Dragon boat paddle frequency and force balance testing device and method under usage scenarios

Abstract

The present invention relates to a device and method for testing dragon boat paddle frequency and force balance in a usage scenario, wherein the testing device includes two parts, a water area and an onshore area. The water area includes multiple sets of grip strength detection devices, multiple sets of soundproof earphones, a Beidou satellite positioning module set on the dragon head of each dragon boat, a dragon ball set at the dragon mouth, and a fireball set at the dragon tail, which serve as markers. The onshore area includes cameras set on the shore at both ends of the waterway where each dragon boat is located, and a PC set on the shore on the left side in the standard vertical direction as a background server. The whole process of dragon boat rowing is coordinated, the paddler's paddling action is dynamically adjusted, and real-time synchronous movement using the real scene of the dragon boat is achieved.

Description

Technical Field

[0001] The present invention relates to a device and method for testing paddle frequency and force balance, and in particular to a dragon boat paddle frequency and force balance test under usage scenarios, belonging to the field of sports equipment testing. Background Art

[0002] Paddle rate and power balance are crucial indicators in dragon boat racing, crucial for performance and enabling timely review to identify areas for training effort. Current dragon boat training for these two indicators relies on simulators, with sensors installed on the pool walls. However, actual performance in real-world competitions is affected by factors such as on-site performance, athlete psychology, the quality of regular training, prior training, and weather, making it impossible to linearly compare these indicators to regular simulator training. Simulator training, in other words, cannot accurately reflect the conditions experienced during dragon boat training, especially during competitions. Furthermore, the movement of dragon boats precludes the cost-effective deployment of sensors over long distances. Furthermore, the vast expanse of water at the competition site makes it impossible to accurately calculate the wave noise and energy attenuation caused by paddling, without requiring surface noise.

[0003] In addition, in order to adapt to the competition scene, training on a "real boat" is the best solution. Therefore, how to set up the dragon boat paddle frequency and power balance test equipment on the real boat, as well as the test method, is particularly important. Summary of the Invention

[0004] Due to the above-mentioned unresolved problems in the prior art, the present invention adopts the following solutions in its design: first, a grip force detection device suitable for paddling is provided; second, camera devices are set at both ends of each track to record the time-varying states of grip force and boat body images, which correspond to the corresponding paddling frequency, synchronization, and relationship with the force balance result; third, soundproof earphones are provided for the paddler to prompt each paddler whether they need to pay attention to the speed of the paddling frequency based on the prediction of the force balance result.

[0005] Based on the above design scheme, the present invention provides a dragon boat paddle frequency and force balance test device under the use scenario, including a grip detection device consistent with the number of dragon boat paddles, used to monitor the pressure generated by the fingers gripping the paddle surface over time, so as to calculate the paddle frequency of each paddler, a camera device set at both ends of the track, and a marker set at the head and tail of the dragon boat, used to calculate the yaw degree of the boat body according to the video or multiple images taken by the camera device, and to construct a model to predict the force balance result through the grip history data of each paddler and the calculation result of the yaw degree, as well as a soundproof headset worn by the paddler, a satellite positioning module set on the dragon head, and a background server on the shore of the water area, wherein,

[0006] The grip detection device, camera device, sound-isolating earphone, and satellite positioning module all communicate with the background server, upload and save the change function, video or multiple images in the background server, perform the calculation and construct the model to predict the force balance result in the background server, and send a prompt message to the sound-isolating earphone worn by the rower.

[0007] Optionally, the paddle shaft has a grip force detection portion for installing a grip force detection device, and the grip force detection device includes a flexible substrate, one side of which is provided with a pressure sensor sheet, and the other side is provided with an acquisition amplifier electrically connected to the pressure sensor sheet, for collecting and amplifying the pressure signal of the grip force, and sending it to the backend server through the antenna hole provided with the antenna on the paddle shaft via a wireless transmission module electrically connected thereto; the method for calculating the paddling frequency of each rower is to find the time interval between two adjacent peaks on the variation function , then the propeller frequency .

[0008] It should be understood that the paddling action determines that the fingers corresponding to the main rowing arm need to exert pressure on the surface of the paddle shaft when paddling to overcome the pressure of the water on the paddle blade. Then the paddle blade leaves the water surface and the palm pushes the paddle shaft forward. At this time, the pressure of the fingers corresponding to the main rowing arm on the paddle shaft is reduced, thus forming a periodic law in the time change function, thereby calculating the paddling frequency.

[0009] Preferably, all the soundproofing headsets can be controlled by a background server to play the chant synchronously to coordinate the paddling rhythm of all rowers.

[0010] Optionally, the method for calculating the yaw degree includes:

[0011] S1, the backend server receives the real-time position of the dragon head from the satellite positioning module, and obtains the real-time distance between the camera device and the dragon head according to the position of the camera device set in front of the dragon head outside the water area. , and according to the distance between it and the camera device set behind the dragon's tail outside the water area , and the distance between the dragon head and tail markers , calculate the distance between the marker set at the dragon's tail and the camera device set behind the dragon's tail outside the water area as ;

[0012] S2 obtains the current video frame or the captured image, and respectively takes the camera distance of the camera device set in front of the dragon head and behind the dragon tail outside the water area. and And the horizontal distance of the crosshair in the field of view from the markers set at the dragon's head and tail and , calculate the horizontal distance of the dragon head and tail from the standard direction and , where the direction of the crosshairs projected on the connecting line of the sensor surfaces of the photographic devices and moving toward the boat is defined as the standard direction;

[0013] S3 yaw degree .

[0014] Preferably, ,but .

[0015] Optionally, the method of constructing a model to predict force balance results based on the grip force history data of each rower and the calculation result of the yaw degree specifically includes:

[0016] Q1 builds a long-short-term memory network. For each rower, the grip force history data is proportionally divided into a training set and a validation set. The training set is sequentially input into each node unit of the network constructed in chronological order. The output terminal outputs the current unit's paddling frequency, which is compared with the true value to obtain the corresponding loss function. Each node unit corresponds to at least one paddling action. The first node unit also inputs the initial signal (usually the background signal of the pressure sensor) at the transmission layer.

[0017] In Q2, the validation set is used to verify the accuracy, and the loss function is used to optimize the network parameters. This training and validation cycle is repeated until the loss function is minimized, the accuracy stabilizes, and the training is completed.

[0018] Q3 obtains the yaw degree data at the current moment, inputs the grip strength detection data of all rowers at the current moment into the node unit of the training network corresponding to the rower, and outputs the predicted paddling frequency , Number the rowers when , , then for the person paddling forward to the left, the predicted paddling frequency is , For the number The paddler's current real paddling frequency (i.e., the time interval between two consecutive paddling processes at the current and previous moments) is obtained, and the backend server sends a reminder message to the paddler's soundproof headset to pay attention to increasing the paddling frequency appropriately. Then send a reminder message to reduce the paddling frequency appropriately. No prompt message will be sent; when Similarly, for those paddling forward to the right, a corresponding prompt message to reduce or increase the paddling frequency will be sent, or no prompt message will be sent;

[0019] Q4 After the preset time, new yaw degree data is obtained again. If , then the predicted result of force balance is given, otherwise continue to execute Q3-Q4 until it meets .

[0020] It is easy to understand that the yaw of the boat to one side is caused by the presence of a rower on the other side of the boat with a single rowing frequency that is too fast, a single rowing frequency that is too slow, or a combination of both, especially the rower near the dragon's head is too fast and the rower near the dragon's tail is too slow. Therefore, it is necessary to determine the relationship between the different predicted rowing frequencies and the measured rowing frequency to issue a prompt message of whether to increase or decrease the frequency. The prediction of force balance is based on satisfying Under the premise of predicting that the next moment is still likely to remain satisfied Since it takes time to adjust the propeller frequency and correct the yaw of the boat, the force is balanced in the short term.

[0021] Preferably, once , then the paddler will be prompted to maintain the paddling frequency through the soundproof headset. If the paddler still cannot meet the requirements after two consecutive rounds of Q3, If the conditions are right, the volume of the horn is increased to coordinate the paddling frequency of all rowers.

[0022] Preferably, the preset time is 3-10s.

[0023] Optionally, the backend server includes at least one of a PC, a portable computer, and a tablet computer.

[0024] Another aspect of the present invention provides a method for testing dragon boat paddle frequency and force balance in a usage scenario. The method uses the aforementioned testing device and specifically includes the following steps:

[0025] P1 sets up multiple channels according to the size of the water area, and installs cameras on the shore at both ends of each channel to return each dragon boat to its starting position. The backend server obtains the position information of the satellite positioning module on each dragon head;

[0026] P2 All rowers wear soundproof headsets and hear the start signal from the backend server. Each dragon boat starts moving forward. The backend server immediately obtains the location information of the satellite positioning module, the video frames and images captured by the camera device on the corresponding channel shore, and the grip strength test data of all rowers. It then calculates the paddling frequency based on the grip strength test data.

[0027] P3 calculates the yaw degree of the boat based on the captured video frames and images, and predicts the force balance result through the grip strength detection data of each rower and the calculated yaw degree, and sends a prompt message to the noise-isolating headset worn by the rower accordingly.

[0028] Beneficial effects

[0029] By applying pressure to the oar shaft with the gripping hand, the photography device on the shore captures the dragon head and tail markers deviating from the center of vision, and builds a model to predict the paddling frequency and force balance results. The backend server is used to coordinate the entire dragon boat rowing process. At the same time, a soundproof microphone is worn to prompt the paddler's paddling frequency based on the force balance results, solving the problem of dynamically adjusting the paddler's paddling action, and playing calls to coordinate the paddler at the same time, realizing real-time synchronous movement using the real scene of the dragon boat. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 A schematic diagram of the dragon boat paddle frequency and force balance test device configured in the water under the usage scenario of Example 1 of the present invention, wherein a schematic diagram of the paddler holding the paddle and paddling in the dragon boat wearing a soundproofing headset is provided.

[0031] Figure 2 for Figure 1 The position of the grip force detection part marked by the dotted circle in the rowing process, and the schematic diagram of the grip force detection device structure,

[0032] Figure 3 Schematic diagram of the force analysis of the holding hand of the main stroke arm.

[0033] Figure 4 A schematic diagram of the onshore portion of the dragon boat paddle frequency and force balance test device configured in the use scenario of Example 1 of the present invention, which provides a schematic diagram of the definitions of various distances, and a schematic diagram of the field of view of the camera device installed on the shore corresponding to the head and tail, with a crosshair.

[0034] Figure 5 Schematic diagram of grip force as a function of time, showing the period, i.e. the time interval between two peaks. ,

[0035] Figure 6 Schematic diagram of the process of using grip force history data to train a long short-term memory network to obtain predicted paddling frequency in Example 2 of the present invention. DETAILED DESCRIPTION

[0036] Example 1

[0037] like Figure 1-Figure 4 As shown, this embodiment will describe the dragon boat paddle frequency and force balance test device in the use scenario, which is divided into the water part and the shore part. Specifically, there are 8 sets of grip strength detection devices in the water ( Figure 2 ), 8 sets of soundproofing headsets ( Figure 1 The picture shows one of them as an example), the Beidou satellite positioning module ( Figure 1 Not shown), the dragon ball at the dragon's mouth, i.e., marker A, and the fireball at the dragon's tail, i.e., marker B (see Figure 1 );

[0038] The shore part includes the camera devices installed on the shore at both ends of the channel where each dragon boat is located ( Figure 4 , a total of two, specifically you can choose a high-definition camera with a long lens, with video, image acquisition and wireless transmitter), the standard vertical direction of the left side of the shore set up the PC as the background server ( Figure 4 ). In this embodiment, the camera master distances of the two high-definition cameras are the same. .

[0039] The PC computer communicates with the grip strength detection device, the sound-isolating headset, the Beidou satellite positioning module, and the camera device.

[0040] like Figure 1 As shown, a rower is shown as an example, and a dotted circle is shown for the hand corresponding to the main stroke arm. Figure 2 The grip force detection unit is shown in the figure, which is used to install the grip force detection device. The grip force detection device is shown in the enlarged diagram of the grip force detection unit, which is guided by two dotted lines. Specifically, it includes a flexible substrate (such as silicone) with a pressure sensing sheet on one side and an acquisition amplifier electrically connected to the pressure sensing sheet on the other side. The amplifier is used to collect and amplify the grip force pressure signal and transmit it to the antenna hole on the paddle shaft through the wireless transmission module electrically connected to the antenna hole (see Figure 2 ) is sent to the backend server; the method for calculating the paddling frequency of each paddler is to find the time interval between two adjacent peaks on the variation function , then the propeller frequency .

[0041] like Figure 3 , when the hand holding the oar corresponding to the main rowing arm holds the oar shaft, it moves in the direction of the stroke, Figure 2 When the paddle blade is subjected to water resistance, the four fingers generate a resistive pressure perpendicular to the surface force direction shown in the figure on the pressure sensor sheet on the surface of the grip force detection part. When the paddle blade is out of the water, the gripping hand pushes forward, and the pressure generated at this time is reduced. Figure 5 The period shown The paddle frequency is calculated from the time interval between the signal intensity peaks of the two slides.

[0042] Among them, such as Figure 4 As shown, the distance between the camera lens on the shore near the dragon's head and the marker A is expressed as , the distance between marker A and marker B is The distance between the two cameras at both ends of the dragon head and tail is A crosshair is set in the lens field of view. As shown in the figure, the markers A and B in the first and last lens fields of view have a lateral deviation from the crosshair (that is, both are projected to the right of the standard on the sensing surface) respectively. and .

[0043] The test device can calculate the degree of yaw of the boat based on multiple images taken by the camera device, and build a model to predict the force balance result through the grip force history data of each rower and the calculation results of the yaw degree.

[0044] Example 2

[0045] This example illustrates the calculation and prediction model construction in Example 1.

[0046] In detail, the method for calculating the yaw degree includes:

[0047] S1, the PC on the shore uses a server-level host to receive the real-time position of the dragon head from the Beidou satellite positioning module, and obtains the real-time distance from the dragon head based on the position of the camera device set in front of the dragon head outside the water area. , and according to the distance between it and the camera device set behind the dragon's tail outside the water area , and the distance between the dragon head and tail markers , calculate the distance between the marker set at the dragon's tail and the camera device set behind the dragon's tail outside the water area as ;

[0048] S2 obtains the currently captured image, and respectively captures the image captured by the camera device in front of the dragon head and behind the dragon tail outside the water area. , and the horizontal distance of the crosshair in the field of view from the markers set at the dragon's head and tail and ( Figure 4 ), calculate the horizontal distance of the dragon head and tail from the standard direction and , where the direction of the crosshairs projected on the connecting line of the sensor surfaces of the photographic devices and moving toward the boat is defined as the standard direction;

[0049] S3 yaw degree .

[0050] The grip force history data of each rower and the calculated results of the yaw degree are used to build a model to predict the force balance results. Specifically,

[0051] Q1 builds a long short-term memory network, such as Figure 6As shown in the figure, for each rower, the grip force history data is divided into a training set and a validation set in proportion. The training set is sequentially input into each node unit of the network constructed in chronological order. The output end outputs the paddling frequency of the current unit and compares it with the true value to obtain the corresponding loss function. Each node unit corresponds to 4 paddling actions (two cycles). The first node unit also inputs the initial signal at the transmission layer.

[0052] It's easy to understand that signal strength is a function of time. By taking the inverse function and finding the reciprocal of the time interval between adjacent maximum signal strength values, the paddling frequency can be calculated. Because a paddler's physical strength over a period of time follows a certain pattern, the maximum signal strength fluctuates after different strokes. By finding this pattern, we can obtain the characteristic strength, and thus the corresponding paddling frequency can also be predicted. Based on these two principles, through training with a large number of training sets, the current paddling frequency can be predicted from the adjacent maximum signal.

[0053] In Q2, the validation set is used to verify the accuracy, and the loss function is used to optimize the network parameters. This training and validation cycle is repeated until the loss function is minimized, the accuracy stabilizes, and the training is completed.

[0054] Q3 obtains the yaw degree data at the current moment, inputs the grip strength detection data of all rowers at the current moment into the node unit of the training network corresponding to the rower, and outputs the predicted paddling frequency , Number the rowers when , here is , then for the person paddling forward to the left, the predicted paddling frequency is , For the number The paddler's current real paddling frequency is sent by the backend server to his / her noise-isolating headset, prompting him / her to increase the paddling frequency appropriately. Then send a reminder message to reduce the paddling frequency appropriately. No prompt message will be sent; when Similarly, for those paddling forward to the right, a corresponding prompt message to reduce or increase the paddling frequency will be sent, or no prompt message will be sent;

[0055] Q4 After another 5 seconds, reacquire the new yaw degree data. If , then the predicted result of force balance is given, otherwise continue to execute Q3-Q4 until it meets .

[0056] Moreover, once satisfied , then the paddler will be prompted to maintain the paddling frequency through the soundproof headset. If the paddler still cannot meet the requirements after two consecutive rounds of Q3, If the conditions are right, the volume of the horn is increased to coordinate the paddling frequency of all rowers.

[0057] Example 3

[0058] This embodiment will use the test device of embodiment 1 that adopts the algorithm and model of embodiment 2 to illustrate a method for testing propeller frequency and force balance, which specifically includes the following steps:

[0059] P1 According to Figure 4 The water area is shown in the figure, and multiple channels are set up. Cameras are installed on the shore at both ends of each channel to return each dragon boat to its starting position. The backend server obtains the position information of the Beidou satellite positioning module on each dragon head.

[0060] P2 All rowers wear noise-isolating headsets (e.g. Figure 1 shown), hear Figure 4 The PC on the shore, acting as the backend server, sends the starting command, and each dragon boat starts moving forward. The backend server immediately obtains the location information of the Beidou satellite positioning module, the captured images of the camera device on the shore of the corresponding channel, and the grip strength test data of all rowers, and calculates the paddling frequency based on the grip strength test data;

[0061] P3 calculates the yaw degree of the boat based on the collected images, and predicts the force balance result through the grip strength detection data of each rower and the calculated yaw degree, and sends a prompt message to the noise-isolating headset worn by the rower accordingly.

[0062] Therefore, according to the method of the present invention, rowers can get a real experience consistent with the on-site conditions during training and competition, thereby creating more realistic training conditions for the competition and quickly adapting, laying the foundation for achieving good results.

Claims

1. The dragon boat paddle frequency and force balance test device under the use scenario is characterized by: It includes a grip detection device that is consistent with the number of dragon boat paddles, which is used to monitor the pressure generated by the fingers gripping the paddle surface over time, so as to calculate the paddling frequency of each paddler, a camera device set at both ends of the track, and a marker set at the bow and stern of the dragon boat, which is used to calculate the yaw degree of the boat body according to the video or multiple images taken by the camera device, and to construct a model to predict the force balance result through the grip strength history data of each paddler and the calculation result of the yaw degree, as well as a soundproof headset worn by the paddler, a satellite positioning module set on the dragon head, and a background server on the shore of the water area, wherein, The grip force detection device, camera device, soundproof headset, and satellite positioning module all communicate with the backend server, upload and save the change function, video or multiple images to the backend server, perform the calculation and build a model to predict the force balance result in the backend server, and send a prompt message to the soundproof headset worn by the rower; The method for calculating the yaw degree includes: S1: The backend server receives the position of the dragon head from the satellite positioning module in real time, and obtains the real-time distance between the camera device and the dragon head according to the position of the camera device set in front of the dragon head outside the water area. , and according to the distance between it and the camera device set behind the dragon's tail outside the water area , and the distance between the dragon head and tail markers , calculate the distance between the marker set at the dragon's tail and the camera device set behind the dragon's tail outside the water area as ; S2: Get the current video frame or the captured image, and then use the camera distance of the camera device set in front of the dragon head and behind the dragon tail outside the water area to obtain the current video frame or the captured image. and And the horizontal distance of the crosshair in the field of view from the markers set at the dragon's head and tail and , calculate the horizontal distance of the dragon head and tail from the standard direction and , where the direction of the crosshairs projected on the connecting line of the sensor surfaces of the photographic devices and moving toward the boat is defined as the standard direction; S3: Yaw degree .

2. The testing device according to claim 1, characterized in that The paddle shaft is provided with a grip force detection portion for mounting a grip force detection device, the grip force detection device comprising a flexible substrate, one side of which is provided with a pressure sensing sheet, and the other side of which is provided with an acquisition amplifier electrically connected to the pressure sensing sheet, for collecting and amplifying the grip force pressure signal, and transmitting the signal to a backend server via an antenna hole provided with an antenna on the paddle shaft via a wireless transmission module electrically connected thereto; the method for calculating the paddling frequency of each rower is to find the time interval between two adjacent peaks on the variation function , then the propeller frequency .

3. The testing device according to claim 1, wherein: All of the soundproof headsets can be controlled by a background server to play the chant synchronously to coordinate the paddling rhythm of all rowers.

4. The testing device according to claim 1, wherein: ,but .

5. The testing device according to claim 4, characterized in that: The method of constructing a model to predict force balance results by using the grip force history data of each rower and the calculated results of the yaw degree includes: Q1 builds a long-short-term memory network. For each rower, the grip force history data is divided into a training set and a validation set in proportion. The training set is sequentially input into each node unit of the network constructed in chronological order. The output terminal outputs the current unit's paddling frequency, which is compared with the true value to obtain the corresponding loss function. Each node unit corresponds to at least one paddling action. The first node unit also inputs the initial signal at the transmission layer. In Q2, the validation set is used to verify the accuracy, and the loss function is used to optimize the network parameters. This training and validation cycle is repeated until the loss function is minimized, the accuracy stabilizes, and the training is completed. Q3 obtains the yaw degree data at the current moment, inputs the grip strength detection data of all rowers at the current moment into the node unit of the training network corresponding to the rower, and outputs the predicted paddling frequency , Number the rowers when , , then for the person paddling forward to the left, the predicted paddling frequency is , For the number The paddler's current real paddling frequency is sent by the backend server to the paddler's soundproof headset, reminding him to increase the paddling frequency appropriately. Then send a reminder message to reduce the paddling frequency appropriately. No prompt message will be sent; when Similarly, for those paddling forward to the right, a corresponding prompt message to reduce or increase the paddling frequency will be sent, or no prompt message will be sent; Q4 After the preset time, new yaw degree data is obtained again. If , then the predicted result of force balance is given, otherwise continue to execute Q3-Q4 until it meets .

6. The testing device according to claim 5, characterized in that: Once satisfied , then the paddler will be prompted to maintain the paddling frequency through the soundproof headset. If the paddler still cannot meet the requirements after two consecutive rounds of Q3, If the conditions are right, the volume of the horn is increased to coordinate the paddling frequency of all rowers.

7. The testing device according to claim 5 or 6, characterized in that: The preset time is 3-10s.

8. The testing device according to claim 7, characterized in that: The backend server includes at least one of a PC, a portable computer, and a tablet computer.

9. A method for testing dragon boat paddle frequency and force balance in a usage scenario, characterized in that: The method uses the testing device according to any one of claims 1 to 8, and specifically comprises the following steps: P1 sets up multiple channels according to the size of the water area, and installs cameras on the shore at both ends of each channel to return each dragon boat to its starting position. The backend server obtains the position information of the satellite positioning module on each dragon head; P2 All rowers wear soundproof headsets and hear the start signal from the backend server. Each dragon boat starts moving forward. The backend server immediately obtains the location information of the satellite positioning module, the video frames and images captured by the camera device on the corresponding channel shore, and the grip strength test data of all rowers. It then calculates the paddling frequency based on the grip strength test data. P3 calculates the yaw degree of the boat based on the captured video frames and images, and predicts the force balance result through the grip strength detection data of each rower and the calculated yaw degree, and sends a prompt message to the noise-isolating headset worn by the rower accordingly.

Images