Roller skate collision flash warning method and system

Abstract

The invention relates to the technical field of roller skating, in particular to a roller skate collision flash warning method and system, and the method comprises the steps: obtaining the rule data of a roller skate and a racing track at a target stage, the historical data comprises the rule position information, the rule speed information and the rule posture information, and the target stage is from the starting point to the ending point of the racing track; acquiring real-time position information, real-time speed information and real-time attitude information of the roller skate and the throwing body in the target stage; a certain number of millimeter wave radar devices are installed around the detection space, and the millimeter wave radar devices are used for collecting human body posture data; the angle position relationship between the millimeter wave radar equipment and the human body comprises a front facing relationship and an inclined facing relationship; and obtaining a detection conclusion of human body tumble detection from the plurality of detection results according to the angular position relationship between the millimeter wave radar equipment and the human body. The system has the advantages of being capable of timely informing the participants of deviating from the racing track and timely sending tumbling information, and achieving rapid rescue.

Description

Technical Field

[0001] This invention relates to the field of roller skating technology, and in particular to a method and system for providing flashing warnings of collisions with roller skates. Background Technology

[0002] Roller skating, a sport that combines fitness, entertainment, and competition, offers numerous health benefits. First, it improves coordination and balance, enhancing flexibility. Second, it strengthens cardiovascular function, increases cardiopulmonary endurance, and reduces the risk of heart and respiratory illnesses. Furthermore, roller skating builds muscle strength and endurance, aids in weight loss and body shaping, and improves balance.

[0003] However, roller skating competitions are prone to accidents such as deviating from the track and falling at high speeds. Therefore, a roller skate collision flashing warning method and system are proposed to promptly inform participants of deviations from the track and send out information about falls, enabling rapid rescue. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to overcome the shortcomings of the prior art and provide a method and system for flashing warnings of roller skate collisions, which can promptly inform participants of deviation from the track and promptly send out information about falls, so as to achieve rapid rescue.

[0005] To solve the above-mentioned technical problems, the present invention adopts the following technical solution: A method for flashing warning lights in the presence of roller skates upon collision includes: Acquire the rule data of roller skates and track at the target stage. Historical data includes rule position information, rule speed information, and rule posture information. The target stage is from the start point to the finish point of the track. Acquire real-time position, velocity, and attitude information of the roller skates and the projectile during the target phase; A certain number of millimeter-wave radar devices are installed around the detection space to collect human posture data. The angular relationship between millimeter-wave radar equipment and the human body includes: face-to-face and oblique-face; based on the angular relationship between millimeter-wave radar equipment and the human body, the detection conclusion of human fall detection is obtained from multiple detection results; The attitude angle difference of the target stage is calculated based on real-time attitude information, the spatial distance value of the target stage is calculated based on real-time position information, and the relative velocity value of the target stage is calculated based on real-time velocity information. When the difference in posture angle exceeds the threshold, a warning audio and flashing light are set on the roller skates. If the roller skates do not deviate from the track, the roller skates do not play the warning audio or flashing light. If the roller skates deviate from the track and are not in an acceleration state, the roller skates play the warning audio. If the roller skates deviate from the track and are in an acceleration state, the roller skates play the warning audio and flashing light.

[0006] Preferably, the facing relationship is such that the angle between the upper body and the center of gravity is less than 30 degrees, and the angle between the lower body and the ground is greater than 45 degrees. A slanted plane is a plane where the angle between the upper body and the center of gravity is greater than 30 degrees, and the angle between the lower body and the ground is less than 45 degrees.

[0007] As a preferred approach, human posture data is labeled to obtain labeled human motion data, which is then used as the dataset for training the human fall detection model.

[0008] As a preferred method, human posture data and rule data are matched in the time domain to generate a synchronous data stream; Based on synchronous data streams, a taxiing scene feature space is constructed through orthogonal projection. The taxiing driving scene feature space includes multiple taxiing scene feature vectors. Based on the feature space of the gliding driving scenario, an online incremental learning algorithm is used to dynamically generate a set of scene-adaptive interaction strategies; based on the set of interaction strategies, warning audio is generated through roller skates.

[0009] As a preferred option, the roller skates have a built-in controller. The controller runs a human fall detection model, inputs the human posture data collected by each millimeter-wave radar device into the human fall detection model, and obtains multiple detection results equal to the preset number. Based on the angular positional relationship between the millimeter-wave radar device and the human body, the final detection conclusion of human fall detection is obtained from the multiple detection results.

[0010] As a preferred option, the millimeter-wave radar equipment can be replaced with an infrared laser transmitter or an ultrasonic transmitter, and simultaneously receive reflected optical signals and scattered acoustic signals from the human body.

[0011] As a preferred approach, the feature vector of the gliding scene is input into a pre-trained convolutional spiking neural network. The local temporal pattern is extracted through the convolutional sub-network, the global spatiotemporal correlation is captured through the spiking sub-network, and the local temporal pattern and global spatiotemporal correlation are fused together with attention weights to form a gliding pattern classifier.

[0012] A roller skate collision flashing warning system includes a history module for acquiring rule data of roller skates and track at a target stage. The historical data includes rule position information, rule speed information, and rule posture information. The target stage is from the start point to the finish point of the track. The real-time module is used to acquire the real-time position, speed, and attitude information of the roller skates and the projectile during the target phase. The installation module is used to install a certain number of millimeter-wave radar devices around the detection space and use the millimeter-wave radar devices to collect human posture data. The calculation module is used to determine the angular relationship between the millimeter-wave radar device and the human body, including: face-to-face and oblique face; based on the angular relationship between the millimeter-wave radar device and the human body, the detection conclusion of human fall detection is obtained from multiple detection results; The attitude angle difference of the target stage is calculated based on real-time attitude information, the spatial distance value of the target stage is calculated based on real-time position information, and the relative velocity value of the target stage is calculated based on real-time velocity information. When the difference in posture angle exceeds the threshold, a warning audio and flashing light are set on the roller skates. If the roller skates do not deviate from the track, the roller skates do not play the warning audio or flashing light. If the roller skates deviate from the track and are not in an acceleration state, the roller skates play the warning audio. If the roller skates deviate from the track and are in an acceleration state, the roller skates play the warning audio and flashing light.

[0013] A computer device includes: a memory and a processor, and a computer program stored in the memory, which, when executed on the processor, implements the roller skate collision flashing warning method described above.

[0014] A computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the roller skate collision flashing warning method described above.

[0015] The beneficial effects of using the present invention are as follows: 1. In terms of the judgment process, the system comprehensively assesses separation safety by acquiring real-time information on the participant's position, speed, and attitude relative to the track, and comparing this information with revised theoretical thresholds across multiple indicators. If any safety conditions are not met, a tiered intervention mechanism is immediately activated, progressing from single actions to combined interventions to ensure effective risk control. An emergency termination procedure is initiated if necessary. Furthermore, the early warning function further enhances safety capabilities, providing advance warnings of potential collision risks.

[0016] Acquire the rule data of roller skates and track at the target stage. Historical data includes rule position information, rule speed information, and rule posture information. The target stage is from the start point to the finish point of the track. Acquire real-time position, velocity, and attitude information of the roller skates and the projectile during the target phase; A certain number of millimeter-wave radar devices are installed around the detection space to collect human posture data. The angular relationship between millimeter-wave radar equipment and the human body includes: facing directly and facing at an angle; A face-to-face relationship is defined as an angle of less than 30 degrees between the upper body and the center of gravity, and an angle of more than 45 degrees between the lower body and the ground. A slanted plane is a plane where the angle between the upper body and the center of gravity is greater than 30 degrees, and the angle between the lower body and the ground is less than 45 degrees.

[0017] At the same time, falling includes falling forward and falling backward. Therefore, when facing each other, the angle between the upper body and the center of gravity of the human body is in the range of [-30°, 30°], and the angle between the lower body and the ground is in the range of [45°, 90°]. Therefore, the angle between the upper body and the center of gravity of the human body is in the range of [30°, 150°], and the angle between the lower body and the ground is in the range of [0°, 45°].

[0018] Based on the angular relationship between the millimeter-wave radar device and the human body, the detection conclusion of human fall detection is obtained from multiple detection results. This invention installs a preset number of millimeter-wave radar devices around the detection space, fully considering the angular relationship between the millimeter-wave radar and the human body, thereby improving the comprehensiveness of the collected data and providing a strong basis for subsequent calculation of the final detection conclusion.

[0019] The feature vector of the gliding scene is input into a pre-trained convolutional spiking neural network. The local temporal pattern is extracted through the convolutional sub-network, the global spatiotemporal correlation is captured through the spiking sub-network, and the local temporal pattern and global spatiotemporal correlation are fused together with attention weights to form a gliding pattern classifier.

[0020] CSNN combines the feature extraction capabilities of Convolutional Neural Networks (CNNs) with the temporal processing capabilities of Spiking Neural Networks (SNNs), effectively handling human pose data with temporal characteristics. A large amount of participant human pose data was collected, labeled accordingly, and used to train the CSNN model. During training, a supervised learning algorithm was employed to optimize model parameters by minimizing the error between predicted and true labels. The trained CSNN model was validated using independent test datasets to evaluate its classification accuracy and generalization ability. Based on the validation results, the model was optimized and adjusted, such as by adjusting the network structure and adding regularization terms, to improve its performance.

[0021] These features and advantages of the present invention will be disclosed in detail in the following specific embodiments and accompanying drawings. Attached Figure Description

[0022] The invention will be further described below with reference to the accompanying drawings: Figure 1This is a schematic diagram of the collision flashing warning method according to Embodiment 1 of the present invention. Detailed Implementation

[0023] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application.

[0024] The concepts involved in this application will first be explained with reference to the accompanying drawings. It should be noted that the following explanation of each concept is only to make the content of this application easier to understand and does not imply any limitation on the scope of protection of this application.

[0025] Example 1: A method for flashing warning lights in the event of a roller skate collision, such as... Figure 1 As shown, it includes: Acquire the rule data of roller skates and track at the target stage. Historical data includes rule position information, rule speed information, and rule posture information. The target stage is from the start point to the finish point of the track. Acquire real-time position, velocity, and attitude information of the roller skates and the projectile during the target phase; A certain number of millimeter-wave radar devices are installed around the detection space to collect human posture data. The angular relationship between millimeter-wave radar equipment and the human body includes: facing directly and facing at an angle; A face-to-face relationship is defined as an angle of less than 30 degrees between the upper body and the center of gravity, and an angle of more than 45 degrees between the lower body and the ground. A slanted plane is a plane where the angle between the upper body and the center of gravity is greater than 30 degrees, and the angle between the lower body and the ground is less than 45 degrees.

[0026] At the same time, falling includes falling forward and falling backward. Therefore, when facing each other, the angle between the upper body and the center of gravity of the human body is in the range of [-30°, 30°], and the angle between the lower body and the ground is in the range of [45°, 90°]. Therefore, the angle between the upper body and the center of gravity of the human body is in the range of [30°, 150°], and the angle between the lower body and the ground is in the range of [0°, 45°].

[0027] Based on the angular relationship between the millimeter-wave radar device and the human body, the detection conclusion of human fall detection is obtained from multiple detection results. The present invention installs a preset number of millimeter-wave radar devices around the detection space, fully considering the angular relationship between the millimeter-wave radar and the human body, thereby improving the comprehensiveness of the collected data and providing a strong basis for subsequent calculation of the final detection conclusion. The attitude angle difference of the target stage is calculated based on real-time attitude information, the spatial distance value of the target stage is calculated based on real-time position information, and the relative velocity value of the target stage is calculated based on real-time velocity information. In terms of the judgment process, the system acquires real-time information on the participant's position, speed, and attitude relative to the track, and compares this information with revised theoretical thresholds using multiple indicators to comprehensively assess separation safety. If any safety conditions are not met, a tiered intervention mechanism is immediately activated, progressing from single actions to combined interventions to ensure effective risk control. An emergency termination procedure is initiated if necessary. Furthermore, the early warning system further enhances safety capabilities, providing advance warnings of potential collision risks.

[0028] When the difference in posture angle exceeds the threshold, a warning audio and flashing light are set on the roller skates. If the roller skates do not deviate from the track, the roller skates do not play the warning audio or flashing light. If the roller skates deviate from the track and are not in an acceleration state, the roller skates play the warning audio. If the roller skates deviate from the track and are in an acceleration state, the roller skates play the warning audio and flashing light.

[0029] Human pose data is labeled to obtain labeled human motion data. The labeled human motion data is used as the dataset for training the human fall detection model. The training parameters include a learning rate of 0.001 and Adam optimization.

[0030] The roller skates have a built-in controller that runs a human fall detection model. The controller inputs human posture data collected by each millimeter-wave radar device into the human fall detection model and obtains multiple detection results equal to the preset number. Based on the angular positional relationship between the millimeter-wave radar device and the human body, the final detection conclusion of human fall detection is obtained from the multiple detection results.

[0031] The dataset is input into the human fall detection model, generating multiple detection results equal to the number of preset millimeter-wave radar devices. A final detection conclusion is then determined from these results. Furthermore, based on the two positional relationships between the millimeter-wave radar device and the human body—facing directly or at an angle—the detection conclusion corresponding to the angled position is prioritized as the final conclusion. If neither a direct nor angled relationship exists, the detection conclusion corresponding to the closest facing position is selected, and so on. This process improves the accuracy of fall detection and avoids false alarms.

[0032] Human posture data is matched with rule data in the time domain to generate a synchronous data stream; Based on synchronous data streams, a taxiing scene feature space is constructed through orthogonal projection. The taxiing driving scene feature space includes multiple taxiing scene feature vectors. Based on the feature space of the gliding driving scenario, an online incremental learning algorithm is used to dynamically generate a set of scenario-adaptive interaction strategies. Based on this set of interaction strategies, warning audio is generated via roller skates. Furthermore, spatial audio coding technology is employed to encode the audio signals according to parameters such as sound source location, distance, and direction. Real-time broadcasts are then delivered through headphones, simulating the sound propagation characteristics in a real environment. This allows medical personnel to provide timely assistance, reducing injuries or fatalities, or to provide participants with race schedule and hazard warnings, thereby improving race safety.

[0033] Millimeter-wave radar equipment can be replaced with infrared laser emitters or ultrasonic emitters, and simultaneously receive reflected optical signals and scattered acoustic signals from the human body, increasing selectivity.

[0034] The feature vector of the gliding scene is input into a pre-trained convolutional spiking neural network. The local temporal pattern is extracted through the convolutional sub-network, the global spatiotemporal correlation is captured through the spiking sub-network, and the local temporal pattern and global spatiotemporal correlation are fused together with attention weights to form a gliding pattern classifier.

[0035] CSNN combines the feature extraction capabilities of Convolutional Neural Networks (CNNs) with the temporal processing capabilities of Spiking Neural Networks (SNNs), effectively handling human pose data with temporal characteristics. A large amount of participant human pose data was collected, labeled accordingly, and used to train the CSNN model. During training, a supervised learning algorithm was employed to optimize model parameters by minimizing the error between predicted and true labels. The trained CSNN model was validated using independent test datasets to evaluate its classification accuracy and generalization ability. Based on the validation results, the model was optimized and adjusted, such as by adjusting the network structure and adding regularization terms, to improve its performance.

[0036] A roller skate collision flashing warning system includes a history module for acquiring rule data of roller skates and track at a target stage. The historical data includes rule position information, rule speed information, and rule posture information. The target stage is from the start point to the finish point of the track. The real-time module is used to acquire the real-time position, speed, and attitude information of the roller skates and the projectile during the target phase. The installation module is used to install a certain number of millimeter-wave radar devices around the detection space and use the millimeter-wave radar devices to collect human posture data. The computing module is used to determine the angular positional relationship between the millimeter-wave radar equipment and the human body, including: face-to-face relationship and oblique face relationship. Based on the angular relationship between the millimeter-wave radar equipment and the human body, the detection conclusion of human fall detection is obtained from multiple detection results; the attitude angle difference of the target stage is calculated based on real-time attitude information, the spatial distance value of the target stage is calculated based on real-time position information, and the relative speed value of the target stage is calculated based on real-time speed information. When the difference in posture angle exceeds the threshold, a warning audio and flashing light are set on the roller skates. If the roller skates do not deviate from the track, the roller skates do not play the warning audio or flashing light. If the roller skates deviate from the track and are not in an acceleration state, the roller skates play the warning audio. If the roller skates deviate from the track and are in an acceleration state, the roller skates play the warning audio and flashing light.

[0037] A computer device includes: a memory and a processor, and a computer program stored in the memory, which, when executed on the processor, implements the roller skate collision flashing warning method described above.

[0038] A computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the roller skate collision flashing warning method described above.

[0039] Those skilled in the art will recognize that the units (or modules, steps, etc., hereinafter the same) of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0040] In the embodiments provided by the present invention, it should be understood that the division of units is only a logical functional division. In actual implementation, there may be other division methods, such as multiple units can be combined into one unit, one unit can be split into multiple units, or some features can be ignored.

[0041] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0042] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, read-only memory (ROM), random access memory (RAM), portable hard drives, magnetic disks, or optical disks.

[0043] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention, and they should all be covered within the scope of the claims and specification of the present invention.

Claims

1. A roller skate collision flash warning method, characterized by, The application relates to a wheel-skating shoe and a method for detecting a human body falling down. The application comprises the following steps: acquiring rule data of a wheel-skating shoe and a track at a target stage, the historical data comprising rule position information, rule speed information and rule posture information, the target stage being from a starting point to an end point of the track; acquiring real-time position information, real-time speed information and real-time posture information of the wheel-skating shoe and a throw-off body at the target stage; installing a certain number of millimeter-wave radar devices around a detection space, and collecting human body posture data by using the millimeter-wave radar devices; the angle position relationship between the millimeter-wave radar device and the human body comprises a front-facing relationship and an oblique-facing relationship; a detection conclusion of human body falling down is obtained from a plurality of detection results according to the angle position relationship between the millimeter-wave radar device and the human body; 2. A skatmg shoe collision flash warning method as claimed in claim 1, characteπzed by, a posture angle difference value of the target stage is calculated according to the real-time posture information, a spatial distance value of the target stage is calculated according to the real-time position information, and a relative speed value of the target stage is calculated according to the real-time speed information; when the posture angle difference value is greater than a threshold value, a warning audio and a flash light are set on the wheel-skating shoe to give a warning, if the wheel-skating shoe does not deviate from the track, the wheel-skating shoe does not play the warning audio and the flash light, if the wheel-skating shoe deviates from the track and the wheel-skating shoe is not in an accelerating state, the wheel-skating shoe plays the warning audio, if the wheel-skating shoe deviates from the track and the wheel-skating shoe is in the accelerating state, the wheel-skating shoe plays the warning audio and flashes.

3. A skatboarding shoe collision flash warning method according to claim 1, wherein the front-facing relationship is that an angle between the upper body of the human body and the center of gravity of the human body is less than 30 degrees, and an angle between the lower body of the human body and the ground is greater than 45 degrees; 4. The roller skate collision flash warning method of claim 1, wherein, the oblique-facing relationship is that the angle between the upper body of the human body and the center of gravity of the human body is greater than 30 degrees, and the angle between the lower body of the human body and the ground is less than 45 degrees. the human body posture data is labeled to obtain labeled human body action data, and the labeled human body action data is used as a data set for training the human body falling down detection model. the human body posture data and the rule data are matched in a time domain to generate a synchronous data stream; 5. A skatboarding shoe collision flash warning method as claimed in claim 1, wherein, a sliding scene feature space is constructed by orthogonal projection based on the synchronous data stream, and the sliding driving scene feature space comprises a plurality of sliding scene feature vectors; 6. A skatboarding shoe collision flash warning method as claimed in claim 1, wherein, an interactive strategy set adapted to a scene is dynamically generated by using an online incremental learning algorithm based on the sliding driving scene feature space; and a warning audio is generated by the wheel-skating shoe based on the interactive strategy set.

7. A skatboarding shoe collision flash warning method according to claim 4, wherein a controller is built in the wheel-skating shoe, the controller runs the human body falling down recognition model, the human body posture data collected by each millimeter-wave radar device is input into the human body falling down recognition model, a plurality of detection results equal to the preset number are obtained, and a final detection conclusion of human body falling down is obtained from the plurality of detection results according to the angle position relationship between the millimeter-wave radar device and the human body. the millimeter-wave radar device can be replaced by an infrared laser emitter or an ultrasonic emitter, and the millimeter-wave radar device synchronously receives reflected optical signals and scattered acoustic signals of the human body. the sliding scene feature vector is input into a pre-trained convolutional pulse neural network, a local time mode is extracted by a convolutional subnetwork, a global space-time correlation is captured by a pulse subnetwork, and the local time mode and the global space-time correlation are fused in combination with attention weight to form a sliding mode classifier.

8. A roller skate impact flash warning system characterized by, The history module is configured to acquire rule data of the roller skate and the race track at a target stage, the history data including rule position information, rule speed information, and rule posture information, the target stage being from a starting point to an end point of the race track. The real-time module is configured to acquire real-time position information, real-time speed information, and real-time posture information of the roller skate and the throw-off body at the target stage. The installation module is configured to install a certain number of millimeter wave radar devices around the detection space, and collect human posture data by using the millimeter wave radar devices. The operation module is configured to determine an angle position relationship between the millimeter wave radar device and the human body, including a front-facing relationship and an oblique-facing relationship, and obtain a detection conclusion of human fall detection from a plurality of detection results according to the angle position relationship between the millimeter wave radar device and the human body. The posture angle difference value of the target stage is calculated according to the real-time posture information, the spatial distance value of the target stage is calculated according to the real-time position information, and the relative speed value of the target stage is calculated according to the real-time speed information. When the posture angle difference value is greater than a threshold value, the roller skate is warned by setting a warning audio and a flash light, if the roller skate does not deviate from the race track, the roller skate does not play the warning audio and the flash light, if the roller skate deviates from the race track and the roller skate is not in an accelerating state, the roller skate plays the warning audio, if the roller skate deviates from the race track and the roller skate is in the accelerating state, the roller skate plays the warning audio and flashes.

9. A computer device, comprising: The memory and the processor and the computer program stored in the memory, when the computer program is executed on the processor, realize the roller skate collision flash warning method as claimed in any one of claims 1 to 7. The computer program is stored on the processor, and when the computer program is executed on the processor, the roller skate collision flash warning method as claimed in any one of claims 1 to 7 is realized.

10. A computer-readable storage medium, characterized in that, ​

Images