CN116682171A

CN116682171A - Method for training a panic-tense gait recognition model, gait recognition method and related device

Info

Publication number: CN116682171A
Application number: CN202310547846.6A
Authority: CN
Inventors: 王铁鑫; 叶汉银; 黄鹏; 李剑
Original assignee: Shenzhen H&T Intelligent Control Co Ltd
Current assignee: Shenzhen H&T Intelligent Control Co Ltd
Priority date: 2023-05-15
Filing date: 2023-05-15
Publication date: 2023-09-01

Abstract

The embodiment of the application relates to the technical field of gait recognition, and discloses a method for training a panic tension gait recognition model, a gait recognition method and a related device. And dividing the gait samples according to the periodicity of the gait samples to obtain a plurality of gait cycle wave bands. And (3) performing iterative training on the two classification algorithm models by adopting a plurality of gait cycle bands to obtain a recognition model of the panic-tense gait. In this embodiment, the gait samples are divided according to their periodicity so that each resulting gait cycle band can reflect the motion data of a complete step, conforming to the walking cycle. Therefore, the training-obtained recognition model of the panic-tense gait can accurately recognize the panic-tense gait based on analysis and mining of potential characteristics of each gait cycle wave band.

Description

Method for training a panic-tense gait recognition model, gait recognition method and related device

Technical Field

The embodiment of the application relates to the technical field of gait recognition, in particular to a method for training a panic tension gait recognition model, a gait recognition method and a related device.

Background

Parkinsonism is a chronic neurological disorder disease which is frequently seen in the elderly, and in the early stages of the disease, gait abnormalities are mainly manifested by changes in stride and walking rhythm. The panic Gait (oblique Gait) is one of the most typical and unique Gait disorders of parkinson patients. The panic gait is characterized by slight flexion of hip joints and knee joints, forward movement of body weight, downward movement of neck, reduced swing of upper limbs, landing of toes or feet during stepping, small steps, small step size, slow starting and rapid falling. As the condition increases, a frozen gait develops gradually. Frozen gait is frequently seen in middle and late stages of the disease, and is mainly manifested by transient retardation of movement, difficulty in lifting foot and stepping.

The parkinsonism is hidden in early onset and is not easily perceived, and many patients have reached mid-term after diagnosis, so early detection is very necessary. However, at present, some sensors are mainly arranged on the sole of the foot, so that abnormal gait is detected, and the accuracy is low.

Disclosure of Invention

The technical problem to be solved by the embodiment of the application is to provide a method for training a panic-tense gait recognition model, a gait recognition method and a related device, wherein the panic-tense gait recognition model obtained through training can accurately recognize the panic-tense gait.

In a first aspect, an embodiment of the present application provides a method for training a recognition model of a panic-tense gait, including:

acquiring a plurality of gait samples, wherein the gait samples comprise acceleration signal waves and angular velocity signal waves acquired by an inertial measurement unit for acquiring walking steps of feet, and one gait sample is marked with a label which comprises a panic gait or a normal gait;

dividing the gait sample according to the periodicity of the gait sample to obtain a plurality of gait cycle wave bands;

and (3) performing iterative training on the two classification algorithm models by adopting a plurality of gait cycle wave bands obtained through division to obtain a panic-tense gait recognition model.

In some embodiments, the dividing the gait sample according to the periodicity of the gait sample to obtain a plurality of gait cycle bands includes:

acquiring a peak of the acceleration signal wave in the motion direction;

determining a start index and an end index of each gait cycle wave band according to each wave crest;

Intercepting gait samples according to the starting index and the ending index of each gait cycle wave band to obtain a plurality of gait cycle wave bands.

In some embodiments, determining the start index and the end index of each gait cycle band from each peak includes:

and sequentially sliding the middle position of the sliding window to each wave crest, and determining a start index and an end index of a gait cycle wave band according to the target wave band covered by the sliding window positioned at one wave crest.

In some embodiments, determining the start index and the end index of a gait cycle band according to the target band covered by the sliding window at a peak includes:

performing first-order differential processing on the target wave band to obtain a differential sequence;

according to the differential sequence, determining an index corresponding to a first stable wave band positioned at the left side of the wave crest of the target wave band, and determining a starting index of the gait cycle wave band according to the index corresponding to the first stable wave band;

and determining an index corresponding to a second stable wave band positioned on the right side of the wave crest of the target wave band according to the differential sequence, and determining an ending index of the gait cycle wave band according to the index corresponding to the second stable wave band.

In some embodiments, the determining, according to the differential sequence, an index corresponding to a first stationary band located on the left side of the peak in the target band includes:

screening indexes of which absolute values of values in the differential sequence positioned at the left side of the wave crest are smaller than or equal to a preset threshold value, and taking the indexes as a first candidate index set;

and taking the k continuous indexes in the first candidate index set as indexes corresponding to the first stable wave band.

In some embodiments, determining the start index of the gait cycle band according to the index corresponding to the first stationary band includes:

determining a first stable wave band from the target wave band according to the index corresponding to the first stable wave band; and obtaining the maximum value of the signal value in the first stable wave band, and taking the index corresponding to the maximum value as the starting index of the gait cycle wave band.

In some embodiments, the determining, according to the differential sequence, an index corresponding to a second stable band located on the right side of the peak in the target band, and determining, according to the index corresponding to the second stable band, an end index of the gait cycle band includes:

screening indexes with absolute values of values in the differential sequence positioned on the right side of the wave crest smaller than or equal to a preset threshold value as a second candidate index set;

Taking the continuous g indexes in the second candidate index set as indexes corresponding to the second stable wave band;

and determining a second stable wave band from the target wave band according to the index corresponding to the second stable wave band, acquiring the minimum value of the signal value in the second stable wave band, and taking the index corresponding to the minimum value as the ending index of the gait cycle wave band.

In some embodiments, before the step of performing iterative training on the binary algorithm model to obtain the recognition model of the panic-tense gait by using the divided gait cycle bands, the method further includes:

the length of each gait cycle band is adjusted to the target length.

In some embodiments, adjusting the length of each gait cycle band to the target length includes:

and (3) adopting linear interpolation to adjust the length of each gait cycle wave band to be the target length.

In some embodiments, before the step of dividing the gait sample according to the periodicity of the gait sample to obtain the plurality of gait cycle bands, the method further comprises:

and normalizing the gait samples.

In a second aspect, an embodiment of the present application provides a gait recognition method, including:

Acquiring a test gait sample, wherein the test gait sample comprises an acceleration signal wave and an angular velocity signal wave which are acquired by an inertial measurement unit for the walking steps of the foot;

dividing the test gait samples according to the periodicity of the test gait samples to obtain a plurality of test gait cycle wave bands;

and inputting a plurality of test gait cycle bands into a panic-tense gait recognition model, and outputting the gait category to which each step belongs, wherein the panic-tense gait recognition model is trained by the method of the first aspect.

In a third aspect, an embodiment of the present application provides an electronic device, including:

at least one processor; and

a memory communicatively coupled to the at least one processor; wherein the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the method of the first aspect.

In a fourth aspect, embodiments of the present application provide a computer-readable storage medium having stored thereon computer-executable instructions for causing a computer device to perform the method of the first aspect.

The embodiment of the application has the beneficial effects that: in contrast to the situation in the prior art, the method for training the recognition model of the panic-tense gait provided by the embodiment of the application firstly obtains a plurality of gait samples, wherein the gait samples comprise acceleration signal waves and angular velocity signal waves obtained by collecting walking steps of feet by the inertial measurement unit, and one gait sample is marked with a label, and the label comprises the panic-tense gait or the normal gait. And dividing the gait samples according to the periodicity of the gait samples to obtain a plurality of gait cycle wave bands. And (3) performing iterative training on the two classification algorithm models by adopting a plurality of gait cycle wave bands obtained through division to obtain a panic-tense gait recognition model. In this embodiment, the gait samples are divided according to their periodicity so that each resulting gait cycle band can reflect the motion data of a complete step, conforming to the walking cycle. Therefore, the two classification algorithm models are trained by the gait cycle bands, and the obtained recognition model of the panic-tension gait can accurately recognize the panic-tension gait based on analysis and mining of potential characteristics of each gait cycle band. In addition, the panic tension gait recognition model is deployed on electronic equipment such as a microcontroller, so that the force pressure can be effectively calculated, the data transmission delay is reduced, the real-time detection recognition is realized, the accuracy of an algorithm is improved, and the robustness and generalization capability of the model are enhanced.

Drawings

One or more embodiments are illustrated by way of example and not limitation in the figures of the accompanying drawings, in which like references indicate similar elements, and in which the figures of the drawings are not to be taken in a limiting sense, unless otherwise indicated.

FIG. 1 is a schematic diagram of a gait recognition system in accordance with some embodiments of the application;

FIG. 2 is a schematic diagram of an electronic device according to some embodiments of the application;

FIG. 3 is a flow chart of a method of training a recognition model of a panic tension gait in some embodiments of the application;

FIG. 4 is a schematic diagram of gait cycle bands after normalization in some embodiments of the present application;

FIG. 5 is a schematic diagram of acceleration signal waves in a motion direction according to some embodiments of the present application;

fig. 6 is a flow chart of a gait recognition method according to some embodiments of the application.

Detailed Description

The present application will be described in detail with reference to specific examples. The following examples will assist those skilled in the art in further understanding the present application, but are not intended to limit the application in any way. It should be noted that variations and modifications could be made by those skilled in the art without departing from the inventive concept. These are all within the scope of the present application.

The present application will be described in further detail with reference to the drawings and examples, in order to make the objects, technical solutions and advantages of the present application more apparent. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the application.

It should be noted that, if not in conflict, the features of the embodiments of the present application may be combined with each other, which is within the protection scope of the present application. In addition, while functional block division is performed in a device diagram and logical order is shown in a flowchart, in some cases, the steps shown or described may be performed in a different order than the block division in the device, or in the flowchart. Moreover, the words "first," "second," "third," and the like as used herein do not limit the data and order of execution, but merely distinguish between identical or similar items that have substantially the same function and effect.

Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used in the description of the application herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the application. The term "and/or" as used in this specification includes any and all combinations of one or more of the associated listed items.

In addition, the technical features of the embodiments of the present application described below may be combined with each other as long as they do not collide with each other.

In order to facilitate understanding of the method provided in the embodiments of the present application, first, terms related to the embodiments of the present application are described:

(1) Neural network

A neural network may be composed of neural units, and is understood to mean, in particular, a neural network having an input layer, an hidden layer, and an output layer, where in general, the first layer is the input layer, the last layer is the output layer, and the middle layers are all hidden layers. Among them, the neural network with many hidden layers is called deep neural network (deep neural network, DNN). The operation of each layer in the neural network can be described by the mathematical expression y=a (w·x+b), from the physical level, and can be understood as the completion of the transformation of the input space into the output space (i.e., the row space into the column space of the matrix) by five operations on the input space (set of input vectors), including 1, dimension up/down; 2. zoom in/out; 3. rotating; 4. translating; 5. "bending". Wherein the operations of 2, 3 are done by "w·x", the operations of 4 are done by "+b", and the operations of 5 are done by "a ()" where the expression "space" is used in two words because the object to be classified is not a single thing but a class of things, space refers to the collection of all individuals of such things, where W is the weight matrix of the layers of the neural network, each value in the matrix representing the weight value of one neuron of that layer. The matrix W determines the spatial transformation of the input space into the output space described above, i.e. the W of each layer of the neural network controls how the space is transformed. The purpose of training the neural network is to finally obtain the weight matrix of all layers of the trained neural network. Thus, the training process of the neural network is essentially a way to learn and control the spatial transformation, and more specifically to learn the weight matrix.

It should be noted that in the embodiments of the present application, the neural network is essentially based on the model employed by the machine learning task. Common components in the neural network comprise a convolution layer, a pooling layer, a normalization layer and the like, a model is designed by assembling the common components in the neural network, and the model converges when model parameters (weight matrixes of all layers) are determined so that model errors meet preset conditions or the number of adjusted model parameters reaches a preset threshold.

Before describing the embodiments of the present application, a gait recognition method known to the present inventors will be briefly described, so that the embodiments of the present application will be conveniently understood later.

In some schemes, gait of a patient is detected by a gait detection sensor, then a time domain analysis module analyzes the detection signal to obtain a time domain index, and a frequency domain analysis module analyzes the detection signal to obtain a frequency domain index. The frequency domain calculating module judges the gait of the patient according to the obtained frequency domain index, and outputs a frequency domain judging result when the patient is judged to be abnormal gait. The time domain calculation module judges the gait of the patient according to the time domain index acquired by the time domain analysis module and the preset parameter and outputs the abnormal gait detection result in real time, so that the detection result is ensured to have higher real-time performance. The correction module generates a correction factor for correcting the preset parameter of the time domain calculation module according to the calculation results of the time domain calculation module and the frequency domain calculation module, so that the time domain calculation module corrects the preset parameter according to the correction factor, the accuracy of the output of the time domain calculation module is improved, and the detection requirement can be met.

In this scheme, by manually setting the threshold, the accuracy is not high. And when the feature space has a high dimension, the threshold adjustment is very cumbersome, making gait detection inaccurate.

In view of the above problems, an embodiment of the present application provides a method for training a recognition model of a panic-tense gait, in which a plurality of gait samples are first acquired, the gait samples include a triaxial acceleration signal wave and a triaxial angular velocity signal wave obtained by acquiring a walking pace of a foot by an inertial measurement unit, and a label is labeled on one gait sample, and the label includes a panic-tense gait or a normal gait. And dividing the gait samples according to the periodicity of the gait samples to obtain a plurality of gait cycle wave bands. And (3) performing iterative training on the two classification algorithm models by adopting a plurality of gait cycle wave bands obtained through division to obtain a panic-tense gait recognition model. In this embodiment, the gait samples are divided according to their periodicity so that each resulting gait cycle band can reflect the motion data of a complete step, conforming to the walking cycle. Therefore, the two classification algorithm models are trained by the gait cycle bands, and the obtained recognition model of the panic-tension gait can accurately recognize the panic-tension gait based on analysis and mining of potential characteristics of each gait cycle band. In addition, the panic tension gait recognition model is deployed on electronic equipment such as a microcontroller, so that the force pressure can be effectively calculated, the data transmission delay is reduced, the real-time detection recognition is realized, the accuracy of an algorithm is improved, and the robustness and generalization capability of the model are enhanced.

Exemplary applications of the electronic device for training a panic-tense gait recognition model or for gait recognition provided by embodiments of the present application are described below. The electronic device provided by the embodiment of the application can be a server, for example, a server deployed at a cloud end. The electronic device provided by some embodiments of the present application may be a notebook computer, a desktop computer, or a mobile device.

As an example, referring to fig. 1, fig. 1 is a schematic view of an application scenario of a gait recognition system according to an embodiment of the application. The terminal 10 is connected to the server 20 via a network, which may be a wide area network or a local area network, or a combination of both.

The terminal 10 may be used to acquire training data and construct a classification model, for example, by those skilled in the art downloading prepared training data on the terminal and constructing a network structure of the classification model. Wherein the training data includes a plurality of gait samples reflecting foot walking steps. It will be appreciated that the terminal 10 may also be used to obtain test samples, for example, the inertial measurement unit may send the collected test data to the terminal 10, whereby the terminal 10 obtains the test data. In some embodiments, when the terminal 10 is used to detect gait, the terminal 10 may be integrated with an inertial measurement unit. In some embodiments, the terminal 10 is configured as a foot worn device that incorporates an inertial measurement unit, such that the inertial measurement unit is worn on the foot. After the inertial measurement unit collects the test data, the test data is sent to the microcontroller in the terminal 10, so that the microcontroller detects the test data by adopting a built-in panic-tense gait recognition model and obtains a detection result. In this embodiment, the whole device is simple in structure, lightweight, easy to wear, and can be used outdoors or indoors. In some embodiments, the device is also in communication connection with a mobile terminal (e.g., a smart phone or tablet, etc.), and the device sends the result of the detection of the panic gait to the mobile terminal in real time, facilitating daily monitoring of gait anomalies.

In some embodiments, the terminal 10 locally performs the method for training the recognition model of the panic-tense gait according to the embodiment of the present application to complete training the designed classification model by using training data, and determines the final model parameters, so that the classification model configures the final model parameters, and the recognition model of the panic-tense gait can be obtained. In some embodiments, the terminal 10 may also send the training data and the constructed classification model stored on the terminal by the person skilled in the art to the server 20 through the network, the server 20 receives the training data and the classification model, trains the classification model using the training data, determines the final model parameters, and then sends the final model parameters to the terminal 10, and the terminal 10 saves the final model parameters, so that the classification model configures the final model parameters, and the recognition model of the panic gait can be obtained.

Next, the structure of the electronic device according to the embodiment of the present application is described, and fig. 2 is a schematic structural diagram of the electronic device 500 according to the embodiment of the present application, where the electronic device 500 includes at least one processor 510 and a memory 550. The various components in electronic device 500 are coupled together by bus system 540. It is appreciated that the bus system 540 is used to enable connected communications between these components. The bus system 540 includes a power bus, a control bus, and a status signal bus in addition to the data bus. The various buses are labeled as bus system 540 in fig. 2 for clarity of illustration.

The processor 510 may be an integrated circuit chip with signal processing capabilities such as a general purpose processor, such as a microprocessor or any conventional processor, or the like, a digital signal processor (DSP, digital Signal Processor), or other programmable logic device, discrete gate or transistor logic, discrete hardware components, or the like.

Memory 550 includes volatile memory or nonvolatile memory, and may also include both volatile and nonvolatile memory. The nonvolatile Memory may be a Read Only Memory (ROM), and the volatile Memory may be a random access Memory (RAM, random Access M emory). The memory 550 described in embodiments of the present application is intended to comprise any suitable type of memory. Memory 550 may optionally include one or more storage devices physically located remote from processor 510.

In some embodiments, memory 550 is capable of storing data to support various operations, examples of which include programs, modules and data structures, or subsets or supersets thereof, as exemplified below. An operating system 551 including system programs for handling various basic system services and performing hardware-related tasks, such as a framework layer, a core library layer, a driver layer, etc., for implementing various basic services and handling hardware-based tasks; network communication module 552 is used to reach other computing devices via one or more (wired or wireless) network interfaces 520, exemplary network interfaces 520 include bluetooth, wireless compatibility authentication (WiFi), and universal serial bus (USB, universal Serial Bus), among others.

In some embodiments, the electronic device further comprises an inertial measurement unit. The electronic device may be worn or disposed on the instep of the user. And after the inertial measurement unit collects the test data, the test data are sent to the processor, so that the processor calls the recognition model of the panic tension gait stored in the memory to detect the test data, and the detection result is obtained. In this embodiment, the electronic device is simple in structure, lightweight, easy to wear by a human body, and can be used outdoors or indoors. In some embodiments, the electronic device is also in communication connection with a mobile terminal (e.g., a smart phone or tablet, etc.), and the electronic device sends the result of the detection of the panic gait to the mobile terminal in real time, so as to facilitate daily monitoring of gait anomalies.

It will be appreciated from the foregoing that the method for training the panic tension gait recognition model provided by the embodiments of the present application may be implemented by various types of electronic devices having processing capabilities, such as by a processor of the electronic device or by other devices having computing processing capabilities. Other devices with computing processing capabilities may be smart terminals or servers or the like communicatively coupled to the electronic device.

The method for training the panic tension gait recognition model provided by the embodiment of the application is described below in connection with exemplary applications and implementations of the electronic device provided by the embodiment of the application. Referring to fig. 3, fig. 3 is a flowchart illustrating a method for training a recognition model of a panic swing gait according to an embodiment of the application. It will be appreciated that the subject of execution of the training method may be one or more processors of the electronic device.

Referring again to fig. 3, the method S100 may specifically include the following steps:

s10: a plurality of gait samples are obtained.

The gait sample comprises a triaxial acceleration signal wave and a triaxial angular velocity signal wave which are acquired by an inertial measurement unit on the walking steps of the foot, and the gait sample is marked with a label which comprises a panic gait or a normal gait.

In some embodiments, the inertial measurement unit includes a tri-axis accelerometer and tri-axis gyroscope that can acquire acceleration signals and angular velocity signals of the moving object in the x-axis, y-axis, and z-axis, respectively. In some embodiments, one inertial measurement unit is disposed on a left foot of a human body and one inertial measurement unit is disposed on a right foot of the human body, such that acceleration signals and angular velocity signals of the left foot in the x-axis, the y-axis, and the z-axis, respectively, and acceleration signals and angular velocity signals of the right foot in the x-axis, the y-axis, and the z-axis, respectively, are acquired.

It will be appreciated that an inertial sensor may be integrated with a microcontroller to form a monitoring device disposed on an upper. After wearing the monitoring equipment, a plurality of test wearers walk according to a specified path. In some embodiments, with reference to the parkinsonism standard test, the test wearer walks straight ahead 5 meters and then turns straight back 5 meters to the origin.

After the monitoring device collects a plurality of gait samples, the gait samples are sent to the electronic device. It will be appreciated that in some embodiments, the gait categories of these test persons are not exactly the same, some are in a panic gait and some are in a normal gait, i.e. a non-parkinsonian gait. Thus, the electronic device may obtain a plurality of gait samples. Each gait sample corresponds to a gait class. I.e., each gait sample is labeled with a label that includes a panic gait or a normal gait.

The gait sample is based on a triaxial acceleration signal and a triaxial angular velocity signal obtained by acquiring a walking pace of a foot by an inertial measurement unit in a period of time, so that one gait sample has a signal wave with 6 dimensions, for example, including: left foot x-axis acceleration signal wave, left foot y-axis acceleration signal wave, left foot z-axis acceleration signal wave, left foot x-axis angular velocity signal wave, left foot y-axis angular velocity signal wave, left foot z-axis angular velocity signal wave. Further examples include: the device comprises a right foot x-axis acceleration signal wave, a right foot y-axis acceleration signal wave, a right foot z-axis acceleration signal wave, a right foot x-axis angular velocity signal wave, a right foot y-axis angular velocity signal wave and a right foot z-axis angular velocity signal wave.

In some embodiments, if the y-axis coincides with the walking direction, the z-axis is the vertical ground direction, and the x-axis is the in-foot direction, the y-axis acceleration signal wave reflects the acceleration of foot walking in the walking direction. Based on the walking characteristics, the in-foot direction has no acceleration or no reference to acceleration clutter, in which embodiment the x-axis acceleration signal wave may be discarded. Thus, one gait sample comprises 5-dimensional signal waves.

In some embodiments, a gait sample with at least 100 labels as a panic gait is selected from the acquired gait samples as a positive sample training set and a gait sample with at least 100 labels as a normal gait is selected as a negative sample training set. And selecting gait samples with at least 20 labels of the gait with the panic tension gait as a positive sample test set, and selecting gait samples with at least 20 labels of the gait with the normal gait as a negative sample test set.

In some embodiments, before step S20, further comprising: and normalizing the gait samples.

For example, if the gait sample includes 5-dimensional signal waves of a y-axis acceleration signal wave, a z-axis acceleration signal wave, and three-axis angular velocity signal waves, the y-axis is the walking direction, i.e., the movement direction, and the z-axis is the vertical ground direction. In this embodiment, signal waves of 5 dimensions of the gait sample are normalized respectively.

In some embodiments, the gait samples are normalized using the following formula, scaling the signal values of the individual signal waves to the [0,1] interval.

Wherein X is any signal value before normalization processing in a signal wave band, and X _norm To normalize the processed signal value, X _max For the signal value with the largest value in the signal wave band, X _min Is the signal value with the smallest value in the signal wave band.

Referring to fig. 4, fig. 4 is a schematic diagram of a normalized signal wave. The signal value of the signal wave is scaled to the [0,1] interval, so that the influence on the model caused by overlarge data difference can be reduced, and the accuracy of the model is improved.

In this embodiment, the normalization processing is performed on the signal wave of each dimension of each gait sample, so that adverse effects of the singular sample data on subsequent training can be effectively reduced, and model convergence is accelerated.

S20: and dividing the gait samples according to the periodicity of the gait samples to obtain a plurality of gait cycle wave bands.

Each gait sample is divided according to the periodicity of the gait samples, so that each obtained gait cycle wave band can reflect the motion data of the complete steps, accords with the walking cycle, and is beneficial to the subsequent classification model to learn each step characteristic so as to improve the classification accuracy.

In some embodiments, the foregoing step S20 specifically includes:

s21: and acquiring the wave peak of the acceleration signal wave in the motion direction.

S22: and determining a starting index and an ending index of each gait cycle wave band according to each wave peak.

S23: intercepting gait samples according to the starting index and the ending index of each gait cycle wave band to obtain a plurality of gait cycle wave bands.

In some embodiments, the direction of motion may be the inertial measurement unit y-axis direction, i.e., the acquisition y-axis signal wave. Due to the periodicity of walking, the acceleration signal wave also has periodicity in the direction of movement, and one period is the acceleration signal corresponding to one step. In the acceleration signal wave in the movement direction, there is a peak in the acceleration signal fluctuation in each period along with the ground separation and landing characteristics of the footstep. Typically the peak is at the highest point of the footstep during the step off to landing.

In this embodiment, in order to accurately divide the gait cycle band, each peak value of the acceleration wave in the movement direction is first acquired. Based on each peak falling in one gait cycle band, the start index and the end index of each gait cycle band can be determined from each peak. It will be appreciated that the signal waves of each dimension in the gait sample have a time sequence, the index of which may be a time stamp. Here, the start index and the end index of the gait cycle band are determined from the time stamp corresponding to each cycle in the acceleration signal wave in the movement direction. Finally, intercepting the gait samples according to the starting index and the ending index of each gait cycle wave band to obtain a plurality of gait cycle wave bands.

In this embodiment, in the acceleration signal wave based on the movement direction, along with the ground-leaving and landing characteristics of the footstep, each peak value is located at the highest point of the footstep, which is separated from the ground in the ground-leaving to landing process, and each peak value falls in one gait cycle band, so that the start index and the end index of each gait cycle band are determined according to each peak value, and the divided gait cycle bands are more accurate.

In some embodiments, the step S22 specifically includes:

s221: and sequentially sliding the middle position of the sliding window to each wave crest, and determining a start index and an end index of a gait cycle wave band according to the target wave band covered by the sliding window positioned at one wave crest.

Wherein the sliding window is a virtual data partitioning window. It will be appreciated that in some embodiments, the sliding window may be 2s in length. In some embodiments, a sliding window of 50hz and 2s acquisition frequency of the inertial measurement unit may cover 100 data.

After each wave crest is found, the sliding window slides on the acceleration signal wave in the moving direction, and when the middle position of the sliding window slides to one wave crest, the starting index and the ending index of a gait cycle wave band are determined according to the target wave band covered by the sliding window positioned at one wave crest. It will be appreciated that the target band is larger than the gait cycle band, so that the start index and the end index can be determined by analysing the trend of the signal values in the target band.

In this embodiment, the start index and the end index of a gait cycle band are determined by analyzing the target bands on the left and right sides of the peak, so that the start index and the end index are more accurate, thereby facilitating the accuracy of the gait cycle band.

In some embodiments, the foregoing "determining the start index and the end index of a gait cycle band according to the target band covered by the sliding window at a peak" includes:

(1) And performing first-order differential processing on the target wave band to obtain a differential sequence.

(2) And determining an index corresponding to a first stable wave band positioned at the left side of the wave crest of the target wave band according to the differential sequence, and determining a starting index of the gait cycle wave band according to the index corresponding to the first stable wave band.

(3) And determining an index corresponding to a second stable wave band positioned on the right side of the wave crest of the target wave band according to the differential sequence, and determining an ending index of the gait cycle wave band according to the index corresponding to the second stable wave band.

In this embodiment, the first-order differential processing refers to subtracting the previous signal value from the next signal value in the target band at the time stamp. And performing first-order differential processing on the target wave band to obtain a differential sequence. Thus, the differential sequence can reflect the trend of the change in the signal value in the target band. For example, if the signal value does not fluctuate much, the corresponding difference is relatively small, and if the signal value fluctuates much, the corresponding difference is relatively large.

The first stationary band is a band in which the signal value located on the left side of the peak in the target band does not fluctuate much. In the first stationary band, the acceleration signal value fluctuates less, corresponding to the stay phase after the footfall. Thus, according to the differential sequence, the index corresponding to the first stationary band located on the left side of the peak in the target band is determined.

In some embodiments, the foregoing "determining, from the differential sequence, the index corresponding to the first stationary band located to the left of the peak in the target band" includes: screening indexes of which absolute values of values in the differential sequence positioned at the left side of the wave crest are smaller than or equal to a preset threshold value, and taking the indexes as a first candidate index set; and taking the k continuous indexes in the first candidate index set as indexes corresponding to the first stable wave band.

It will be appreciated that the portion of the differential sequence to the left of the peak is determined from the index of the peak. In the part, if the absolute value of the difference value is smaller than or equal to a preset threshold value, the corresponding signal value is indicated to be not greatly fluctuated and is stable. The preset threshold is an empirical value preset, for example, may be 100. And forming a first candidate index set by indexes of the differences with absolute values smaller than a preset threshold value. And if the first candidate index set has k continuous indexes, taking the k indexes as indexes corresponding to the first stable wave band. Where k is a preset empirical value, for example, k may be 5.

It can be understood that if k indexes in the first candidate index set are continuous, the acceleration signal wave bands corresponding to the k indexes are smooth, the fluctuation is small, and the stationary phase when the foot falls to the ground is corresponding. Therefore, k consecutive indexes in the first candidate index set are used as indexes corresponding to the first stable wave band, so that the first stable wave band is more accurate. Furthermore, by setting k, the possibility of occurrence of consecutive sequence numbers in the motion state can be effectively shielded.

After the index corresponding to the first stationary band is obtained, a start index of the gait cycle band may be determined according to the index. For example, one of indexes corresponding to the first stationary band is used as a start index of the gait cycle band.

In some embodiments, the foregoing "determining the start index of the gait cycle band according to the index corresponding to the first plateau band" includes: determining a first stable wave band from the target wave band according to the index corresponding to the first stable wave band; and obtaining the maximum value of the signal value in the first stable wave band, and taking the index corresponding to the maximum value as the starting index of the gait cycle wave band.

In this embodiment, the signal value included in the first stationary band is found and determined from the target band based on the index corresponding to the first stationary band. And taking the index corresponding to the maximum signal value in the first stable wave band as the starting index of the gait cycle wave band. As shown in fig. 5, the index corresponding to the point B on the left side of the peak a is the start index.

It is understood that the maximum signal value is the last signal value when the footstep falls to the stationary phase, and can be understood as the signal value corresponding to the moment when the footstep starts to start to leave the ground. Starting from this point in time, which is the last signal value in the resting phase, the next signal value will fluctuate considerably as the foot starts to move. Therefore, the index corresponding to the maximum signal value in the first stable wave band is used as the starting index of the gait cycle wave band, accords with the step characteristics, and enables the starting index to be more accurate.

In some embodiments, the determining the index corresponding to the second stable band located on the right side of the peak in the target band according to the differential sequence, and determining the end index of the gait cycle band according to the index corresponding to the second stable band, includes: screening indexes with absolute values of values in the differential sequence positioned on the right side of the wave crest smaller than or equal to a preset threshold value as a second candidate index set; taking the continuous g indexes in the second candidate index set as indexes corresponding to the second stable wave band; and determining a second stable wave band from the target wave band according to the index corresponding to the second stable wave band, acquiring the minimum value of the signal value in the second stable wave band, and taking the index corresponding to the minimum value as the ending index of the gait cycle wave band.

It will be appreciated that the portion of the differential sequence to the right of the peak is determined from the index of the peak. In the part, if the absolute value of the difference value is smaller than or equal to a preset threshold value, the corresponding signal value is indicated to be not greatly fluctuated and is stable. The preset threshold is an empirical value preset, for example, may be 100. And forming a second candidate index set by indexes of the differences with absolute values smaller than a preset threshold value. And if the second candidate index set has continuous g indexes, taking the g indexes as indexes corresponding to the second stable wave band. Where g is a preset empirical value, for example, g may be 4.

It can be understood that if g indexes in the second candidate index set are continuous, the acceleration signal wave bands corresponding to the g indexes are smooth, the fluctuation is small, and the stationary phase when the foot falls to the ground is corresponding. Therefore, the continuous g indexes in the second candidate index set are used as indexes corresponding to the second stable wave band, so that the second stable wave band is more accurate. Furthermore, by setting g, the possibility of occurrence of consecutive sequence numbers in the motion state can be effectively shielded.

And after the index corresponding to the second stable wave band is acquired, determining the ending index of the gait cycle wave band according to the index. For example, one of indexes corresponding to the second stable band is set as an end index of the gait cycle band.

And finding and determining the signal value included in the second stable wave band from the target wave band based on the index corresponding to the second stable wave band. And taking the index corresponding to the minimum signal value in the second stable wave band as the ending index of the gait cycle wave band. As shown in fig. 5, the index corresponding to the point C on the right side of the peak a is the end index.

It will be appreciated that the minimum signal value is the signal value corresponding to the time the footstep falls to ground, after which the footstep enters the rest phase. Therefore, the index corresponding to the minimum signal value in the second stable wave band is used as the ending index of the gait cycle wave band, accords with the step characteristics, and ensures that the ending index is more accurate.

By the method, accurate start indexes and accurate end indexes can be obtained, so that gait samples can be intercepted according to the start indexes and the end indexes of each gait cycle wave band, and the accurate gait cycle wave band can be obtained.

In some embodiments, before step S30, further comprising:

s40: the length of each gait cycle band is adjusted to the target length.

It will be appreciated that the gait cycle bands are determined based on the characteristics of the foot step characteristics reflected in the direction of motion acceleration signal waves, and that the lengths of these gait cycle bands may be different. In this embodiment, a target length is preset, and it is understood that the target length is an empirical value set by those skilled in the art according to the actual step size. Specifically, the signal waves of 5 dimensions in the gait cycle wave band are all adjusted to the target length.

The lengths of the gait cycle wave bands are unified, and the lengths are uniformly adjusted to be target lengths. The method is beneficial to reducing the interference caused by the length of a pair of subsequent classification models.

In some embodiments, the step S40 specifically includes: and (3) adopting linear interpolation to adjust the length of each gait cycle wave band to be the target length.

The linear interpolation is to insert new signal values with zero error into each interpolation node based on the original signal values by adopting interpolation functions, such as a first order polynomial. It will be appreciated that linear interpolation is of the prior art and the interpolation process is not described in detail here.

For any one dimension of the gait cycle wave bands, for example, the acceleration signal wave band in the movement direction, linear interpolation is adopted to adjust the length of the signal wave band to the target length.

It can be understood that the linear interpolation can approximately replace the original signal wave under a certain allowable error, and the original waveform characteristics can be effectively maintained. For example, the single step time of a person walking is extremely short, the length of the corresponding signal wave band is also short, and the linear interpolation processing is adopted to fill in the short signal wave which is the signal wave with the target length of the self-set standard on the basis of basically keeping the original waveform characteristics.

In this embodiment, by unifying the lengths of the gait cycle bands to be the target lengths, on one hand, the method is beneficial to reducing the interference caused by a pair of subsequent classification models with different lengths, and quickening the model convergence, and on the other hand, the method can effectively maintain the original waveform characteristics and does not bring new interference.

S30: and (3) performing iterative training on the two classification algorithm models by adopting a plurality of gait cycle wave bands obtained through division to obtain a panic-tense gait recognition model.

It will be appreciated that in some embodiments, the number of gait cycle bands are gait cycle bands that are of uniform length. The labels are labeled based on each gait sample, and thus each gait cycle band is also labeled, including a panic gait or a normal gait. I.e. a part of the gait cycle band belongs to the panic-tense gait, the labeled label is 1,1 represents the panic-tense gait. The other part of gait cycle wave band belongs to normal gait, and the labeled label is 0, and 0 represents normal gait.

Here, a plurality of gait cycle bands are used as a training set of the classification algorithm model, the preset classification algorithm model is trained, parameters of the classification algorithm model are continuously adjusted, and under the constraint of a loss function, the predicted gait class output by the classification algorithm model is more and more close to the real gait class (label). When the loss calculated by the loss function fluctuates in a certain range or reaches a certain value, the two classification algorithm models converge, and parameters during convergence are taken as model parameters to obtain the gait recognition model.

In some embodiments, the classification algorithm model may be a logistic regression model. I.e., invoking a logistic regression model to train the training set. The logistic regression model is a conditional probability distribution as follows;

wherein x is a feature vector obtained by feature mapping extraction of a gait cycle wave band, w is a weight vector, and b is a bias. It will be appreciated that w and b are model parameters of the logistic regression model.

The logistic regression model predicts the labels using the following activation functions;

the basic linear regression form is y=w ^T x+b；

Replacing x with y to obtain a logistic regression model:

wherein x is a feature vector obtained by feature mapping extraction of a gait cycle wave band, it can be understood that if x=0, y=0.5; if x is less than 0, y is less than 0.5, which indicates that the feature vector x is judged as one type; if x > 0, y > 0.5, indicating that the feature vector x is discriminated as another class. Thus, the predictive label can be judged and obtained through the formula.

And calculating the loss between each prediction tag and the real tag by adopting a loss function, and adjusting model parameters of the logistic regression model based on the loss, wherein if the loss and the fluctuation in a certain range or a certain value are reached, the logistic regression model converges. The logistic regression model is configured with model parameters in convergence to obtain the recognition model of the panic gait.

In some embodiments, a test set comprising positive and negative samples is used to evaluate the accuracy and similarity of a converged logistic regression model (a recognition model of a panic gait). In some embodiments, the test set false positive rate of the negative sample reaches 96.77%. The test set accuracy of the positive samples reached 100%. Therefore, the recognition model of the panic tension gait obtained by training in the mode has higher accuracy.

To sum up, in some embodiments of the present application, a plurality of gait samples are first acquired, where the gait samples include a triaxial acceleration signal wave and a triaxial angular velocity signal wave acquired by an inertial measurement unit for a walking pace of a foot, and a gait sample is labeled with a tag, and the tag includes a panic gait or a normal gait. And dividing the gait samples according to the periodicity of the gait samples to obtain a plurality of gait cycle wave bands. And (3) performing iterative training on the two classification algorithm models by adopting a plurality of gait cycle wave bands obtained through division to obtain a panic-tense gait recognition model. In this embodiment, the gait samples are divided according to their periodicity so that each resulting gait cycle band can reflect the motion data of a complete step, conforming to the walking cycle. Therefore, the two classification algorithm models are trained by the gait cycle bands, and the obtained recognition model of the panic-tension gait can accurately recognize the panic-tension gait based on analysis and mining of potential characteristics of each gait cycle band. In addition, the panic tension gait recognition model is deployed on electronic equipment such as a microcontroller, so that the force pressure can be effectively calculated, the data transmission delay is reduced, the real-time detection recognition is realized, the accuracy of an algorithm is improved, and the robustness and generalization capability of the model are enhanced.

After the method for training the recognition model of the panic-tense gait is used for training to obtain the recognition model of the panic-tense gait, the recognition model of the panic-tense gait can be used for recognizing the panic-tense gait. The gait recognition method provided by the embodiment of the application can be implemented by various types of electronic equipment with calculation processing capacity, such as an intelligent terminal, a server or monitoring equipment with an inertia measurement unit.

The gait recognition method provided by the embodiment of the application is described below in connection with exemplary applications and implementations of the monitoring device provided by the embodiment of the application. Specifically, the monitoring device comprises an inertial measurement unit and a microcontroller, and the gait recognition method is deployed in a program form in the microcontroller. The monitoring device is small and wearable on a foot, such as an upper. And the microcontroller receives the test gait sample sent by the inertial measurement unit, analyzes potential characteristics and obtains the gait class.

Referring to fig. 6, fig. 6 is a flow chart of a gait recognition method according to an embodiment of the application. The method S200 comprises the steps of:

s201: and acquiring a test gait sample, wherein the test gait sample comprises a triaxial acceleration signal wave and a triaxial angular velocity signal wave which are acquired by an inertial measurement unit for the walking steps of the foot.

It can be understood that the test gait sample is a signal wave of triaxial acceleration and a signal wave of triaxial angular velocity acquired by the inertial measurement unit for walking steps of the foot, such as left foot and/or right foot, in an actual application scene. The inertial measurement unit sends the test gait sample to the microprocessor, whereby the microprocessor obtains the test gait sample.

S202: and dividing the test gait samples according to the periodicity of the test gait samples to obtain test gait cycle wave bands.

Referring to the method of dividing the gait sample in step S20, the test gait sample is divided into a plurality of test gait cycle bands. The specific dividing manner is referred to the detailed description about step S20, and will not be described in detail herein.

S203: and inputting a plurality of test gait cycle bands into the panic-tense gait recognition model, and outputting the gait category of each step.

The recognition model of the panic-tense gait is trained by the method for training the recognition model of the panic-tense gait in any embodiment.

It can be understood that the microprocessor is internally provided with a gait recognition application program, the panic tension gait recognition model is packaged in the gait recognition application program, the panic tension gait recognition model is called to perform gait recognition on the test gait cycle wave bands, and after a plurality of columns of calculation processing, the gait categories corresponding to the test gait cycle wave bands are output. The gait recognition model is obtained by training the method for training the panic tension gait recognition model in any one of the above embodiments, and has the same structure and function as the panic tension gait recognition model in the above embodiment, and will not be described in detail herein.

In this embodiment, the characteristic of movement of one step of walking is reflected based on the test gait cycle band, so that the recognition model of the panic-tense gait can recognize the gait class of each step of walking of the user, and in some embodiments, the recognition result is 1 indicating the panic-tense gait, the recognition result is 0 indicating the normal gait, and if in the recognition result, the recognition result is a probability, for example, the probability of outputting the panic-tense gait is 0.8, the similarity of the panic-tense gait of this step is 80%. By identifying each step of the user and determining the gait category or similarity, the panic gait can be accurately detected, and the user can be reminded of timely finding parkinsonism.

In some embodiments, the microcontroller and inertial measurement unit are integrated into a monitoring device that is small and wearable on the foot, such as an upper. The microcontroller can receive the test gait sample sent by the inertial measurement unit in real time, analyze potential characteristics and obtain gait categories. On one hand, the calculation pressure is reduced, the data transmission delay is reduced, and the real-time detection can be performed. On the other hand, the method has higher detection accuracy, stronger algorithm robustness and generalization capability.

Embodiments of the present application also provide a computer readable storage medium storing computer executable instructions for causing an electronic device to perform a method for training a panic-tense gait recognition model provided by the embodiments of the present application, for example, a method for training a panic-tense gait recognition model as shown in fig. 3-5, or a virtual fitting method provided by the embodiments of the present application, for example, a gait recognition method as shown in fig. 6.

In some embodiments, the storage medium may be FRAM, ROM, PROM, EPROM, EE PROM, flash memory, magnetic surface memory, optical disk, or CD-ROM; but may be a variety of devices including one or any combination of the above memories.

In some embodiments, the executable instructions may be in the form of programs, software modules, scripts, or code, written in any form of programming language (including compiled or interpreted languages, or declarative or procedural languages), and they may be deployed in any form, including as stand-alone programs or as modules, components, subroutines, or other units suitable for use in a computing environment.

As an example, the executable instructions may, but need not, correspond to files in a file system, may be stored as part of a file that holds other programs or data, such as in one or more scripts in a hypertext markup language (html, hyper TextMarkup Language) document, in a single file dedicated to the program in question, or in multiple coordinated files (e.g., files that store one or more modules, sub-programs, or portions of code).

As an example, executable instructions may be deployed to be executed on one computing device (including devices such as smart terminals and servers) or on multiple computing devices located at one site, or on multiple computing devices distributed across multiple sites and interconnected by a communication network.

Embodiments of the present application also provide a computer-readable storage medium storing a computer program comprising program instructions that, when executed by a computer, cause the computer to perform a method of training a panic-tense gait recognition model or a gait recognition method as in the previous embodiments.

It should be noted that the above-described apparatus embodiments are merely illustrative, and the units described as separate units may or may not be physically separate, and units shown as units may or may not be physical units, may be located in one place, or may be distributed over a plurality of network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

From the above description of embodiments, it will be apparent to those skilled in the art that the embodiments may be implemented by means of software plus a general purpose hardware platform, or may be implemented by hardware. Those skilled in the art will appreciate that all or part of the processes implementing the methods of the above embodiments may be implemented by a computer program for instructing relevant hardware, where the program may be stored in a computer readable storage medium, and where the program may include processes implementing the embodiments of the methods described above. The storage medium may be a magnetic disk, an optical disk, a Read-Only Memory (ROM), a random access Memory (Random Access Memory, RAM), or the like.

Finally, it should be noted that: the above embodiments are only for illustrating the technical solution of the present application, and are not limiting; the technical features of the above embodiments or in the different embodiments may also be combined within the idea of the application, the steps may be implemented in any order, and there are many other variations of the different aspects of the application as described above, which are not provided in detail for the sake of brevity; although the application has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit of the application.

Claims

1. A method of training a panic-tense gait recognition model, comprising:

acquiring a plurality of gait samples, wherein the gait samples comprise acceleration signal waves and angular velocity signal waves acquired by an inertial measurement unit for acquiring walking steps of feet, one gait sample is marked with a label, and the label comprises a panic gait or a normal gait;

and performing iterative training on the two classification algorithm models by adopting a plurality of divided gait cycle wave bands to obtain the panic-tense gait recognition model.

2. The method of claim 1, wherein dividing the gait sample according to the periodicity of the gait sample results in a plurality of gait cycle bands, comprising:

acquiring a peak of the acceleration signal wave in the motion direction;

determining a start index and an end index of each gait cycle wave band according to each wave peak;

intercepting the gait samples according to the starting index and the ending index of each gait cycle wave band to obtain a plurality of gait cycle wave bands.

3. The method of claim 2, wherein determining a start index and an end index for each of the gait cycle bands from each of the peaks comprises:

and sequentially sliding the middle position of the sliding window to each wave crest, and determining a start index and an end index of the gait cycle wave band according to the target wave band covered by the sliding window positioned at one wave crest.

4. A method according to claim 3, wherein said determining a start index and an end index of the gait cycle band from the target band covered by the sliding window at a peak comprises:

determining an index corresponding to a first stable wave band positioned at the left side of the wave crest in the target wave band according to the differential sequence, and determining a starting index of the gait cycle wave band according to the index corresponding to the first stable wave band;

and determining an index corresponding to a second stable wave band positioned on the right side of the wave crest in the target wave band according to the differential sequence, and determining an ending index of the gait cycle wave band according to the index corresponding to the second stable wave band.

5. The method of claim 4, wherein determining, from the differential sequence, an index corresponding to a first stationary band of the target bands that is to the left of the peak comprises:

screening out indexes of which absolute values of values in a differential sequence positioned at the left side of the wave crest are smaller than or equal to a preset threshold value, and taking the indexes as a first candidate index set;

6. The method of claim 5, wherein determining the start index of the gait cycle band from the index corresponding to the first plateau band comprises:

determining the first stable wave band from the target wave band according to the index corresponding to the first stable wave band; and obtaining the maximum value of the signal value in the first stable wave band, and taking the index corresponding to the maximum value as the starting index of the gait cycle wave band.

7. The method of claim 4, wherein the determining, from the differential sequence, an index corresponding to a second plateau band located to the right of the peak in the target band, and determining, from the index corresponding to the second plateau band, an ending index of the gait cycle band comprises:

screening indexes of which absolute values of values in the differential sequence positioned on the right side of the wave crest are smaller than or equal to a preset threshold value, and taking the indexes as a second candidate index set;

And determining the second stable wave band from the target wave band according to the index corresponding to the second stable wave band, acquiring the minimum value of the signal value in the second stable wave band, and taking the index corresponding to the minimum value as the ending index of the gait cycle wave band.

8. The method of claim 1, wherein prior to the step of iteratively training a binary classification algorithm model using the partitioned ones of the gait cycle bands to obtain the recognition model of a panic gait, the method further comprises:

and adjusting the length of each gait cycle wave band to be a target length.

9. The method of claim 8, wherein said adjusting the length of each of the gait cycle bands to a target length comprises:

and (3) adopting linear interpolation to adjust the length of each gait cycle wave band to be a target length.

10. The method of claim 1, wherein prior to the step of dividing the gait sample according to the periodicity of the gait sample to obtain a plurality of gait cycle bands, the method further comprises:

and normalizing the gait sample.

11. A gait recognition method, comprising:

dividing the test gait sample according to the periodicity of the test gait sample to obtain a plurality of test gait cycle bands;

inputting the plurality of test gait cycle bands into a panic-tense gait recognition model and outputting a gait class to which each step belongs, wherein the panic-tense gait recognition model is trained by the method for training the panic-tense gait recognition model according to any one of claims 1-10.

12. An electronic device, comprising:

at least one processor; and

a memory communicatively coupled to the at least one processor; wherein the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the method of any one of claims 1-11.

13. A computer readable storage medium storing computer executable instructions for causing a computer device to perform the method of any one of claims 1-11.