Skiing motion behavior recognition method and device based on a single inertial measurement unit

Abstract

The present invention relates to the technical field of behavior recognition, and in particular to a skiing motion behavior recognition method and device based on a single inertial measurement unit. The technical solution provided by the present invention collects inertial data of a skier during skiing through a single inertial measurement unit installed on the skier's leg, and recognizes the skiing motion behavior of the skier based on the acceleration and angular velocity in the three-dimensional space included in the inertial data. In this way, both the recognition of skiing technical actions and the recognition of skiing fall behaviors can be realized. Therefore, the above technical solution can effectively recognize skiing motion behaviors by using a single inertial measurement unit.

Description

Technical Field

[0001] The present invention relates to the technical field of behavior recognition, and in particular, to a skiing motion behavior recognition method and device based on a single inertial measurement unit. Background Art

[0002] Skiing is a high-speed and highly skilled sport, and it is extremely easy to fall or collide with others, thus causing injuries. Therefore, the recognition of motion behaviors such as technical actions and falling behaviors in skiing is particularly important. Currently, the research on skiing motion behavior recognition mostly focuses on image-based recognition. However, due to the high speed, long distance, and large drop of skiing, it is difficult for a camera to capture the behavior of skiers throughout the process, which causes difficulties in the recognition of technical actions and falling behaviors.

[0003] With the development of microelectronics technology, wearable inertial measurement units have become a popular choice for motion behavior recognition. However, the existing skiing motion behavior recognition based on inertial measurement units requires a large number of inertial measurement units, and lacks the recognition of falling behaviors, and cannot meet the recognition requirements of skiing for portability, high precision, and safety.

[0004] Based on this, the present invention proposes a skiing motion behavior recognition method and device based on a single inertial measurement unit to solve the above technical problems. Summary of the Invention

[0005] The present invention describes a skiing motion behavior recognition method and device based on a single inertial measurement unit, which can effectively recognize skiing motion behaviors by using a single inertial measurement unit.

[0006] According to a first aspect, the present invention provides a skiing motion behavior recognition method based on a single inertial measurement unit, including:

[0007] Obtaining inertial data collected by a single inertial measurement unit installed on the leg of a skier; wherein, the inertial data includes acceleration and angular velocity in a three-dimensional space;

[0008] Based on the acceleration and the angular velocity, recognizing the skiing motion behavior of the skier; wherein, the skiing motion behavior includes skiing technical actions and skiing falling behaviors.

[0009] According to a second aspect, the present invention provides a skiing motion behavior recognition device based on a single inertial measurement unit, including:

[0010] An obtaining unit, configured to obtain inertial data collected by a single inertial measurement unit installed on the leg of a skier; wherein, the inertial data includes acceleration and angular velocity in a three-dimensional space;

[0011] An identification unit, configured to identify the skiing motion behavior of the skier based on the acceleration and the angular velocity; wherein, the skiing motion behavior includes skiing technical actions and skiing falling behaviors.

[0012] In a third aspect, an embodiment of the present specification further provides an electronic device, including a memory and a processor. A computer program is stored in the memory, and when the processor executes the computer program, the method described in any embodiment of the present specification is implemented.

[0013] In a fourth aspect, an embodiment of the present specification further provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed on a computer, the computer is made to execute the method described in any embodiment of the present specification.

[0014] According to the skiing motion behavior identification method and device based on a single inertial measurement unit provided by the present invention, inertial data of the skier during skiing is collected by a single inertial measurement unit installed on the skier's leg, and based on the acceleration and angular velocity in the three-dimensional space included in the inertial data, the skiing motion behavior of the skier is identified. In this way, both the identification of skiing technical actions and the identification of skiing falling behaviors can be achieved. Therefore, the above technical solution can effectively identify skiing motion behaviors by using a single inertial measurement unit. Description of the Drawings

[0015] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0016] Figure 1 Shows a schematic flowchart of a skiing motion behavior identification method based on a single inertial measurement unit according to an embodiment;

[0017] Figure 2 Shows a schematic block diagram of a skiing motion behavior identification device based on a single inertial measurement unit according to an embodiment;

[0018] Figure 3 Shows a schematic diagram of the placement position of an inertial measurement unit according to an embodiment. Detailed Embodiments

[0019] The following describes the solution provided by the present invention in conjunction with the drawings.

[0020] Figure 1The flowchart shows a method for identifying skiing behavior based on a single inertial measurement unit according to an embodiment. It can be understood that this method can be executed by any device, equipment, platform, or cluster of devices with computing and processing capabilities. As Figure 1 shown, the method includes:

[0021] Step 100: Obtain inertial data collected by a single inertial measurement unit installed on the skier's leg; wherein, the inertial data includes acceleration and angular velocity in three-dimensional space;

[0022] Step 102: Identify the skiing behavior of the skier based on the acceleration and angular velocity; wherein, the skiing behavior includes skiing technical actions and skiing fall behavior.

[0023] In this embodiment, the inertial data of the skier during skiing is collected by a single inertial measurement unit installed on the skier's leg, and the skiing behavior of the skier is identified based on the acceleration and angular velocity in three-dimensional space included in the inertial data. In this way, both the identification of skiing technical actions and the identification of skiing fall behavior can be realized. Therefore, the above technical solution can effectively identify skiing behavior using a single inertial measurement unit.

[0024] Considering that most of the technical actions in alpine skiing are related to lower limb movements, the inertial data collected from the lower limbs can better reflect the characteristics of different technical actions. Therefore, a single-node inertial measurement unit is installed on the skier's leg (for example, the outer side of the calf, either left or right) to collect the skier's inertial data, which includes acceleration [a x , a y , a z and three-dimensional angular velocity [w x , w y , w z in three-dimensional space. The inertial measurement unit follows the "right-handed Cartesian coordinate system" rule. Taking the installation on the outer side of the right calf as an example, the x-axis points to the right of the body, the y-axis points to the front of the body, and the z-axis points to the right side of the body (see Figure 3 ).

[0025] In an embodiment of the present invention, the step of "identifying the skiing behavior of the skier based on the acceleration and angular velocity" may specifically include:

[0026] Respectively judge the acceleration and angular velocity against a first preset set of thresholds to obtain a first judgment result;

[0027] Based on the first judgment result, determine whether the skier has a skiing fall behavior;

[0028] If it appears, based on the first judgment result, determine the specific behavior of the skiing fall behavior; wherein, the specific behaviors include left - right fall, rotational fall, and front - back fall.

[0029] If it does not appear, then based on the acceleration and angular velocity, identify the skiing technical actions of the skier.

[0030] In this embodiment, due to the complex skiing environment, there are many reasons for falls, and the resulting fall types are diverse. Moreover, due to the relatively high skiing speed, the instantaneous values of acceleration and angular velocity during a fall are far beyond those of other fall scenarios. Therefore, the inventor considered that a threshold discrimination method can be used to determine the fall situation of the skier.

[0031] In an embodiment of the present invention, the first judgment result is determined by the following formula:

[0032]

[0033] In the formula, a xij 、a yij 、a zij respectively represent the acceleration values of the j - th data point in the i - th sample in the x - axis, y - axis, and z - axis directions. The sample is obtained by segmenting the inertial data through a sliding window. a x 、a y 、a z respectively represent the first acceleration thresholds preset in the x - axis, y - axis, and z - axis directions. w xij 、w yij 、w zij respectively represent the angular velocity values of the j - th data point in the i - th sample in the x - axis, y - axis, and z - axis directions. β x 、β y 、β z respectively represent the first angular velocity thresholds preset in the x - axis, y - axis, and z - axis directions.

[0034] If F i = 1, then determine the first judgment result as a left - right fall (i.e., a left - right tilt fall).

[0035] If F i = 2, then determine the first judgment result as a rotational fall.

[0036] If F i = 3, then determine the first judgment result as a front - back fall (i.e., a forward or backward fall).

[0037] If F i = 0, then determine the first judgment result as no skiing fall behavior occurred, and then continue to identify the skiing technical actions of the skier based on the acceleration and angular velocity.

[0038] In this example, falling to the left or right is mostly caused by insufficient pushing force of the skiing leg when the skier makes a turning motion. Spinning falls are mostly caused by collisions between skiers. Falling forward or backward is mostly caused by insufficient core strength of novice skiers. Therefore, based on this consensus, the inventor creatively thought of determining thresholds for acceleration and angular velocity in each direction to identify the above-mentioned falling behaviors.

[0039] In an embodiment of the present invention, after the step of "determining the specific behavior of the skiing fall", the following is further included:

[0040] After a preset duration of determining the specific behavior of the skiing fall, the acceleration and angular velocity are respectively judged against a second preset threshold set to obtain a second judgment result;

[0041] Based on the second judgment result, it is determined to cancel the danger warning or issue a danger alarm.

[0042] In this embodiment, when determining the specific behavior of the skiing fall, the danger warning mode can be triggered. After a preset duration of determining the specific behavior of the skiing fall, it can be further continuously monitored whether the skier can get up by himself. If the skier can get up by himself, the danger warning is cancelled; otherwise, the danger situation alarm is triggered for emergency handling. Such a setting can better ensure the personal safety of the skier.

[0043] In an embodiment of the present invention, the second judgment result is determined by the following formula:

[0044]

[0045] In the formula, γ x , γ y , γ z respectively represent the second acceleration thresholds preset in the x-axis, y-axis, and z-axis directions, and δ x , δ y , δ z respectively represent the second angular velocity thresholds preset in the x-axis, y-axis, and z-axis directions;

[0046] If G i = 1, the second judgment result is determined to cancel the danger warning;

[0047] If G i = 0, the second judgment result is determined to issue a danger alarm.

[0048] In this embodiment, by setting the three-axis acceleration thresholds γ x , γ y , γ z and the three-axis angular velocity thresholds δ x , δy , δ z , it can be determined whether to lift the danger warning or issue a danger alarm.

[0049] In one embodiment of the present invention, α x =0.5g, α y =0.5g, α z =1.5g, β x =2rad / s, β y =2rad / s, β z =2rad / s,γ x x , γ y y , γ z z , δ x <β y , δ y <β z , δ z <β x .

[0050] In this embodiment, considering that the skiing speed is fast, the instantaneous values ​​of acceleration and angular velocity are large when falling, so α is set x =0.5g, α y =0.5g, α z =1.5g, β x =2rad / s, β y =2rad / s, β z =2rad / s; and since the climbing action is not as violent as the falling action, the angular velocity and angular velocity value of the climbing action are both smaller than those of the falling action, that is, γ x x , γ y y , γ z z , δ x <β y , δ y <β z , δ z <β x .

[0051] In one embodiment of the present invention, the step of “identifying the skiing technique of the skier based on the acceleration and angular velocity” may specifically include:

[0052] Perform matrix processing on the acceleration and angular velocity to obtain time series data; the time series data includes M samples, each sample has a dimension of 7×n, and n is the length of the sliding window;

[0053] ​​​​​​Input the time series data into the first extraction model and the second extraction model respectively, and perform feature fusion on the results output by the first extraction model and the second extraction model to obtain fusion data; wherein, the first extraction model is used to sequentially extract the spatial features and temporal features of the time series data, and the second extraction model is used to extract the time domain features of the time series data.

[0054] Input the fusion data into the action recognition model, and output the skiing technical actions of the skier; wherein, the skiing technical actions include parallel slalom, snowplow slalom, parallel skiing, and snowplow skiing.

[0055] In this embodiment, by performing feature fusion on the spatial features, temporal features, and frequency domain features of the extracted time series data, the spatio-temporal features (i.e., spatial features and temporal features) of the input data can be better captured, thereby improving the accuracy of skiing behavior recognition.

[0056] In some embodiments, the first extraction model includes a spatial information extraction module and a temporal information extraction module. Among them, the spatial information extraction module may include multiple convolutional layers, a batch normalization layer, and a max pooling layer. The convolutional layer can well mine the spatial interaction features between data. Each convolutional layer contains multiple convolutional kernels for extracting rich spatial interaction features in the input data; the batch normalization layer is used for data normalization, subtracting the mean and normalizing the variance to accelerate the model convergence speed and improve the model stability; the max pooling layer is used for dimensionality reduction, reducing the computational amount, retaining important features, and enhancing the generalization ability of the model. Since the convolutional network in the spatial information extraction module does not have the temporal memory function, in order to better extract the temporal features contained in the input data, a temporal information extraction module is designed. The temporal information extraction module may include two double-layer gated recurrent units and a flattening layer. In skiing, the data at each moment is affected not only by past data but also by future data. In order to enhance the capture of forward and backward temporal features and improve the recognition accuracy, double-layer gated recurrent units are adopted.

[0057] In some embodiments, among the four alpine skiing actions, the calf rhythm of each action presents different characteristics, which are specifically manifested in the time domain features of acceleration and angular velocity. Therefore, extract the skiing technique-related features in sample X i including seven-dimensional data [a x , a y , a z , a, w x , w y , w zEight characteristic indicators: maximum value, minimum value, average value, peak-to-peak value, variance, standard deviation, root mean square, and range, a total of 56 characteristics, and then a second extraction model is constructed. The second extraction model may include two fully connected layers, and a dropout layer is added after each fully connected layer to prevent overfitting.

[0058] In some embodiments, the action recognition model is responsible for classifying skiing technical actions. The action recognition model may include two fully connected layers, and a dropout layer is added after each layer to avoid overfitting; then, the output layer using the Softmax activation function can calculate the scores of each behavior category and output the activity category with the highest probability score.

[0059] In summary, the present invention has the following advantages: 1) It can use only one inertial measurement unit to identify four alpine skiing techniques, three falling behaviors, and getting-up behaviors. For the monitored technical actions, they can be included in the technical action statistics; for the behavior of not getting up after falling, early warning and alarm can be given to improve skiing safety; 2) A single inertial measurement unit is light, easy to carry, and low in cost, suitable for skiing sports with large-scale movement, does not interfere with the normal movement of skiers, and overcomes the problem of troublesome wearing of multi-node inertial measurement units; 3) This method does not require attitude solution, constructs the first extraction model and the second extraction model, and directly captures the spatio-temporal relationship in the acceleration and angular velocity data in three-dimensional space by using convolutional neural networks and gated recurrent units in deep learning to improve the recognition accuracy, overcoming the problem of low recognition accuracy using a single inertial measurement unit.

[0060] The above describes specific embodiments of the present invention. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims may be performed in a different order than in the embodiments and still achieve the desired results. Additionally, the processes depicted in the drawings do not necessarily require the specific order or sequential order shown to achieve the desired results. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0061] According to an embodiment of another aspect, the present invention provides a skiing motion behavior recognition device based on a single inertial measurement unit. Figure 2 A schematic block diagram showing a skiing motion behavior recognition device based on a single inertial measurement unit according to an embodiment is shown. It can be understood that the device can be implemented by any device, equipment, platform, and device cluster having computing and processing capabilities. As Figure 2 shown, the device includes: an acquisition unit 200 and an identification unit 202. The main functions of each component unit are as follows:

[0062] An acquisition unit 200, configured to acquire inertial data collected by a single inertial measurement unit installed on the leg of a skier; wherein, the inertial data includes acceleration and angular velocity in a three-dimensional space;

[0063] An identification unit 202, configured to identify the skiing motion behavior of the skier based on the acceleration and the angular velocity; wherein, the skiing motion behavior includes skiing technical actions and skiing falling behaviors.

[0064] As a preferred implementation manner, the identification unit is used to perform the following operations:

[0065] Respectively judge the acceleration and the angular velocity against a first preset threshold set to obtain a first judgment result;

[0066] Based on the first judgment result, determine whether the skier has a skiing falling behavior;

[0067] If so, based on the first judgment result, determine the specific behavior of the skiing falling behavior; wherein, the specific behavior includes left and right falls, rotational falls, and front and back falls;

[0068] If not, identify the skiing technical actions of the skier based on the acceleration and the angular velocity.

[0069] As a preferred implementation manner, the first judgment result is determined by the following formula:

[0070]

[0071] In the formula, a xij 、a yij 、a zij respectively represent the acceleration values of the j-th data point in the i-th sample in the x-axis, y-axis, and z-axis directions, the sample is obtained by segmenting the inertial data through a sliding window, a x 、a y 、a z respectively represent the first acceleration thresholds preset in the x-axis, y-axis, and z-axis directions, w xij 、w yij 、w zij respectively represent the angular velocity values of the j-th data point in the i-th sample in the x-axis, y-axis, and z-axis directions, β x 、β y 、β z respectively represent the first angular velocity thresholds preset in the x-axis, y-axis, and z-axis directions;

[0072] If F iIf F

[0073] If F i If F

[0074] If F i If F

[0075] If F i If F

[0076] As a preferred embodiment, it further includes:

[0077] A judgment unit, configured to judge the acceleration and the angular velocity respectively against a second preset threshold set after a preset duration for determining the specific behavior of the skiing fall behavior, to obtain a second judgment result; based on the second judgment result, determine to cancel the danger warning or issue a danger alarm.

[0078] As a preferred embodiment, the second judgment result is determined by the following formula:

[0079]

[0080] In the formula, γ x 、γ y 、γ z respectively represent second acceleration thresholds preset in the x-axis, y-axis, and z-axis directions, and δ x 、δ y 、δ z respectively represent second angular velocity thresholds preset in the x-axis, y-axis, and z-axis directions;

[0081] If G i If G

[0082] If G i If G

[0083] As a preferred embodiment, α x = 0.5g, α y = 0.5g, α z = 1.5g, β x = 2rad / s, β y = 2rad / s, β z = 2rad / s, γ x <a x ,γ y <a y, γ z <a z , δ x <δ y , δ y <β z , δ z <β x 。

[0084] As a preferred embodiment, when performing the recognition of the skiing technical actions of the skier based on the acceleration and the angular velocity, the recognition unit is configured to perform the following operations:

[0085] Perform matrix processing on the acceleration and the angular velocity to obtain time series data; wherein, the time series data includes M samples, and the dimension of each sample is 7×n, where n is the sliding window length;

[0086] Input the time series data into a first extraction model and a second extraction model respectively, and perform feature fusion on the results output by the first extraction model and the second extraction model to obtain fusion data; wherein, the first extraction model is used to sequentially extract the spatial features and the time features of the time series data, and the second extraction model is used to extract the time domain features of the time series data;

[0087] Input the fusion data into an action recognition model, and output the skiing technical actions of the skier; wherein, the skiing technical actions include parallel turning, snowplow turning, parallel skiing, and snowplow skiing.

[0088] According to an embodiment of another aspect, there is also provided a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed in a computer, the computer is made to execute the method described in combination with Figure 1 。

[0089] According to an embodiment of still another aspect, there is also provided an electronic device, including a memory and a processor, where an executable code is stored in the memory, and when the processor executes the executable code, the method described in combination with Figure 1 is implemented.

[0090] The embodiments in the present invention are all described in a progressive manner. For the same or similar parts among the embodiments, reference can be made to each other. Each embodiment focuses on the differences from other embodiments. In particular, for the device embodiments, since they are basically similar to the method embodiments, the description is relatively simple, and for the relevant parts, reference can be made to the partial description of the method embodiments.

[0091] Those skilled in the art should be able to realize that in one or more of the above examples, the functions described in the present invention can be implemented by hardware, software, firmware, or any combination thereof. When implemented using software, these functions can be stored in a computer-readable medium or transmitted as one or more instructions or codes on a computer-readable medium.

[0092] The specific embodiments described above have further elaborated on the purpose, technical solutions, and beneficial effects of the present invention. It should be understood that the above description is only for the specific embodiments of the present invention and is not used to limit the protection scope of the present invention. Any modifications, equivalent replacements, improvements, etc. made on the basis of the technical solutions of the present invention should be included within the protection scope of the present invention.

Claims

1. A skiing motion behavior recognition method based on a single inertial measurement unit, characterized in that: include: Acquiring inertial data collected by a single inertial measurement unit mounted on the skier's leg; wherein the inertial data includes acceleration and angular velocity in three-dimensional space; Based on the acceleration and the angular velocity, the skiing motion behavior of the skier is identified; wherein the skiing motion behavior includes skiing technical movements and skiing falls; The identifying the skiing motion behavior of the skier based on the acceleration and the angular velocity includes: respectively judging the acceleration and the angular velocity against a first preset threshold set to obtain a first judgment result; Based on the first judgment result, determining whether the skier has fallen while skiing; If it occurs, then based on the first judgment result, the specific behavior of the skiing fall behavior is determined; wherein the specific behavior includes left-right fall, rotation fall and front-back fall; If not, identifying the skiing technique of the skier based on the acceleration and the angular velocity; The first judgment result is determined by the following formula: In the formula, a xij 、a yij 、a zij represents the acceleration value of the jth data point in the ith sample in the x-axis, y-axis, and z-axis directions, respectively. The sample is obtained by segmenting the inertial data through a sliding window. x 、a y 、a z represents the first acceleration thresholds preset in the x-axis, y-axis, and z-axis directions, respectively, xij 、w yij 、w zij Respectively represent the angular velocity value of the jth data point in the ith sample in the x-axis, y-axis, and z-axis directions, β x , β y , β z Respectively represent the first angular velocity thresholds preset in the x-axis, y-axis, and z-axis directions; If F i =1, the first judgment result is determined as a left or right fall; If F i =2, the first judgment result is determined as a rotation fall; If F i =3, the first judgment result is determined as a front-to-back fall; If F i =0, the first judgment result is determined as no skiing fall occurs.

2. The method according to claim 1, characterized in that After determining the specific behavior of the skiing fall behavior, the method further includes: After determining the preset duration of the specific behavior of the skiing fall behavior, respectively judging the acceleration and the angular velocity with a second preset threshold set to obtain a second judgment result; Based on the second judgment result, it is determined whether to cancel the danger warning or issue a danger alarm.

3. The method according to claim 2, characterized in that The second judgment result is determined by the following formula: In the formula, γ x , γ y , γ z Respectively represent the second acceleration thresholds preset in the x-axis, y-axis, and z-axis directions, δ x , δ y , δ z Respectively represent the second angular velocity thresholds preset in the x-axis, y-axis, and z-axis directions; If G i =1, the second judgment result is determined as the danger warning being lifted; If G i =0, the second judgment result is determined as issuing a danger alarm.

4. The method according to claim 3, characterized in that a x =0.5g,a y =0.5g,a z =1.5g,β x =2rad / s, β y =2rad / s, β z =2rad / s,γ x x ,c y y ,c z z ,b x <b y ,d y <b z ,d z <b x 。​​​ 5. The method according to any one of claims 1 to 4, characterized in that The identifying the skiing technical action of the skier based on the acceleration and the angular velocity includes: Performing matrix processing on the acceleration and the angular velocity to obtain time series data; wherein the time series data includes M samples, each of which has a dimension of 7×n, where n is a sliding window length; The time series data are input into a first extraction model and a second extraction model respectively, and the results output by the first extraction model and the second extraction model are subjected to feature fusion to obtain fused data; wherein the first extraction model is used to sequentially extract the spatial features and the temporal features of the time series data, and the second extraction model is used to extract the temporal features of the time series data; The fusion data is input into a motion recognition model, and the skiing technical movements of the skier are output; wherein the skiing technical movements include parallel slalom, pear turn, parallel downhill and pear downhill.

6. A skiing motion behavior recognition device based on a single inertial measurement unit, characterized in that: include: An acquisition unit configured to acquire inertial data collected by a single inertial measurement unit installed on the skier's leg; wherein the inertial data includes acceleration and angular velocity in three-dimensional space; an identification unit configured to identify the skiing motion behavior of the skier based on the acceleration and the angular velocity; wherein the skiing motion behavior includes skiing technical movements and skiing falls; The identification unit is used to perform the following operations: respectively judging the acceleration and the angular velocity against a first preset threshold set to obtain a first judgment result; Based on the first judgment result, determining whether the skier has fallen while skiing; If it occurs, then based on the first judgment result, the specific behavior of the skiing fall behavior is determined; wherein the specific behavior includes left-right fall, rotation fall and front-back fall; If not, identifying the skiing technique of the skier based on the acceleration and the angular velocity; The first judgment result is determined by the following formula: In the formula, a xij 、a yij 、a zij represents the acceleration value of the jth data point in the ith sample in the x-axis, y-axis, and z-axis directions, respectively. The sample is obtained by segmenting the inertial data through a sliding window. x 、a y 、a z represents the first acceleration thresholds preset in the x-axis, y-axis, and z-axis directions, respectively, xij 、w yij 、w zij Respectively represent the angular velocity value of the jth data point in the ith sample in the x-axis, y-axis, and z-axis directions, β x , β y , β z Respectively represent the first angular velocity thresholds preset in the x-axis, y-axis, and z-axis directions; If F i =1, the first judgment result is determined as a left or right fall; If F i =2, the first judgment result is determined as a rotation fall; If F i =3, the first judgment result is determined as a front-to-back fall; If F i =0, the first judgment result is determined as no skiing fall occurs.

7. An electronic device, characterized in that: The method comprises a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the method according to any one of claims 1 to 5 is implemented.

8. A computer-readable storage medium, characterized in that: A computer program is stored thereon, and when the computer program is executed in a computer, the computer is caused to execute the method according to any one of claims 1 to 5.

Images