Photoelectric multi-mode gait recognition method, system, equipment and medium

The photoelectric multimodal gait recognition method enhances accuracy and practicality by combining electromyographic and pressure signals, addressing limitations of computer vision-based systems with lower-cost sensors for improved gait recognition.

CN119073964APending Publication Date: 2024-12-06NORTH CHINA UNIVERSITY OF TECHNOLOGY
View PDF 0 Cites 2 Cited by

Patent Information

Application Number
CN202411228686.X
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-09-03
Publication Date
2024-12-06

AI Technical Summary

Technical Problem

Current gait recognition methods using computer vision are limited by hardware costs, camera viewing angles, and light changes, leading to reduced recognition accuracy and poor practicality, hindering widespread application.

Method used

A photoelectric multimodal gait recognition method that combines electromyographic signals from leg muscles and pressure signals from the sole using fiber optic sensors, processing high-frequency and low-frequency signals separately for enhanced feature extraction and recognition.

Benefits of technology

Improves gait recognition accuracy and practicality by utilizing lower-cost electromyographic and fiber optic pressure sensors, providing robust data for comprehensive gait analysis and reducing hardware requirements.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119073964A_ABST
    Figure CN119073964A_ABST
Patent Text Reader

Abstract

The invention relates to the technical field of gait recognition, and discloses a photoelectric multi-mode gait recognition method, system and device and a medium. The photoelectric multi-mode gait recognition method comprises the following steps: acquiring an electromyographic signal of a leg of a target object, wherein the electromyographic signal is used for representing a muscle activity condition of the leg of the target object in a gait process; acquiring a pressure signal of the sole of the target object in the gait process through an optical fiber pressure sensor; extracting a high-frequency signal and a low-frequency signal in the electromyographic signal, and performing feature extraction on the high-frequency signal and the low-frequency signal of the electromyographic signal to obtain a first gait feature of the target object; extracting a high-frequency signal and a low-frequency signal in the pressure signal, and performing feature extraction on the high-frequency signal and the low-frequency signal of the pressure signal to obtain a second gait feature of the target object; and performing gait recognition on the target object in combination with the first gait feature and the second gait feature. According to the photoelectric multi-mode gait recognition method, the accuracy and practicability of gait recognition can be improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of gait recognition, and in particular to a photoelectric multimodal gait recognition method, system, device and medium. Background Art

[0002] As people's demand for health monitoring increases, gait recognition technology has gradually become an important research direction in the field of medical health. By analyzing the gait characteristics of the human body when walking, gait recognition can help doctors and rehabilitation therapists evaluate patients' motor functions, monitor disease progression, and provide a scientific basis for postoperative rehabilitation.

[0003] Most gait recognition methods use computer vision methods, which use cameras to capture continuous videos or image sequences, use image processing algorithms to analyze the movement trajectory of the human body, and then recognize gait. However, its recognition performance will be severely reduced in the case of light changes, shadow interference, and limited camera viewing angle; and this method is limited by the clarity and image processing capabilities of the camera equipment. Building such a system often requires high hardware costs and powerful computing resources, including high-definition cameras, powerful computers, and complex image processing algorithms. These factors limit the popularization of this technology and the expansion of its application scenarios.

[0004] In summary, the current gait recognition methods have low recognition accuracy and poor practicality. Summary of the invention

[0005] The purpose of the present invention is to provide a photoelectric multimodal gait recognition method, system, device and medium, which can improve the accuracy and practicality of gait recognition.

[0006] To solve the above technical problems, an embodiment of the present invention provides a photoelectric multimodal gait recognition method, comprising the following steps:

[0007] Acquire electromyographic signals of the target object's legs, where the electromyographic signals are used to characterize muscle activity of the target object's legs during gait;

[0008] The pressure signal of the sole of the target subject during gait is obtained by using a fiber optic pressure sensor;

[0009] Extract high-frequency signals and low-frequency signals from the electromyographic signal, and perform feature extraction on the high-frequency signals and low-frequency signals of the electromyographic signal respectively to obtain the first step characteristics of the target object;

[0010] Extracting high-frequency signals and low-frequency signals from the pressure signal, and performing feature extraction on the high-frequency signals and low-frequency signals of the pressure signal respectively, to obtain a second gait feature of the target object;

[0011] The first gait feature and the second gait feature are combined to perform gait recognition on the target object.

[0012] In some optional embodiments, obtaining the electromyographic signal of the target object's leg includes:

[0013] The EMG signals of the target object’s legs are oversampled by the analog-to-digital converter built into the Nrf52810 chip.

[0014] In some optional embodiments, the electromyographic signals include electromyographic signals of the tibialis anterior muscle and the gastrocnemius muscle of the target subject's leg.

[0015] In some optional embodiments, the pressure signals include pressure signals of the heel, big toe and little toe of the sole of the target object.

[0016] In some optional embodiments, the first gait feature and the second gait feature are obtained by processing the electromyographic signal and the pressure signal respectively by using a wavelet packet decomposition method; wherein the wavelet packet decomposition method adopts a three-layer decomposition.

[0017] In some optional embodiments, before the electromyographic signal and the pressure signal are processed by the wavelet packet decomposition method, the method further includes:

[0018] The sliding window cutting method is used to synchronize the electromyographic signal and the pressure signal.

[0019] In some optional embodiments, the result of the gait recognition includes at least: sitting, standing, squatting, walking, running, going up stairs and going down stairs.

[0020] An embodiment of the present invention further provides a photoelectric multimodal gait recognition system, comprising:

[0021] A signal acquisition module is used to acquire electromyographic signals of the target object's legs, wherein the electromyographic signals are used to characterize the muscle activity of the target object's legs during gait; and to acquire pressure signals of the soles of the target object's feet during gait through an optical fiber pressure sensor;

[0022] A feature extraction module is used to extract high-frequency signals and low-frequency signals from the electromyographic signal, and perform feature extraction on the high-frequency signals and low-frequency signals of the electromyographic signal respectively to obtain the first gait feature of the target object; and to extract high-frequency signals and low-frequency signals from the pressure signal, and perform feature extraction on the high-frequency signals and low-frequency signals of the pressure signal respectively to obtain the second gait feature of the target object;

[0023] The gait recognition module is used to combine the first gait feature and the second gait feature to perform gait recognition on the target object.

[0024] An embodiment of the present invention also provides a computer device, comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the above-mentioned optoelectronic multimodal gait recognition method.

[0025] An embodiment of the present invention further provides a computer-readable storage medium storing a computer program, wherein the computer program implements the above-mentioned optoelectronic multimodal gait recognition method when executed by a processor.

[0026] The photoelectric multimodal gait recognition method provided by the present invention has at least the following beneficial effects:

[0027] Not only the electromyographic signals of the target object's legs, that is, the muscle activity status of the target object's legs during gait, are obtained, but also the pressure signals of the target object's soles during gait are obtained through fiber optic pressure sensors, thereby obtaining optoelectronic multimodal data used to reflect the target object's gait, which provides important data support for the comprehensive analysis of gait characteristics, ensuring that the subsequently extracted gait features can more accurately reflect the target object's gait, thereby improving the accuracy of gait recognition.

[0028] Moreover, when extracting features from electromyographic signals and pressure signals, the high-frequency signals and low-frequency signals therein are processed separately as gait features of the target object, that is, signals of different sizes and resolutions are processed separately, thereby improving the processing effect of gait data.

[0029] Furthermore, the hardware equipment includes electromyographic signal acquisition equipment and fiber optic pressure sensors, which have lower costs than computer vision methods, are conducive to the popularization of this technology and the expansion of application scenarios, and are more practical. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] One or more embodiments are exemplarily described by the pictures in the corresponding drawings, and these exemplary descriptions do not constitute limitations on the embodiments.

[0031] Figure 1 is a flow chart of a photoelectric multimodal gait recognition method provided according to an embodiment of the present invention;

[0032] Figure 2 is a schematic diagram of the positions of the tibialis anterior muscle and the gastrocnemius muscle in the lower limbs of the human body provided according to an embodiment of the present invention;

[0033] Figure 3 It is a schematic diagram of a design structure of a wearable wireless myoelectric acquisition system with a high degree of freedom provided according to an embodiment of the present invention;

[0034] Figure 4 is a schematic diagram of pressure monitoring at three points of a heel, a thumb and a little finger provided according to an embodiment of the present invention;

[0035] Figure 5 is a schematic diagram of wearing positions of an EMG device and optical sensors provided according to an embodiment of the present invention;

[0036] Figure 6 is a schematic diagram of a gait pattern provided according to an embodiment of the present invention;

[0037] Figure 7 is a schematic diagram of a sliding window data processing effect provided according to an embodiment of the present invention;

[0038] Figure 8 is a schematic diagram of an optoelectronic multimodal gait recognition system provided according to an embodiment of the present invention;

[0039] Fig. 9 It is a structural schematic diagram of a computer device provided according to an embodiment of the present invention. DETAILED DESCRIPTION

[0040] In order to make the purpose, technical scheme and advantages of the embodiments of the present invention clearer, the embodiments of the present invention will be described in detail below in conjunction with the accompanying drawings. However, it will be appreciated by those skilled in the art that in the embodiments of the present invention, many technical details are proposed in order to enable the reader to better understand the present invention. However, even without these technical details and various changes and modifications based on the following embodiments, the technical scheme claimed in the present invention can be implemented. The division of the following embodiments is for the convenience of description and should not constitute any limitation on the specific implementation of the present invention. The various embodiments can be combined and referenced with each other without contradiction.

[0041] An embodiment of the present invention relates to a photoelectric multimodal gait recognition method. The implementation details of the photoelectric multimodal gait recognition method of this embodiment are described in detail below. The following content is only the implementation details provided for the convenience of understanding and is not necessary for the implementation of this solution.

[0042] The specific process of the optoelectronic multimodal gait recognition method of this embodiment can be as follows: Figure 1 As shown, including:

[0043] Step 101 : Acquire electromyographic signals of the target object's legs, where the electromyographic signals are used to characterize the muscle activity status of the target object's legs during gait.

[0044] Step 102: Obtain pressure signals of the sole of the target object during gait through a fiber optic pressure sensor.

[0045] Step 103, extracting the high-frequency signal and the low-frequency signal in the electromyographic signal, and performing feature extraction on the high-frequency signal and the low-frequency signal of the electromyographic signal respectively, to obtain the first step posture feature of the target object.

[0046] Step 104 , extracting the high-frequency signal and the low-frequency signal from the pressure signal, and performing feature extraction on the high-frequency signal and the low-frequency signal of the pressure signal respectively, to obtain a second gait feature of the target object.

[0047] Step 105: Combining the first gait feature and the second gait feature, gait recognition is performed on the target object.

[0048] In this embodiment, not only the electromyographic signal of the target object's leg, that is, the muscle activity of the target object's leg during the gait process, is obtained, but also the pressure signal of the sole of the target object's foot during the gait process is obtained through the optical fiber pressure sensor, thereby obtaining the photoelectric multimodal data used to reflect the gait of the target object, providing important data support for the comprehensive analysis of gait characteristics, ensuring that the gait characteristics extracted later can more accurately reflect the gait of the target object, thereby improving the accuracy of gait recognition. In addition, when extracting features from the electromyographic signal and the pressure signal, the high-frequency signal and the low-frequency signal therein are processed respectively as the gait features of the target object, that is, the signals of different sizes and resolutions are processed respectively, which improves the processing effect of the gait data. Furthermore, the hardware equipment is the acquisition equipment of the electromyographic signal and the optical fiber pressure sensor, and the cost is lower than that of the computer vision method, which is conducive to the popularization of the technology and the expansion of application scenarios, and has strong practicality.

[0049] In step 101, the target object is a human body whose gait needs to be recognized, and the target object's leg muscles are mainly exerting force during the gait process. Therefore, it is necessary to obtain the electromyographic signal of the target object's leg to determine the muscle activity status of the target object's leg during the gait process, thereby reflecting the muscle coordination and force output of the target object's leg during the gait process.

[0050] In a specific implementation, the human lower limbs contain many complex muscle groups, among which the changes in electrical activity of the tibialis anterior (TA) and gastrocnemius (GT) during gait can clearly reflect the muscle coordination and force output in different action states. Compared with monitoring the entire lower limb muscle group, monitoring only the tibialis anterior and gastrocnemius is not only more convenient, but also has no obvious loss in the accuracy and practicality of gait analysis. It can achieve a small amount of muscle monitoring with clear functions. Therefore, the electromyographic signals obtained in this embodiment include the electromyographic signals of the tibialis anterior and gastrocnemius of the target object's leg. Among them, the positions of the tibialis anterior and gastrocnemius in the lower limbs of the human body can be referred to. Figure 2 .

[0051] In some embodiments, the present invention designs a high-degree-of-freedom wearable wireless electromyography acquisition system, which acquires the electromyography signal of the target object's leg in the following manner:

[0052] Since the amplitude of human electromyographic signals is very weak, usually between 0.01mV and 5mV, mainly concentrated in 0.05mV to 1mV, and the signal frequency range is 10Hz to 300Hz, mainly concentrated in 10Hz to 150Hz, there are strict requirements on the hardware accuracy and sampling rate of the high-freedom wearable wireless electromyographic acquisition system. In order to meet these requirements, this embodiment uses the Nrf52810 chip, the built-in analog-to-digital converter (Analog-to-Digital Converter ADC) of the chip supports high-resolution oversampling function, thereby improving the signal acquisition accuracy.

[0053] At the same time, in view of the weak characteristics of electromyographic signals, the high-degree-of-freedom wearable wireless electromyographic acquisition system adopts a differential input method to effectively eliminate the common-mode signal caused by the body's own noise. The electromyographic EMG signal is first processed by the electrostatic protection circuit and the amplification circuit, and then input into the ADC for sampling. This design ensures that the signal can maintain high quality and stability during the acquisition process. In order to further improve the acquisition accuracy, a gyroscope module is also integrated into the circuit to assist in precision calibration. In addition, in order to achieve efficient wireless data transmission, the high-degree-of-freedom wearable wireless electromyographic acquisition system adopts a 2.4GHz private protocol for communication, which not only effectively avoids interference from other radio frequency signals, but also enhances the security of communication. The overall high-degree-of-freedom wearable wireless electromyographic acquisition system design structure of this embodiment can be as follows Figure 3 shown.

[0054] In step 102, the optical fiber pressure sensor is an intelligent gait monitoring insole based on elastomeric optical fiber, focusing on obtaining various gait parameters to meet the needs of medical care. Through a curved composite optical fiber sensing structure integrated with flexible transparent silicone rubber (Solaris), a smart phone is used as a signal generator and receiver at the same time to collect plantar gait signals without the need for additional experimental equipment, which greatly improves the portability and scalability of the system. Therefore, this embodiment obtains the optical fiber signal of the sole of the target object through the optical fiber pressure sensor, so as to reflect the pressure information of the sole of the target object during the gait process through the strength of the optical fiber signal.

[0055] In the specific implementation, for the sole pressure perception, this embodiment selects the heel, little toe and little toe three-point pressure monitoring (Heel, Little toe And Big toe), which can effectively capture the pressure distribution of the sole during walking and exercise, that is, the pressure signal obtained by this embodiment includes the pressure signals of the heel, little toe and little toe of the sole of the target object. Among them, the schematic diagram of the pressure monitoring of the heel, little toe and little toe can be seen in Figure 4 .

[0056] In one example, the fiber optic pressure sensor uses flexible optical fiber to obtain the strength signal of the light path to respond to pressure changes. Compared with traditional electric pressure sensors, flexible optical fiber is not only more economical, but also has the advantages of high biocompatibility, light weight and not susceptible to electromagnetic radiation interference. By monitoring these key areas, the dynamic changes in plantar pressure during gait can be accurately obtained, providing important data support for the comprehensive analysis of gait characteristics.

[0057] In practical applications, the hardware devices (i.e., EMG device and optical sensors) for obtaining the target object's leg electromyographic signals and sole optical fiber signals are shown in FIG. Figure 5 .

[0058] In step 103 and step 104, after obtaining the electromyographic signals of the target object's legs and the pressure signals of the soles of the feet, that is, the gait data of the target object, it is necessary to extract the gait features in these gait data for subsequent gait recognition.

[0059] Specifically, there are many ways to extract gait features, such as the gait-based time domain analysis method, which mainly focuses on the motion characteristics of the gait pattern, which are usually reflected in the local feature points of the overall waveform. Although parameters are often used to describe the gait time domain characteristics with clear physical meanings, the number of effective features is limited, and the details useful for classification are often ignored. The wavelet transform analysis method can be used to process the local details. The wavelet transform analysis method is applied to gait data for multi-level decomposition, time-frequency feature extraction, gait signal reconstruction and denoising. Wavelet analysis adjusts the window in real time, uses narrow window analysis at high frequencies, and uses wide window analysis at low frequencies, and flexibly processes signals of different sizes and resolutions. This adaptive feature is particularly suitable for processing time-varying and non-stationary signals, which improves the processing effect and classification accuracy of gait data. The wavelet packet decomposition method has the characteristics of low entropy, multi-resolution and decorrelation, and can further decompose the unprocessed high-frequency part, thereby extracting more effective information and more detailed spectral features. Although the low-frequency part of the original wavelet transform signal has been analyzed, the high-frequency part has not been processed. Since the high-frequency gait signal as a biomedical signal often contains a lot of useful information, this embodiment uses the wavelet packet decomposition method for in-depth analysis.

[0060] Specifically, when the present embodiment uses the wavelet packet decomposition method to process the electromyographic signal and the pressure signal, the first gait feature of the target object is obtained by extracting the high-frequency signal and the low-frequency signal in the electromyographic signal, and performing feature extraction on the high-frequency signal and the low-frequency signal of the electromyographic signal respectively. The second gait feature of the target object is obtained by extracting the high-frequency signal and the low-frequency signal in the pressure signal, and performing feature extraction on the high-frequency signal and the low-frequency signal of the pressure signal respectively. The first gait feature and the second gait feature serve as the basis for subsequent gait recognition.

[0061] Compared with the wavelet analysis method, the decomposition and reconstruction of wavelet packets provide more detailed calculations and higher resolution accuracy, and can better distinguish high-frequency and low-frequency signals. Among them, the wavelet packet transform decomposes the signal into multiple frequency bands by recursively decomposing all frequency bands. For a signal x[n], the wavelet packet coefficient is calculated as:

[0062]

[0063] In the formula, represents the wavelet packet coefficients at node level m, frequency scale j, and time position k; represents the wavelet packet basis function, which is constructed by scaling and shifting the mother wavelet ψ[n]; x[n] represents the discrete input signal, which is analyzed and decomposed.

[0064] Feature extraction usually uses energy features and entropy features to describe different characteristics of signals.

[0065]

[0066] in:

[0067]

[0068] In the formula, represents the energy characteristics at node level m and frequency scale j; It represents the entropy characteristics at node level m and frequency scale j, reflecting the complexity or uncertainty of the signal; represents the normalized coefficient probability, reflecting the contribution of each coefficient to the total energy.

[0069] In the specific implementation, although the calculation speed is faster, the denoising effect is limited; while the higher decomposition level can achieve the ideal effect, the calculation amount is huge and the time is long, so it is very important to choose the appropriate decomposition level. This embodiment uses wavelet packet decomposition to extract features from gait data, and selects 3-layer decomposition (m=3) to take into account both extraction effect and calculation efficiency, avoiding excessive running time.

[0070] In one example, before using the wavelet packet decomposition method to extract the features of the high-frequency signal and the low-frequency signal in the electromyographic signal and the pressure signal, the sliding window cutting method is first used to synchronize the electromyographic signal and the pressure signal. Since the original multimodal signal (i.e., the leg electromyographic signal and the plantar pressure signal) usually contains a large amount of redundant information, the method of sliding window cutting of the collected signal of the present invention first ensures the timing synchronization of the photoelectric signal channel within the window length, and then performs effective gait feature extraction based on the signal within the window length, thereby obtaining more regular matrix values, which accurately represent the features of the timing signals of different channels, and help to improve the efficiency of data processing and calculation.

[0071] In step 104, after extracting gait features (i.e., the first gait feature and the second gait feature) from the electromyographic signal and the pressure signal, the gait of the target object is recognized by combining the two. The result of gait recognition is the gait pattern of the target object, which at least includes: sitting, standing, squatting, walking, running, climbing stairs, and descending stairs.

[0072] In one embodiment, plantar pressure and electromyographic signal multimodal data are obtained by simulating daily activities (ADL) to study and analyze gait characteristics under different postures, and to evaluate the performance of the photoelectric multimodal gait recognition method of the present invention in practical applications. This embodiment recruited eight healthy subjects without gait disorders to participate. Each subject was required to wear a detection device (including electromyographic signal acquisition equipment and optical fiber signal acquisition equipment) on the right foot in the test site. The subject was required to simulate Figure 6 and the 7 common daily activities shown in Table 1 .

[0073] Table 1

[0074]

[0075]

[0076] After obtaining the EMG signal and the Fiber signal, the data needs to be cropped to facilitate classification and identification using a machine learning algorithm. A 2-second sliding window can be used to capture the data, and the sliding window step is set to 1 second. The sliding window processing method can effectively smooth the data, reduce the impact of noise, and maintain the continuity of the data. Figure 7 Valid data segments were extracted from each type of action, and a total of 500 valid segments were obtained. For all 7 types of actions, a total of 3,500 training samples were obtained, which will be used for subsequent gait feature analysis and performance evaluation of signal acquisition equipment.

[0077] In this embodiment, the K-nearest neighbor (KNN) algorithm is used to classify gait patterns. The data set is divided into a training set and a validation set with a ratio of 8:2, and the average classification accuracy is calculated using 10 cross-validation methods. The superiority of the signal acquisition device and algorithm is verified by changing the two dimensions. The evaluation index uses accuracy (ACC) to reflect the proportion of correctly classified events in the total events in the classification results.

[0078]

[0079] In the formula, TP (True Positives) is the number of samples correctly classified as positive; TN (True Negatives) is the number of samples correctly classified as negative; FP (False Positives) is the number of negative samples incorrectly classified as positive (false positives); FN (False Negatives) is the number of positive samples incorrectly classified as negative (false negatives).

[0080] The classification verification results using the KNN algorithm are shown in Table 2. Combined with the gait classification accuracy table, it can be seen that the accuracy of gait recognition using a single electromyographic signal is only 81.973%, and the accuracy of using a fiber optic pressure sensor is 85.578%. The use of an optoelectronic fusion multimodal sensor (combination of two signals) can significantly improve the gait classification accuracy to 92.109%.

[0081]

[0082] The photoelectric multimodal gait recognition method of the present invention has broad application prospects in the fields of medical health, sports training, rehabilitation therapy, etc., especially in the postoperative monitoring of rehabilitation patients. This technology can provide more scientific and accurate data support to help medical practitioners make more effective treatment decisions. It has the following advantages:

[0083] (1) Convenience of wearing and monitoring:

[0084] Not only does it focus on improving recognition accuracy, but it also takes into account the user's wearing comfort and ease of operation. The flexible fiber optic pressure sensor is arranged in the insole and combined with the wireless electromyography acquisition system. The entire system is lightweight and does not affect daily activities.

[0085] (2) Robustness of data collection:

[0086] By fusing data from different modalities, the present invention can still maintain high robustness in complex environments. Especially in noisy environments, the multimodal system can effectively filter redundant information and ensure the stability of data collection.

[0087] (3) Gait recognition effect is significant:

[0088] The accuracy of gait recognition is significantly better than that of a single sensor mode, whether it is normal walking, running or other complex gait patterns, while significantly reducing the overall cost, it still maintains excellent gait recognition performance. This cost-effective solution not only greatly enhances the market value of the technology, but also lays the foundation for its large-scale application and popularization. By precisely controlling costs and improving performance, the technology of the present invention shows broad market prospects and can meet diverse application needs.

[0089] The step division of the various methods above is only for clear description. When implemented, they can be combined into one step or some steps can be split and decomposed into multiple steps. As long as they include the same logical relationship, they are all within the protection scope of the present invention. Adding insignificant modifications or introducing insignificant designs to the algorithm or process without changing the core design of the algorithm and process are all within the protection scope of the invention.

[0090] Another embodiment of the present invention relates to a photoelectric multimodal gait recognition system. The implementation details of the photoelectric multimodal gait recognition system of this embodiment are described in detail below. The following content is only for the convenience of understanding the implementation details, and is not necessary for the implementation of this solution. The schematic diagram of the photoelectric multimodal gait recognition system of this embodiment can be as follows: Figure 8 As shown, it includes: a signal acquisition module 801, a feature extraction module 802 and a gait recognition module 803.

[0091] Specifically, the signal acquisition module 801 is used to acquire the electromyographic signal of the target object's leg, which is used to characterize the muscle activity of the target object's leg during gait; and to acquire the pressure signal of the target object's sole during gait through a fiber optic pressure sensor.

[0092] The feature extraction module 802 is used to extract the high-frequency signal and the low-frequency signal in the electromyographic signal, and perform feature extraction on the high-frequency signal and the low-frequency signal of the electromyographic signal respectively to obtain the first gait feature of the target object; and to extract the high-frequency signal and the low-frequency signal in the pressure signal, and perform feature extraction on the high-frequency signal and the low-frequency signal of the pressure signal respectively to obtain the second gait feature of the target object.

[0093] The gait recognition module 803 is used to perform gait recognition on the target object by combining the first gait feature and the second gait feature.

[0094] It is not difficult to find that this embodiment is a system embodiment corresponding to the above method embodiment, and this embodiment can be implemented in conjunction with the above method embodiment. The relevant technical details and technical effects mentioned in the above embodiment are still valid in this embodiment, and in order to reduce repetition, they are not repeated here. Accordingly, the relevant technical details mentioned in this embodiment can also be applied in the above embodiment.

[0095] It is worth mentioning that all modules involved in this embodiment are logic modules. In practical applications, a logic unit can be a physical unit, a part of a physical unit, or a combination of multiple physical units. In addition, in order to highlight the innovative part of the present invention, this embodiment does not introduce units that are not closely related to solving the technical problem proposed by the present invention, but this does not mean that there are no other units in this embodiment.

[0096] Another embodiment of the present invention relates to a computer device, such as Fig. 9 As shown, it includes: at least one processor 901; and a memory 902 that is communicatively connected to the at least one processor 901; wherein the memory 902 stores instructions that can be executed by the at least one processor 901, and the instructions are executed by the at least one processor 901 so that the at least one processor 901 can execute the optoelectronic multimodal gait recognition method in the above-mentioned embodiments.

[0097] Among them, the memory and the processor are connected in a bus manner, and the bus may include any number of interconnected buses and bridges, and the bus connects various circuits of one or more processors and memories together. The bus can also connect various other circuits such as peripherals, voltage regulators, and power management circuits, which are well known in the art and are therefore not further described herein. The bus interface provides an interface between the bus and the transceiver. The transceiver can be one element or multiple elements, such as multiple receivers and transmitters, providing a unit for communicating with various other devices on a transmission medium. The data processed by the processor is transmitted on a wireless medium via an antenna, and further, the antenna also receives data and transmits the data to the processor.

[0098] The processor is responsible for managing the bus and general processing, and can also provide various functions, including timing, peripheral interfaces, voltage regulation, power management, and other control functions. Memory can be used to store data used by the processor when performing operations.

[0099] Another embodiment of the present invention relates to a computer-readable storage medium storing a computer program, which implements the above method embodiment when executed by a processor.

[0100] That is, those skilled in the art can understand that all or part of the steps in the above-mentioned embodiment method can be completed by instructing the relevant hardware through a program, and the program is stored in a storage medium, including a number of instructions for a device (which can be a single-chip microcomputer, chip, etc.) or a processor (processor) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (Read-Only Memory, referred to as: ROM), random access memory (Random Access Memory, referred to as: RAM), disk or optical disk and other media that can store program codes.

[0101] Those skilled in the art will appreciate that the above embodiments are specific embodiments for implementing the present invention, and in actual applications, various changes may be made thereto in form and detail without departing from the spirit and scope of the present invention.

Claims

1. A photoelectric multimodal gait recognition method, characterized in that: The method comprises: Acquire electromyographic signals of the target object's legs, where the electromyographic signals are used to characterize muscle activity of the target object's legs during gait; The pressure signal of the sole of the target subject during gait is obtained by using a fiber optic pressure sensor; Extract high-frequency signals and low-frequency signals from the electromyographic signal, and perform feature extraction on the high-frequency signals and low-frequency signals of the electromyographic signal respectively to obtain the first step characteristics of the target object; Extracting high-frequency signals and low-frequency signals from the pressure signal, and performing feature extraction on the high-frequency signals and low-frequency signals of the pressure signal respectively, to obtain a second gait feature of the target object; The first gait feature and the second gait feature are combined to perform gait recognition on the target object.

2. The optoelectronic multimodal gait recognition method according to claim 1, characterized in that: The step of obtaining the electromyographic signal of the target object's leg includes: The EMG signals of the target object’s legs are oversampled by the analog-to-digital converter built into the Nrf52810 chip.

3. The photoelectric multimodal gait recognition method according to claim 2, characterized in that: The electromyographic signals include electromyographic signals of the tibialis anterior muscle and the gastrocnemius muscle of the target object's leg.

4. The photoelectric multimodal gait recognition method according to any one of claims 1 to 3, characterized in that: The pressure signals include pressure signals of the heel, big toe and little toe of the sole of the target object.

5. The optoelectronic multimodal gait recognition method according to claim 1, characterized in that: The first gait feature and the second gait feature are obtained by processing the electromyographic signal and the pressure signal respectively by using a wavelet packet decomposition method; wherein the wavelet packet decomposition method adopts a three-layer decomposition.

6. The optoelectronic multimodal gait recognition method according to claim 5, characterized in that: Before the myoelectric signal and the pressure signal are processed by the wavelet packet decomposition method, the method further includes: The sliding window cutting method is used to synchronize the electromyographic signal and the pressure signal.

7. The optoelectronic multimodal gait recognition method according to claim 1, characterized in that: The results of the gait recognition include at least: sitting, standing, squatting, walking, running, going up stairs and going down stairs.

8. An optoelectronic multimodal gait recognition system, characterized in that: The system comprises: A signal acquisition module is used to acquire electromyographic signals of the target object's legs, wherein the electromyographic signals are used to characterize the muscle activity of the target object's legs during gait; and to acquire pressure signals of the soles of the target object's feet during gait through an optical fiber pressure sensor; A feature extraction module is used to extract high-frequency signals and low-frequency signals from the electromyographic signal, and perform feature extraction on the high-frequency signals and low-frequency signals of the electromyographic signal respectively to obtain the first gait feature of the target object; and to extract high-frequency signals and low-frequency signals from the pressure signal, and perform feature extraction on the high-frequency signals and low-frequency signals of the pressure signal respectively to obtain the second gait feature of the target object; The gait recognition module is used to combine the first gait feature and the second gait feature to perform gait recognition on the target object.

9. A computer device, characterized in that: include: at least one processor; And, a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the optoelectronic multimodal gait recognition method as described in any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the optoelectronic multimodal gait recognition method according to any one of claims 1 to 7 is implemented.

Citation Information

Cited By

  • Abnormal gait correction method and device based on foot electrostatic gait signal

    CN119385557A

  • Device and method for recognizing human behaviors in ecological protection area based on gait feature recognition

    CN120470319A