure values in the dynamic fingerprint data are normalized to generate a dynamic fingerprint data feature vector. The dynamic fingerprint data feature vector includes dynamic fingerprint position, pressure value and time stamp.

[0101] The static fingerprint data feature vector and the dynamic fingerprint data feature vector are normalized and spliced to form a multi-dimensional fusion data vector. A fingerprint feature template is generated based on the multi-dimensional fusion data vector and stored in the database.

[0102] By collecting the static fingerprint data and the dynamic fingerprint data of the user and preprocessing them, the fingerprint information of the user can be efficiently stored in the fingerprint database, realizing high-precision collection and storage of user fingerprint data, providing a basis for subsequent fingerprint matching and dynamic adjustment. Through high-precision data storage, the system can provide stable and reliable fingerprint recognition performance in different environments, thereby achieving the effect of reducing false touch.

[0103] S2, based on the real-time collected dynamic fingerprint data of the user, an elliptical dynamic fingerprint window is generated;

[0104] Further, based on the real-time collected dynamic fingerprint data of the user, an elliptical dynamic fingerprint window is generated, including

[0105] Calculate the dynamic fingerprint window parameters, set the current touch point position of the user's finger as (x i-1 , y i-1 ), and the next touch point position as (x i , y i );

[0106] Collect the user's earphone historical touch data and collect N touch point data;

[0107] Calculate the distance difference between each touch:

[0108]

[0109] Where (x i ', y i ') is the position of the i-th touch of the user's finger in the earphone historical touch data;

[0110] Calculate the mean value of the distance between all touch points

[0111]

[0112] Use the PDR algorithm to predict the next touch point (x i , y i ) position after the user's finger moves:

[0113]

[0114] Where θ i is the direction angle of the user's finger movement;

[0115] Calculate the straight line distance c i between the two touch points:

[0116]

[0117] Use the finger movement distance c i and the unit time interval Δt to calculate the current speed V i :

[0118]

[0119] Calculate all the speed values during the finger movement and record the maximum value V max ;

[0120] Calculate the short semi-axis b i :

[0121] b i = (V max -V i ) × Δt + e

[0122] wherein b i is the vertical distance of the user touch point position change at the current speed, V max is the maximum speed, V i is the current speed, and e is the empirical error;

[0123] The major axis a i is calculated.

[0124]

[0125] wherein a i is the maximum expansion distance of the elliptical window;

[0126] According to the major axis a i , the minor axis b i , and the focal distance c i , an elliptical dynamic fingerprint window is generated, with its center being (x i , y i ).

[0127] The user's dynamic fingerprint data is collected in real time through the earphone touch sensor, and an elliptical dynamic fingerprint window is generated according to the real-time dynamic fingerprint data, so as to predict the moving track of the finger, thereby realizing accurate tracking and recognition of the user's dynamic behavior, facilitating the user to monitor and dynamically adjust the touch action in real time. The elliptical dynamic fingerprint window can adapt to fingerprints of various angles, avoiding recognition errors caused by the angle of fingerprint contact, thereby enhancing the adaptability of the system to complex touch operations, improving the accuracy and real-time performance of fingerprint matching, reducing the occurrence of misoperation, and effectively preventing false fingerprint attacks, thereby improving the convenience and satisfaction of the user when using the earphone.

[0128] S3, converting the elliptical dynamic fingerprint window into a rectangular dynamic fingerprint window for fingerprint data matching;

[0129] Further, converting the elliptical dynamic fingerprint window into a rectangular dynamic fingerprint window includes,

[0130] The calculation of the range of the rectangular dynamic fingerprint window includes the number of left-moving grids, the number of right-moving grids, the number of upward-moving grids, and the number of downward-moving grids;

[0131] The number of left-moving grids:

[0132]

[0133] The number of right-moving grids:

[0134]

[0135] The number of upward-moving grids:

[0136]

[0137] Down grid number:

[0138]

[0139] Where, N ileft is the grid number of the left boundary, N iright is the grid number of the right boundary, N ibottom is the grid number of the upper boundary, N itop is the grid number of the lower boundary, and alpha is the rotation angle, and cellsize is the size of the grid.

[0140] According to the boundaries of the rectangular dynamic fingerprint window, the elliptical dynamic fingerprint window data is mapped into the rectangular dynamic fingerprint window to generate a rectangular dynamic fingerprint window containing the elliptical dynamic fingerprint window.

[0141] By converting the elliptical dynamic fingerprint window into a rectangular dynamic fingerprint window, the fingerprint matching process can be simplified, the speed and accuracy of fingerprint matching can be improved, and the delay in the identification process can be reduced. Through this conversion, the computational complexity is reduced, the processing efficiency is improved, the consumption of system resources, the pressure of data transmission and storage are reduced, fast and accurate fingerprint matching is realized, the waiting time of the user in the touch operation is reduced, and at the same time, by mapping the data to a rectangular window, the consistency and standardization of data processing are enhanced, the reliability and stability of the system are improved, and efficient anti-mis touch function is ensured, thereby improving the user experience and satisfaction.

[0142] Furthermore, the fingerprint data matching is performed;

[0143] The fingerprint feature point matching includes,

[0144] Collecting real-time dynamic fingerprint data, constructing a rectangular dynamic fingerprint window, recording all rectangular dynamic fingerprint windows changing with the finger, selecting a rectangular dynamic fingerprint window closest to the center point of the rectangular dynamic fingerprint window, recording the upper left corner coordinates (x min , y min ) and the lower right corner coordinates (x max , y max ), checking the extracted feature point coordinates, and retaining the feature points that satisfy the conditions x min <x<x max and y min <y<y max , and filtering the feature points that do not satisfy the conditions.

[0145] Extract the fingerprint feature points p in real-time dynamic fingerprint data and the fingerprint feature points q in the fingerprint database respectively, calculate the Euclidean distance between each feature point p and feature point q, select the feature point pair with the smallest Euclidean distance as the final matching feature point pair, and calculate the fingerprint matching degree S based on the Euclidean distance between the final matching feature point pair:

[0146]

[0147] Where M is the number of matching feature point pairs, dist(p f , q f ) is the Euclidean distance between the matching feature point pair (p f , q f ), p f is the real-time dynamic fingerprint data feature point in the fth matching feature point pair, and q f is the fingerprint feature template feature point in the fth matching feature point pair.

[0148] Set the fingerprint matching degree threshold St, and compare it with the fingerprint matching degree S:

[0149] If S≥St, it is determined that the user identity verification is successful, and the verification result is sent to the mobile phone application program through Bluetooth, and the mobile phone application program displays a verification success message and allows the user to continue operation;

[0150] If S<St, it is determined that the user identity verification fails, the earphone automatically triggers the anti-mis-touch mechanism, locks the touch operation, and restores the touch function after 5 seconds, and sends the verification information to the mobile phone application program through Bluetooth, and sends a warning to the user.

[0151] By matching the fingerprint data in the rectangular dynamic fingerprint window, an efficient fingerprint verification process is realized, the feature points are extracted from the real-time dynamic fingerprint data and limited in the rectangular window, and the matching degree calculation is performed with the fingerprint template in the database, which improves the accuracy of fingerprint recognition and reduces the possibility of false matching, and can distinguish the fingerprints of different users, ensuring high precision and efficiency of fingerprint recognition, preventing unauthorized access of users, ensuring user data security and privacy protection. In addition, fast response and efficient identity verification of user operation are realized, the consumption of computing resources is reduced, the running efficiency of the earphone system is improved, the service life of the earphone is prolonged, the maintenance cost is reduced, and the user experience of the earphone is enhanced.

[0152] S4, collect user earphone usage data, build a touch scene analysis model, and obtain an earphone usage scene;

[0153] Further, the user earphone usage data is collected, a touch scene analysis model is built, and an earphone usage scene is obtained, including,

[0154] Collect earphone historical use data of mobile phone application, generate earphone historical use data set and divide it into training set and test set, the earphone historical use data set includes feature parameters and label data, the feature parameters include average acceleration, angular velocity of earphone rotation, ambient light intensity, ambient noise level and heart rate variation, the label data indicates that each touch event corresponds to earphone use scenario including emergency scenario, sports scenario and static scenario;

[0155] Extract feature parameters in earphone historical use data set to form high-dimensional feature matrix Q:

[0156]

[0157] Wherein is the average acceleration, w is the angular velocity of earphone rotation, O is the ambient light intensity, U is the ambient noise level, and ΔHR is the heart rate variation, m is the number of samples;

[0158] Use convolutional neural network to build touch scenario analysis model, including input layer, convolutional layer, pooling layer, fully connected layer and output layer;

[0159] The input content of the input layer is high-dimensional feature matrix Q;

[0160] The convolutional layer extracts features through multiple convolutional layers, each convolutional layer is composed of multiple convolutional kernels and activation functions, and generates feature maps;

[0161] The pooling layer pools the feature maps output by the convolutional layer, reduces the dimension of the feature maps, and extracts main features;

[0162] The output of the fully connected layer is flattened, and the feature combination and classification are performed through the fully connected layer;

[0163] The output layer uses Softmax activation function for multi-classification, and outputs the probability distribution of earphone use scenario;

[0164] Use training set data to train the model, use cross-entropy loss function to calculate the model error, and use Adam optimizer to update the parameters;

[0165] Use test set to evaluate the performance of the model, and optimize the model through hyperparameter tuning and cross-validation;

[0166] Input the real-time collected earphone use data of the user into the trained model to obtain the earphone use scenario.

[0167] By collecting the use data of the user's earphone, a touch scene analysis model is constructed, the operation logic of the earphone can be dynamically adjusted according to different use scenes, historical use data of the earphone is collected, a convolutional neural network is used for model training and optimization, through this scene analysis model, the current scene of the user is intelligently identified, the intelligent identification and classification of the use scene of the user are realized, a personalized earphone control experience is provided, the mis-touch is effectively reduced, especially in the case of exercise or emergency, the scene perception ability is particularly important, thereby the intelligent level of the earphone is improved.

[0168] S5, dynamically adjusting the operation logic of the earphone according to the use scene of the earphone;

[0169] Further, dynamically adjusting the operation logic of the earphone according to the use scene of the earphone includes,

[0170] The use scene of the earphone includes cycling, driving, walking, running and quiet scene, the safety, practicability and demand of the user are taken as the principle of setting the scene priority.

[0171] The user is in the state of cycling and driving, which is an emergency scene, belonging to the highest priority, the user's attention is focused on the road and traffic conditions, the touch control is disabled and the voice control is enabled, the earphone ambient sound mode is turned on, the ambient sound mode can adjust the reception amount of ambient sound according to the user's demand, turning on the ambient sound mode will weaken the noise reduction function of the earphone and enhance the human voice, so as to carry out clear communication and voice communication;

[0172] The user is in the state of walking and running, which is a sports scene, belonging to the second highest priority, the user's body and hands are in the state of exercise, the exercise mode is turned on, the music is automatically played when the user starts to exercise, and the music is automatically paused when the user stops exercising;

[0173] The user is in the state of stillness, which is a quiet scene, belonging to the general priority, the user normally uses the earphone, all functions of the earphone are opened, and the user can use the complete earphone function.

[0174] By dynamically adjusting the operation logic according to different use scenes of the earphone, a more intelligent and personalized earphone control experience is provided, it is ensured that the user can get the best feedback in different scenes, when the user is cycling or driving, the earphone disables the touch control to avoid mis-touch, and enables the voice control and ambient sound mode to improve safety, in the sports scene, the earphone simplifies the touch control, automatically plays or pauses the music, enhances the convenience of use, realizes the dynamic adjustment of the operation logic of the earphone according to the use scene of the user, significantly reduces the occurrence of mis-touch, improves the safety and convenience of the earphone, and enhances the user satisfaction.

[0175] Embodiment 2, refer to Figure 3For the second embodiment of the application, which is different from the previous embodiment, a fingerprint recognition-based earphone anti-mis-touch control system is provided, which comprises,

[0176] A data acquisition and preprocessing module is configured to acquire user static fingerprint data and dynamic fingerprint data, and to preprocess the fingerprint data.

[0177] A window conversion module is configured to generate an elliptical dynamic fingerprint window and convert the elliptical dynamic fingerprint window into a rectangular dynamic fingerprint window according to the acquired user fingerprint data.

[0178] A fingerprint matching module is configured to perform fingerprint feature point matching between the real-time acquired dynamic fingerprint data and the fingerprint data in the fingerprint database.

[0179] A data analysis module is configured to construct a touch dynamic analysis model and analyze user earphone use scenarios.

[0180] A dynamic adjustment module is configured to dynamically adjust the operation logic of the earphone according to the earphone use scenarios.

[0181] If the functions are realized in the form of software function units and sold or used as independent products, they can be stored in a computer readable storage medium. Based on this understanding, the technical solutions of the application or the parts of the technical solutions that essentially contribute to the prior art can be embodied in the form of a software product stored in a storage medium, including a plurality of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the application. The aforementioned storage medium includes a U disk, a mobile hard disk, a read-only memory (ROM, Read-Only Memory), a random access memory (RAM, Random Access Memory), a magnetic disk or an optical disk, and various media that can store program codes.

[0182] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a list of executable instructions for implementing logical functions, which can be embodied in any computer readable medium for use by or in conjunction with an instruction execution system, device or apparatus, such as a computer-based system, a system including a processor or other system that can fetch and execute instructions from the instruction execution system, device or apparatus. For the purpose of this specification, the "computer readable medium" can be any device that can contain, store, communicate, propagate or transport programs for use by or in conjunction with the instruction execution system, device or apparatus, or in conjunction with these instruction execution systems, devices or apparatus.

[0183] More specific examples (a non-exhaustive list) of the computer-readable medium include the following: an electrical connection (electronic) having one or more wires, a portable computer diskette (magnetic), a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), an optical fiber device, and a portable compact disc read-only memory (CDROM). Additionally, the computer readable medium can be paper or another suitable medium upon which the program is printed, as the program can be electronically captured, for example, via an optical scanner, then compiled, interpreted, or otherwise processed, as necessary, and stored in a computer memory.

[0184] It should be understood that aspects of the application can be implemented in hardware, software, firmware or combinations thereof. In the embodiments described above, various steps or methods can be implemented, for example, in software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented in hardware, and in another embodiment, any of the following technologies, known in the art, or combinations thereof, can be used: discrete logic circuitry having logic gates for implementing logic functions upon an application of data signals, application specific integrated circuits having appropriate combinational logic gates, programmable gate arrays (PGA), field programmable gate arrays (FPGA), and the like.

[0185] It should be noted that the above examples are merely intended to illustrate the technical solutions of the present application and not to limit the same. Although the present application has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present application can be modified or replaced by equivalents without departing from the spirit and scope of the present application, and all such modifications or replacements should be included in the scope of the claims of the present application.

Claims

1. A method for preventing accidental touches in headphones based on fingerprint recognition, characterized in that: include, Collect user static and dynamic fingerprint data, preprocess them, and store them in a fingerprint database; An elliptical dynamic fingerprint window is generated based on real-time collected user dynamic fingerprint data. The elliptical dynamic fingerprint window is converted into a rectangular dynamic fingerprint window for fingerprint feature point matching. Collect user headphone usage data, build a touch scenario analysis model, and obtain headphone usage scenarios; Dynamically adjust the headphone's operation logic based on the headphone's usage scenario; The process of collecting and preprocessing user static and dynamic fingerprint data and storing it in the fingerprint database refers to the process of collecting the user's fingerprint image using the headset's accompanying application when the user first uses the headset, removing noise from the fingerprint image through a filter, enhancing the contrast of the fingerprint image using image enhancement technology, and converting the grayscale image into a binary image. The Zhang-Suen thinning algorithm is used to thin the ridges in the binary image to a single pixel width. Static feature points, including ridge endpoints and bifurcation points, are extracted from the thinned fingerprint image. Descriptive information, including position and orientation, is generated for each static feature point. The extracted static feature point information is integrated into a unified format, including the position, orientation and type of each static feature point, to generate a static fingerprint data feature vector. The static fingerprint data feature vector includes the position coordinates, orientation angle and number of static feature points. Capacitive touch sensors and pressure sensors are installed on the touch area of ​​the earphone surface; When a user uses headphones, dynamic fingerprint data is collected in real time. The dynamic fingerprint data includes time-series data generated during the user's operation, the current fingerprint image recorded at each time point, the finger position, pressure intensity, and touch type, including light touch, long press, and swipe. The fingerprint position coordinates and pressure values ​​in the dynamic fingerprint data are normalized to generate a dynamic fingerprint data feature vector, which includes the dynamic fingerprint position, pressure value and timestamp. The feature vectors of static fingerprint data and dynamic fingerprint data are normalized and then concatenated to form a multi-dimensional fused data vector. Fingerprint feature templates are generated based on the multi-dimensional fused data vector and stored in the database.

2. The method for preventing accidental touches in headphones based on fingerprint recognition as described in claim 1, characterized in that: The process of generating an elliptical dynamic fingerprint window based on real-time collected user dynamic fingerprint data includes: Calculate the dynamic fingerprint window parameters and set the current touch point position of the user's finger as ( The next touch point location is ( ); Collect historical touch data from the user's headphones, and collect data from N touch points; Calculate the distance difference between each touch: in, It is the user's historical touch data in the headphones, the first finger... The location of the next touch; Calculate the mean distance between all touch points. : Using the PDR algorithm, predict the next touch point after the user's finger moves. )Location: in, It is the direction angle of the user's finger movement; Calculate the straight-line distance between two touch points : Move distance using your finger and unit time interval t calculates the current velocity : Calculate all speed values ​​during finger movement and record the maximum value. ; Calculate the minor semi-axis : in, It is the vertical distance of the change in the user's touch point position at the current speed. That is the maximum speed. This is the current speed. It is an empirical error; Calculate the major semi-axis : in, It is the maximum expansion distance of the elliptical window; Based on the long half-axis short half shaft and focal length Generate an elliptical dynamic fingerprint window, the center of which is ( ).

3. The method for preventing accidental touches in headphones based on fingerprint recognition as described in claim 2, characterized in that: The step of converting the elliptical dynamic fingerprint window into a rectangular dynamic fingerprint window includes, The range of the rectangular dynamic fingerprint window is calculated by including the number of grid cells to the left, right, up, and down. Number of grid cells to shift left: Number of grid cells to shift to the right: Number of grid cells moved up: Number of grid cells to move down: in, It is the number of grid cells on the left boundary. It is the number of grid cells on the right boundary. It is the number of grid cells at the upper boundary. It is the number of grid cells at the lower boundary. It's the rotation angle. It is the size of the grid; Based on the boundaries of the rectangular dynamic fingerprint window, the data of the elliptical dynamic fingerprint window is mapped onto the rectangular dynamic fingerprint window to generate a rectangular dynamic fingerprint window that contains the elliptical dynamic fingerprint window.

4. The method for preventing accidental touches in headphones based on fingerprint recognition as described in claim 3, characterized in that: The fingerprint feature point matching process includes, Collect real-time dynamic fingerprint data, construct a rectangular dynamic fingerprint window, record all rectangular dynamic fingerprint windows that change with the finger, select the rectangular dynamic fingerprint window whose corner is closest to the center point, and record the coordinates of the top left corner. , ) and the coordinates of the lower right corner ( , ), check the coordinates of the extracted feature points, and simultaneously satisfy and Feature points that meet the conditions are retained, while feature points that do not meet the conditions are filtered out. Fingerprint feature points p and q from real-time dynamic fingerprint data and fingerprint database are extracted respectively. The Euclidean distance between each feature point p and feature point q is calculated. The feature point pair with the smallest Euclidean distance is selected as the final matching feature point pair. The fingerprint matching degree S is calculated based on the Euclidean distance between the final matching feature point pairs. Where M is the number of matching feature point pairs. It is a pair of matching feature points The Euclidean distance between them It is the first Real-time dynamic fingerprint data feature points in a pair of matching feature points It is the first Fingerprint feature template feature points in a pair of matching feature points; Set a fingerprint matching threshold St and compare it with the fingerprint matching score S: If S≥St, the user authentication is considered successful. The authentication result is sent to the mobile application via Bluetooth. The mobile application displays a successful authentication message and allows the user to continue. If S < St, the user authentication is deemed to have failed. The headset automatically triggers the anti-accidental touch mechanism, locks the touch operation, and restores the touch function after 5 seconds. The authentication information is then sent to the mobile application via Bluetooth to issue a warning to the user.

5. The method for preventing accidental touches in headphones based on fingerprint recognition as described in claim 4, characterized in that: The process involves collecting user headphone usage data and constructing a touch scenario analysis model to determine headphone usage scenarios, including... Collect historical headphone usage data from mobile applications, generate a headphone historical usage dataset and divide it into training and testing sets. The headphone historical usage dataset includes feature parameters and label data. The feature parameters include average acceleration, angular velocity of headphone rotation, ambient light intensity, ambient noise level and heart rate change. The label data refers to the headphone usage scenario corresponding to each touch event, including emergency scenario, sports scenario and stationary scenario. Extracting the headphone history uses feature parameters from the dataset to construct a high-dimensional feature matrix Q: in It is the average acceleration. It is the angular velocity of the headphone rotation. It is ambient light intensity. It refers to the ambient noise level. It represents heart rate changes, and m is the number of samples; A touch scenario analysis model is built using convolutional neural networks, including an input layer, convolutional layers, pooling layers, fully connected layers, and an output layer. The input layer receives a high-dimensional feature matrix Q. The convolutional layer extracts features through multiple convolutional layers. Each convolutional layer consists of multiple convolutional kernels and activation functions to generate feature maps. The pooling layer pools the feature map output by the convolutional layer, reducing the dimension of the feature map and extracting the main features; The fully connected layer flattens the output of the pooling layer and performs feature combination and classification through the fully connected layer. The output layer uses the Softmax activation function for multi-class classification and outputs the probability distribution of headphone usage scenarios; The model is trained using the training set data, the model error is calculated using the cross-entropy loss function, and the parameters are updated using the Adam optimizer. The model's performance was evaluated using a test set, and the model was optimized through hyperparameter tuning and cross-validation. The real-time user headphone usage data is input into the trained model to obtain the headphone usage scenario.

6. The method for preventing accidental touches in headphones based on fingerprint recognition as described in claim 5, characterized in that: The logic for dynamically adjusting the headphone's operation based on the headphone's usage scenario includes... The headphone usage scenarios include cycling, driving, walking, running, and quiet scenarios, and the headphone usage scenario priority is set; When the user is riding or driving, it is an emergency situation with the highest priority. The user's attention is focused on the road and traffic conditions. Touch control is disabled and voice control is enabled. The headphone ambient sound mode is turned on. When the user is walking or running, it is considered a sports scenario, which is of the second highest priority. When the user's body and hands are in motion, the sports mode is activated. Music is automatically played when the user starts exercising and automatically paused when the user stops exercising. When the user is stationary, it is a quiet scenario, which is of normal priority, and the user can use the headphones normally.

7. A fingerprint recognition-based headphone accidental touch prevention control system according to any one of claims 1-6, characterized in that: include, The data acquisition and preprocessing module is used to acquire static and dynamic fingerprint data from users and to preprocess the fingerprints. The window conversion module is used to generate an elliptical dynamic fingerprint window based on the collected user fingerprint data and convert the elliptical dynamic fingerprint window into a rectangular dynamic fingerprint window. The fingerprint matching module is used to match fingerprint feature points between the real-time acquired dynamic fingerprint data and the fingerprint data in the fingerprint database. The data analysis module is used to build a touch dynamic analysis model and analyze user headphone usage scenarios; The dynamic adjustment module is used to dynamically adjust the operating logic of the headphones based on the usage scenario.

8. A computer device, comprising: Memory and processor; The memory stores a computer program, characterized in that: when the processor executes the computer program, it implements the steps of the fingerprint recognition-based headphone anti-mistouch control method as described in any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by the processor, it implements the steps of the fingerprint recognition-based headphone anti-mistouch control method as described in any one of claims 1 to 6.

Images