WO2017185222A1 - 一种基于球类运动的运动轨迹采集和分析的***与方法 - Google Patents

一种基于球类运动的运动轨迹采集和分析的***与方法 Download PDF

Info

Publication number
WO2017185222A1
WO2017185222A1 PCT/CN2016/080191 CN2016080191W WO2017185222A1 WO 2017185222 A1 WO2017185222 A1 WO 2017185222A1 CN 2016080191 W CN2016080191 W CN 2016080191W WO 2017185222 A1 WO2017185222 A1 WO 2017185222A1
Authority
WO
WIPO (PCT)
Prior art keywords
data
analysis
acquisition
motion
sensor
Prior art date
Application number
PCT/CN2016/080191
Other languages
English (en)
French (fr)
Inventor
黄宇
胡振江
李广贵
Original Assignee
深圳市优宝创科技有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 深圳市优宝创科技有限公司 filed Critical 深圳市优宝创科技有限公司
Priority to PCT/CN2016/080191 priority Critical patent/WO2017185222A1/zh
Priority to CN201680011563.8A priority patent/CN107454970A/zh
Publication of WO2017185222A1 publication Critical patent/WO2017185222A1/zh

Links

Images

Classifications

    • A—HUMAN NECESSITIES
    • A63—SPORTS; GAMES; AMUSEMENTS
    • A63B—APPARATUS FOR PHYSICAL TRAINING, GYMNASTICS, SWIMMING, CLIMBING, OR FENCING; BALL GAMES; TRAINING EQUIPMENT
    • A63B71/00—Games or sports accessories not covered in groups A63B1/00 - A63B69/00
    • A63B71/06—Indicating or scoring devices for games or players, or for other sports activities
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/20—Analysis of motion
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/20—Analysis of motion
    • G06T7/292—Multi-camera tracking

Definitions

  • the invention relates to the technical field of ball motion track monitoring, in particular to a system and method for collecting and analyzing motion track based on ball motion.
  • the existing motion recognition systems are composed of an acquisition module, a communication module and a calculation module.
  • the acquisition module mainly collects and preliminary analyzes data through sensors including acceleration sensors, gyroscopes, magnetometers, and motion and orientation-related sensors, and then transmits the data to the host computer equipment through mobile communication devices, including mobile phones, tablet computers, and the like.
  • the computer performs the final motion analysis. Since the algorithm of motion recognition is based on one or more integrals and spatial transformations of sensor data, the higher the sampling rate, the higher the final motion recognition rate and the higher the trajectory reduction degree. For some ball games with very high speeds, such as badminton, table tennis, baseball, softball, cricket, and football with pitching, the movement rate is very fast, and the sampling rate is not high enough.
  • the present invention provides a ball based motion.
  • Method of motion trajectory acquisition and analysis By using the method provided by the invention, the limitation of communication bandwidth and storage capacity can be solved, and the sampling speed of the sensor can be maximized to achieve the maximum sampling precision.
  • the entire motion process can be analyzed without gap tracking without losing the small motion data. Being able to work independently without relying on the host device greatly simplifies the system architecture, compresses system costs, and simultaneously improves the user experience.
  • a motion trajectory acquisition and analysis system based on ball motion the system comprises an integrated module for data acquisition and analysis, a communication module and a human-machine interface display module, and the integrated module for data acquisition and analysis is connected to a human-machine interface through a communication module.
  • the module, the data acquisition and analysis integration module has a built-in non-volatile memory, an acquisition sensor, and a local data analysis processor, and the acquisition sensor is electrically connected to the local data analysis processor.
  • the invention also includes a method for collecting and analyzing motion trajectories based on ball motion, including data acquisition, data analysis, and data result and important trajectory information output of the ball trajectory, and the data collection and data analysis are collected and analyzed.
  • the integrated module performs data analysis and directly outputs it through the display device that is included in the integrated module of the acquisition and analysis or transmits it to the external terminal device through the communication module.
  • the data acquisition comprises: selecting an applicable acquisition sensor, selecting an appropriate sampling rate, and data acquisition with local data processing capability, multi-axis data fusion through its own processing capability, selecting a required acquisition sensor, and then passing through itself
  • the axis fusion algorithm obtains the linear acceleration affected by the separation of gravity.
  • the acquisition sensor comprises one or more of an acceleration sensor, a gyroscope, a magnetic field sensor, and an audio sensor.
  • the data analysis comprises: the local data analysis processor included in the acquisition and analysis integration module monitors the data collected by the acquisition sensor, determines the current motion type according to the surface data result, and flexibly sets the motion sensor according to the severity of the exercise. Sampling speed.
  • the output of the data result and the important trajectory information comprises: obtaining the analysis result by collecting and analyzing the integrated module data analysis, directly outputting through the display device and the display device on the integrated module, and selectively saving the analysis result in the collection. And analysis of the non-volatile memory that comes with the integrated module The analysis result is filtered according to the predetermined setting condition, and the sound and light and vibration feedback output are output according to the trigger condition, and the user is promptly reminded of the specific event, and the stored data is selectively sent to the upper device through the communication module.
  • the host device can be connected in real time as needed to view real-time results or display for display.
  • the upper device includes a mobile phone, a tablet computer, a PC, and a dedicated terminal.
  • the present invention has the following beneficial effects:
  • the system and method for collecting and analyzing motion path based on ball motion of the present invention can analyze the collected data without gaps because the local data analysis processor does not need to consider the capacity of the data storage and the bandwidth of the transmitted data.
  • the complete analysis of the action solves the limitation of communication bandwidth and storage capacity, maximizes the sampling speed of the sensor, and achieves the maximum sampling accuracy. At the same time, it greatly simplifies the system architecture, compresses the system cost, and simultaneously improves the user experience.
  • FIG. 1 is a schematic structural diagram of a system for collecting and analyzing motion trajectories based on ball motions according to the present invention.
  • a motion trajectory acquisition and analysis system based on ball motion
  • the system includes an integrated module for data acquisition and analysis, a communication module, and a human-machine interface display module, and the data acquisition and analysis integration module communicates
  • the module is connected to the human-machine interface display module, and the data acquisition and analysis integrated module has a built-in non-volatile memory, an acquisition sensor, and a local data analysis processor.
  • the set sensor is electrically connected to the local data analysis processor.
  • the invention also discloses a method for collecting and analyzing motion trajectories based on ball sports, comprising an integrated module for data acquisition and analysis, a communication module and a human-machine interface module, the method comprising data collection, data analysis and data of ball motion trajectories.
  • the result and the important trajectory information output, the data acquisition and the data analysis are performed by the integrated module of the collection and analysis, and after the data analysis, the output device is directly outputted through the collection and analysis integrated module or transmitted to the outside through the communication module.
  • the terminal device uses the appropriate sampling speed for different motions to achieve the highest performance power balance without considering the storage and communication bandwidth.
  • the data acquisition includes: selecting an applicable acquisition sensor, selecting an appropriate sampling rate, data acquisition with local data processing capability, multi-axis data fusion through its own processing capability, selecting a required sensor, and then using its own multi-axis fusion algorithm, The linear acceleration affected by the separation of gravity is obtained, providing an accurate source of data for motion recognition.
  • the multi-axis data fusion process is as follows:
  • the system tilt true angle ⁇ can be used to make a state vector.
  • the accelerometer is used to estimate the gyroscope constant deviation b, and the deviation is used as the state vector to obtain the corresponding state equation and observation equation:
  • ⁇ gyro is the angular velocity output of the gyroscope containing the fixed deviation
  • ⁇ acce is the angle value obtained by the accelerometer after processing
  • ⁇ g is the measurement noise of the gyroscope
  • ⁇ a is the accelerometer measurement noise
  • b is Gyro drift error
  • ⁇ g and ⁇ a are independent of each other.
  • the Kalman filter The recursive operation is performed until the optimal angle value is estimated.
  • the system process noise covariance matrix Q and the covariance matrix R of the measurement error are known to correct the Kalman filter.
  • the form of the Q and R matrix is as follows:
  • q_acce and q_gyro are the covariances measured by the accelerometer and the gyroscope, respectively, and their values represent the degree of trust of the Kalman filter on its sensor data. The smaller the value, the higher the degree of trust. In this system, the value of the gyroscope is closer to the exact value, so the value of q_gyro is less than the value of q_acce.
  • k-1) is the result of k prediction
  • k-1) is the optimal result at time k-1.
  • Equations (1) and (2) update the status of the system.
  • Equations (3), (4), and (5) are Kalman filter state update equations. After calculating the time update equation and the measurement update equation, the posterior estimate obtained from the previous calculation is repeated again as the prior calculation of the next calculation. It is estimated that, in this way, the cycle is repeated and repeated until the optimal result is found.
  • the earth's magnetic field, various magnetic conductive materials in the surface building, and other magnetic field sources together form a unique synthetic magnetic field whose direction is not necessarily the original geomagnetic direction, but is a fixed vector relative to the earth.
  • the 3-axis geomagnetic sensor can measure this vector. Since the geomagnetic sensor and other 6-axis sensor coordinate systems are set to be the same in circuit design, the geomagnetic vector provides an absolute direction reference vector for the resulting attitude after fusion, thus completing the 9-axis. Fusion.
  • the acquisition sensor includes one or more of an acceleration sensor, a gyroscope, a magnetic field sensor, and an audio sensor.
  • the present invention also introduces an audio sensor that analyzes motion characteristics from an audio dimension.
  • the audio sensor is a high-speed, high-precision sensor that provides more precise and detailed motion characteristics than motion sensors.
  • the invention uses an audio sensor to determine the collision time of the collision racquet and the ball, and the sampling speed of up to several tens of KHz ensures that the millisecond-level collision event will not be lost.
  • various features of the collision will be recorded differently, including: the collision position of the ball and the beat, the intensity of the collision, and the characteristics of the racket itself.
  • the data analysis includes: the local data analysis processor included in the acquisition and analysis integration module monitors the data collected by the acquisition sensor, determines the current motion type according to the surface data result, and flexibly sets the sampling speed of the motion sensor according to the severity of the motion.
  • the local data analysis processor will always monitor the data collected by the sensor, and can judge the current motion type according to the surface data result, and flexibly set the sampling speed of the motion sensor according to the severity of the motion to ensure that the power consumption is reduced without losing data, and at the same time Guaranteed sampling accuracy and power control.
  • the original application relies on the simplified migration of algorithms on PCs or other high-speed processors to the embedded system, making integration of acquisition and analysis possible.
  • the acceleration sensor detects that the threshold value b is continuously exceeded, and the angular velocity continuously exceeds the threshold; c, the attitude of the gravity detection or the fusion algorithm exceeds the change threshold d, A predetermined attitude or combination of actions, the system enters a high-speed sampling mode to cope with large dynamic motion analysis and recognition.
  • the data collected from the sensor group includes the original acceleration, angular velocity, and geomagnetic intensity.
  • the multi-axis fusion is combined with the original data to obtain the attitude parameter, the gravity parameter, the gravity-free linear acceleration parameter, and the azimuth of the relative earth.
  • the action type is obtained by the following analysis methods. And characteristics:
  • attitude parameters Through feature comparison and quantification through attitude parameters, gravity parameters, gravity-free linear acceleration parameters, instantaneous values of azimuth angles, and intermediate changes in velocity, displacement, attitude change, motion direction, direction of motion, and trend of change. Discriminate
  • the data result and the important trajectory information output include: obtaining the analysis result through the integrated module data analysis, directly outputting through the display device on the collection and analysis device, and selectively saving the analysis result in the collection analysis and the device.
  • the analysis result is filtered according to predetermined setting conditions, and the sound and light and vibration feedback output are output according to the trigger condition, and the user is promptly reminded of the specific event, and the stored data is selectively sent to the communication module.
  • Upper device According to the needs, you can connect the host device in real time, view the real-time results or display the display, filter the analysis results according to the predetermined setting conditions, and output the sound and light and vibration feedback devices in accordance with the trigger condition, prompting the user to a specific event.
  • the stored data can be selectively sent to a host device through a communication module, such as a mobile phone, a tablet computer, a PC, and a dedicated terminal, for continuing transmission to the cloud server for big data analysis.
  • a communication module such as a mobile phone, a tablet computer, a PC, and a dedicated terminal
  • the communication methods are not limited to: wireless communication modes of various frequency bands such as wifi, Bluetooth, NFC, and proprietary protocols, and wired connections such as USB, USART, 485, and CAN.
  • the upper device includes a mobile phone, a tablet computer, a PC, and a dedicated terminal.
  • the system and method for collecting and analyzing the motion trajectory based on the ball motion of the present invention may have no need to consider the capacity of the data storage and the bandwidth of the transmitted data because the local data analysis processor is used.
  • the gap analyzes the collected data to achieve a complete analysis of the action, solves the limitation of communication bandwidth and storage capacity, maximizes the sampling speed of the sensor, and achieves the maximum sampling accuracy; at the same time, greatly simplifies the system architecture and compresses the system cost. And at the same time improve the user experience.

Landscapes

  • Engineering & Computer Science (AREA)
  • Multimedia (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Health & Medical Sciences (AREA)
  • General Health & Medical Sciences (AREA)
  • Physical Education & Sports Medicine (AREA)
  • Gyroscopes (AREA)
  • Testing Or Calibration Of Command Recording Devices (AREA)

Abstract

一种基于球类运动的运动轨迹采集和分析***，包括数据采集和分析一体化模块、通讯模块和人机界面显示模块，数据采集和分析一体化模块通过通讯模块连接人机界面显示模块，数据采集和分析一体化模块内置非易失存储器、采集传感器以及本地数据分析处理器，采集传感器与本地数据分析处理器电连接。还公开了一种基于球类运动的运动轨迹采集和分析的方法。

Description

一种基于球类运动的运动轨迹采集和分析的***与方法 技术领域
本发明涉及球类运动轨迹监测技术领域，特别涉及一种基于球类运动的运动轨迹采集和分析的***与方法。
背景技术
随着传感器技术、计算机技术的发展以及人们对游戏、体育的需求急剧增长，动作识别有很大的需求。
现有的动作识别***，都是由采集模块，通讯模块和计算模块组成。目前采集模块主要通过传感器包括加速度传感器、陀螺仪、磁力计等和运动以及方位有关的传感器，采集并初步分析数据，然后把数据通过通讯模块传输到上位计算机设备包括手机、平板电脑等移动设备或者计算机进行最终的动作分析。由于动作识别的算法的是基于传感器数据的一次甚至多次积分及空间变换重组，所以相同的噪声前提下，采样率越高，最终动作识别率越高，轨迹还原度越高。而对于某些动作速度非常高的球类运动，例如拍类运动里的羽毛球、乒乓球，带投球动作的棒球、垒球、板球，以及足球等，由于动作变化非常快，不够高的采样率会导致关键数据点缺失，带来的不单只是识别精度的降低问题，严重的会导致动作识别错误，使采集的数据变成一堆无效数据。参与采集数据的传感器数量众多，采集到的原始数据数量庞大，在不干涉动的前提下，通讯模块把原始数据传送到下级单元的手段主要是无线传输，巨大的数据量会对无线通讯的带宽和功耗带来巨大的压力，甚至超出无线信道的承受能力，即使是使用了数据的初步识别和筛选，面对高速的运动分析需要的巨大采集量，原始数据的传输仍然是一个瓶颈，会大大影响到识别范围和用户体验度。
发明内容
本发明为避免上述现有技术所存在的不足之处，提供一种基于球类运动的 运动轨迹采集和分析的方法。利用本发明所给出的方法，能解决通讯带宽和存储容量的限制问题，可以最大限度发挥传感器采样速度的，达到最大的采样精度。可以无间隙跟踪分析整个运动过程，不丢失细小运动数据。能不依靠上位设备独立工作，大大简化***架构，压缩了***成本，并同时提高用户体验。
本发明的技术方案是这样实现的：
一种基于球类运动的运动轨迹采集和分析***，***包括数据采集和分析一体化模块、通讯模块以及人机界面显示模块，所述数据采集和分析一体化模块通过通讯模块连接人机界面显示模块，所述数据采集和分析一体化模块内置有非易失存储器、采集传感器以及本地数据分析处理器，所述采集传感器与本地数据分析处理器电性连接。
本发明还包括一种基于球类运动的运动轨迹采集和分析的方法，包括球类运动轨迹的数据采集、数据分析以及数据结果和重要轨迹信息输出，所述数据采集与数据分析采用采集和分析一体化模块进行，经过数据分析后通过采集和分析一体化模块上自带的显示设备直接输出或是通过通讯模块传送至外部终端设备。
优选地，所述数据采集包括：选择适用的采集传感器、选择适当的采样速率，数据采集带有本地数据处理能力，通过自身处理能力进行多轴数据融合，选用所需采集传感器，然后通过自身多轴融合算法，得到分离重力影响的线性加速度。
优选地，所述采集传感器包括：加速度传感器、陀螺仪、磁场传感器以及声频传感器中的一种或者多种。
优选地，所述数据分析包括：采集和分析一体化模块包括的本地数据分析处理器会一直监测采集传感器采集到的数据，根据表面数据结果判断当前运动类型，根据运动剧烈程度灵活设置运动传感器的采样速度。
优选地，数据结果和重要轨迹信息输出包括：通过采集和分析一体化模块数据分析得到分析结果，通过采集和分析一体化模块上自带的显示设备直接输出，有选择的把分析结果保存在采集和分析一体化模块上自带的非易失存储器 中，根据预定的设置条件对分析结果进行过滤，符合触发条件的情况通过声光、震动反馈输出，及时提醒用户特定事件，有选择的把存储的数据，通过通讯模块发送给上位设备。
优选地，根据需要可以实时连接上位设备，查看实时结果或者进行展示用显示。
优选地，所述上位设备包括手机、平板电脑、PC、专用终端。
与现有技术相比，本发明具有以下有益效果：
本发明的基于球类运动的运动轨迹采集和分析的***与方法，由于使用了本地数据分析处理器不需要考虑数据存储的容量和传输数据的带宽问题，可以无间隙对采集数据进行分析，达到动作的完整分析，解决了通讯带宽和存储容量的限制问题，可以最大限度发挥传感器采样速度的，达到最大的采样精度；同时大大简化了***架构，压缩了***成本，并同时提高用户体验。
附图说明
为了更清楚地说明本发明实施例或现有技术中的技术方案，下面将对实施例或现有技术描述中所需要使用的附图作简单地介绍，显而易见地，下面描述中的附图仅仅是本发明的一些实施例，对于本领域普通技术人员来讲，在不付出创造性劳动性的前提下，还可以根据这些附图获得其他的附图。
图1为本发明基于球类运动的运动轨迹采集和分析的***结构示意图。
具体实施方式
下面将结合本发明实施例中的附图，对本发明实施例中的技术方案进行清楚、完整地描述，显然，所描述的实施例仅仅是本发明一部分实施例，而不是全部的实施例。基于本发明中的实施例，本领域普通技术人员在没有作出创造性劳动前提下所获得的所有其他实施例，都属于本发明保护的范围。
如图1所示，一种基于球类运动的运动轨迹采集和分析***，***包括数据采集和分析一体化模块、通讯模块以及人机界面显示模块，所述数据采集和分析一体化模块通过通讯模块连接人机界面显示模块，所述数据采集和分析一体化模块内置有非易失存储器、采集传感器以及本地数据分析处理器，所述采 集传感器与本地数据分析处理器电性连接。
本发明还公开一种基于球类运动的运动轨迹采集和分析的方法，包括数据采集分析一体化模块、通讯模块、人机界面模块，该方法包括球类运动轨迹的数据采集、数据分析以及数据结果和重要轨迹信息输出，所述数据采集与数据分析采用采集和分析一体化模块进行，经过数据分析后通过采集和分析一体化模块上自带的显示设备直接输出或是通过通讯模块传送至外部终端设备，根据不同的运动使用适合的采样速度，以达到最高的性能功耗平衡，不需要考虑存储和通讯带宽。
所述数据采集包括：选择适用的采集传感器、选择适当的采样速率，数据采集带有本地数据处理能力，通过自身处理能力进行多轴数据融合，选用所需传感器，然后通过自身多轴融合算法，得到分离重力影响的线性加速度，为动作识别提供准确的数据来源。其中多轴数据融合过程如下：
1)利用卡尔曼滤波对3轴加速度和3轴陀螺仪数据进行融合
首先建立***的状态方程和测量方程。由于倾角和倾角角速度存在导数关系，***倾斜真实角度φ可以用来做一个状态向量。在该***中，采用加速度计估计出陀螺仪常值偏差b，以此偏差作为状态向量得到相应的状态方程和观测方程:
Figure PCTCN2016080191-appb-000001
式中,ωgyro为包含固定偏差的陀螺仪输出角速度,φacce为加速度计经处理后得到的角度值，ωg为陀螺仪测量噪声，ωa为加速度计测量噪声，b为 陀螺仪漂移误差，ωg与ωa相互独立，此处假设二者为满足正态分布的白色噪声,令Ts为***采样周期，得到离散***的状态方程和测量方程:
Figure PCTCN2016080191-appb-000002
同时，要估算k时刻的实际角度，就必须根据k-1时刻的角度值，再根据预测得到的k时刻的角度值，得到k时刻的高斯噪声的方差，在此基础之上卡尔曼滤波器进行递归运算直至估算出最优的角度值.在此，须知道***过程噪声协方差阵Q以及测量误差的协方差矩阵R，对卡尔曼滤波器进行校正。Q与R矩阵的形式如下:
Figure PCTCN2016080191-appb-000003
式中，q_acce和q_gyro分别是加速度计和陀螺仪测量的协方差，其数值代表卡尔曼滤波器对其传感器数据的信任程度，值越小，表明信任程度越高。在该***中陀螺仪的值更为接近准确值,因此取q_gyro的值小于q_acce的值.当前状态:
X(k|k-1)＝AX(k-1|k-1)+BU(k)
式中，
Figure PCTCN2016080191-appb-000004
X(k|k-1)是利用k预测的结果,X(k-1|k-1)是k-1时刻的最优结果。
则有对应于X(k|k-1)的协方差为:
P(k|k-1)＝AP(k-1|k-1)AT+Q   (2)式中，
P(k-1|k-1)是X(k-1|k-1)对应的协方差，AT表示A的转置矩阵，Q是***过程的协方差。式子(1)、(2)即对***的状态更新。
则状态k的最优化估算值X(k|k):
X(k|k)＝X(k|k-1)+K(k)(Z(k)-HX(k|k-1))    (3)其中H＝[10]，K为卡尔曼增益(KalmanGain):
K(k)＝P(k|k-1)HT/(HP(k|k-1)HT+R)    (4)此时，我们已经得到了k状态下最优的估算值X(k|k)，但是为了使卡尔曼滤波器不断的运行下去直到找到最优的角度值，我们还要更新k状态下X(k|k)的协方差：
P(k|k)＝(I-Kg(k)H)P(k|k-1)   (5)其中，I为单位阵，对于本***则有，
Figure PCTCN2016080191-appb-000005
当***进入k+1状态时，P(k|k)就是式子(2)的P(k-1|k-1)。(3)、(4)、(5)式为卡尔曼滤波器状态更新方程.计算完时间更新方程和测量更新方程后,再次重复上一次计算得到的后验估计,作为下一次计算的先验估计,这样,周而复始、循环反复地运算下去直至找到最优的结果。
经过6轴融合后，由于得到了比较精确的瞬时姿态，在此基础上可以准确的计算该姿态下各轴受到的重力分量，和原始加速度之间差值即为去除重力影响的线性加速度。
2)结合地磁传感器进行二次方向融合
地球磁场、地表建筑中各种导磁物质、其他磁场源一起会构成一个唯一的合成磁场，其方向不一定是原来地磁的方向，但是相对地球是一个固定的矢量。 3轴地磁传感器可以把这个矢量测量出来，由于地磁传感器和其他6轴传感器坐标系在电路设计时设成相同，所以地磁矢量为融合后得出姿态提供一个绝对方向参考矢量，从而完成9轴的融合。
所述采集传感器包括：加速度传感器、陀螺仪、磁场传感器以及声频传感器中的一种或者多种，本发明还引入了声频传感器，从声频维度分析运动特征。声频传感器为高速高精度传感器，可以得到比运动传感器更为精密细致的运动特征。本发明使用音频传感器用来判断碰撞球拍和球的碰撞时刻，高达数十KHz的采样速度，保证毫秒级的碰撞事件不会被丢失。另一方面，得益于声频传感器的高速高精度采样，碰撞时的多种特征会被有差别的记录下来，包括：球与拍的碰撞位置、碰撞剧烈程度、球拍自身属性特征等。
所述数据分析包括：采集和分析一体化模块包括的本地数据分析处理器会一直监测采集传感器采集到的数据，根据表面数据结果判断当前运动类型，根据运动剧烈程度灵活设置运动传感器的采样速度。本地数据分析处理器会一直监测传感器采集到的数据，可以根据表面数据结果判断当前运动类型，根据运动剧烈程度灵活设置运动传感器的采样速度，保证不丢失数据的前提下达到功耗的降低，同时保证采样精度和功耗控制。通过对算法的高度优化，把原来应用依赖PC或其他高速处理器上的算法简化移植到嵌入式***，使采集分析一体化变得有可能。当检测到声频传感器的碰撞事件或者预定的某些传感器例如a、加速度传感器检测到连续超过阀值b、角速度连续超过阀值；c、重力检测或者融合算法出来的姿态发生超过变化阀值d、预定的某种姿态或者动作组合，***进入高速采样模式，以应付大动态动作分析识别。
从传感器组采集到的数据包括原始加速度、角速度、地磁强度，结合原始数据经过多轴融合得到姿态参数、重力参数、无重力线性加速度参数，相对大地的方位角，通过各如下分析方法得到动作类型和特征：
1、通过姿态参数、重力参数、无重力线性加速度参数、方位角的瞬时值和积分后的速度、位移、姿态变化量、运动方向、运动方向趋势、变化趋势等中间变化量通过特征对比、量化判别；
2、通过上述原始参数和中间变化量的模糊算法识别；
3、通过上述原始参数和中间变化量的相似度判别算法；
4、通过上述原始参数和中间变化量和自适应学习结果的对比。
所述数据结果和重要轨迹信息输出包括：通过一体化模块数据分析得到分析结果，通过采集、分析设备上自带的显示设备直接输出，有选择的把分析结果保存在采集分析、设备上自带的非易失存储器中，根据预定的设置条件对分析结果进行过滤，符合触发条件的情况通过声光、震动反馈输出，及时提醒用户特定事件，有选择的把存储的数据，通过通讯模块发送给上位设备。根据需要可以实时连接上位设备，查看实时结果或者进行展示用显示，根据预定的设置的条件对分析结果进行过滤，符合触发条件的情况通过声光、震动反馈设备输出，及时提醒用户特定事件。可以有选择的把存储的数据，通过通讯模块发送给上位设备例如：手机、平板电脑、PC、专用终端，用于继续传送到云服务器后台进行大数据分析。通讯方式包括并不限于：wifi、蓝牙、NFC、私有协议的各种频段无线通讯方式以及USB、USART、485、CAN等有线连接。所述上位设备包括手机、平板电脑、PC、专用终端。
综合本发明的方法步骤可知，本发明的基于球类运动的运动轨迹采集和分析的***与方法，由于使用了本地数据分析处理器不需要考虑数据存储的容量和传输数据的带宽问题，可以无间隙对采集数据进行分析，达到动作的完整分析，解决了通讯带宽和存储容量的限制问题，可以最大限度发挥传感器采样速度的，达到最大的采样精度；同时大大简化了***架构，压缩了***成本，并同时提高用户体验。
以上所述仅为本发明的较佳实施例而已，并不用以限制本发明，凡在本发明的精神和原则之内，所作的任何修改、等同替换、改进等，均应包含在本发明的保护范围之内。

Claims (8)

  1. 一种基于球类运动的运动轨迹采集和分析***，其特征在于，***包括数据采集和分析一体化模块、通讯模块以及人机界面显示模块，所述数据采集和分析一体化模块通过通讯模块连接人机界面显示模块，所述数据采集和分析一体化模块内置有非易失存储器、采集传感器以及本地数据分析处理器，所述采集传感器与本地数据分析处理器电性连接。
  2. 一种基于球类运动的运动轨迹采集和分析的方法，包括球类运动轨迹的数据采集、数据分析以及数据结果和重要轨迹信息输出，其特征在于，所述数据采集与数据分析采用采集和分析一体化模块进行，经过数据分析后通过采集和分析一体化模块上自带的显示设备直接输出或是通过通讯模块传送至外部终端设备。
  3. 如权利要求2所述的基于球类运动的运动轨迹采集和分析的方法，其特征在于，所述数据采集包括：选择适用的采集传感器、选择适当的采样速率，数据采集带有本地数据处理能力，通过自身处理能力进行多轴数据融合，选用所需采集传感器，然后通过自身多轴融合算法，得到分离重力影响的线性加速度。
  4. 如权利要求3所述的基于球类运动的运动轨迹采集和分析的方法，其特征在于，所述采集传感器包括：加速度传感器、陀螺仪、磁场传感器以及声频传感器中的一种或者多种。
  5. 如权利要求2所述的基于球类运动的运动轨迹采集和分析的方法，其特征在于，所述数据分析包括：采集和分析一体化模块包括的本地数据分析处理器会一直监测采集传感器采集到的数据，根据表面数据结果判断当前运动类型，根据运动剧烈程度灵活设置运动传感器的采样速度。
  6. 如权利要求2所述的基于球类运动的运动轨迹采集和分析的方法，其特征在于，数据结果和重要轨迹信息输出包括：通过采集和分析一体化模块数据分析得到分析结果，通过采集和分析一体化模块上自带的显示设备直接输出，有选择的把分析结果保存在采集和分析一体化模块上自带的非易失存储器中，根据预定的设置条件对分析结果进行过滤，符合触发条件的情况通过声光、震 动反馈输出，及时提醒用户特定事件，有选择的把存储的数据，通过通讯模块发送给上位设备。
  7. 如权利要求6所述的基于球类运动的运动轨迹采集和分析的方法，其特征在于，根据需要可以实时连接上位设备，查看实时结果或者进行展示用显示。
  8. 如权利要求7所述的基于球类运动的运动轨迹采集和分析的方法，其特征在于，所述上位设备包括手机、平板电脑、PC、专用终端。
PCT/CN2016/080191 2016-04-26 2016-04-26 一种基于球类运动的运动轨迹采集和分析的***与方法 WO2017185222A1 (zh)

Priority Applications (2)

Application Number Priority Date Filing Date Title
PCT/CN2016/080191 WO2017185222A1 (zh) 2016-04-26 2016-04-26 一种基于球类运动的运动轨迹采集和分析的***与方法
CN201680011563.8A CN107454970A (zh) 2016-04-26 2016-04-26 一种基于球类运动的运动轨迹采集和分析的***与方法

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/CN2016/080191 WO2017185222A1 (zh) 2016-04-26 2016-04-26 一种基于球类运动的运动轨迹采集和分析的***与方法

Publications (1)

Publication Number Publication Date
WO2017185222A1 true WO2017185222A1 (zh) 2017-11-02

Family

ID=60160613

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2016/080191 WO2017185222A1 (zh) 2016-04-26 2016-04-26 一种基于球类运动的运动轨迹采集和分析的***与方法

Country Status (2)

Country Link
CN (1) CN107454970A (zh)
WO (1) WO2017185222A1 (zh)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112929332A (zh) * 2021-01-19 2021-06-08 北京睿芯高通量科技有限公司 一种多轴物体运动检测***及检测方法
CN114796986A (zh) * 2022-06-08 2022-07-29 深圳市汇泰科电子有限公司 一种识别壶铃运动信息的方法

Families Citing this family (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111767932B (zh) * 2019-04-02 2024-02-06 北京深蓝长盛科技有限公司 动作判定方法及装置、计算机设备及计算机可读存储介质
CN110440799A (zh) * 2019-09-19 2019-11-12 哈尔滨工程大学 一种基于陀螺仪和加速度计的姿态角度测量融合***及方法
CN112957688B (zh) * 2021-01-27 2022-02-01 牛鹤璇 一种智能体育运动数据测量方法及设备

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102553200A (zh) * 2012-03-16 2012-07-11 崔宝生 基于zigbee的高尔夫球落点定位***
CN102989163A (zh) * 2012-11-23 2013-03-27 安徽寰智信息科技股份有限公司 一种基于动作识别的保龄球训练辅助***
CN204395358U (zh) * 2014-12-20 2015-06-17 西安思能网络科技有限公司 一种智能足球装置
CN105056530A (zh) * 2015-08-11 2015-11-18 王燕军 一种实时虚拟再现运动球类
US20160023042A1 (en) * 2014-07-22 2016-01-28 Cleopatrick Cherry Tracking Assembly
CN105435430A (zh) * 2014-09-26 2016-03-30 上海珏芯光电科技有限公司 空间运动轨迹自跟踪高尔夫球及其空间运动轨迹感测方法

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR20060008835A (ko) * 2004-07-24 2006-01-27 삼성전자주식회사 가속도 센서를 이용한 운동량 측정장치 및 방법
CN102023700B (zh) * 2009-09-23 2012-06-06 吴健康 一种三维人机交互***
US8617005B2 (en) * 2010-08-05 2013-12-31 Active Mind Technologies R&D Limited Golf data collection

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102553200A (zh) * 2012-03-16 2012-07-11 崔宝生 基于zigbee的高尔夫球落点定位***
CN102989163A (zh) * 2012-11-23 2013-03-27 安徽寰智信息科技股份有限公司 一种基于动作识别的保龄球训练辅助***
US20160023042A1 (en) * 2014-07-22 2016-01-28 Cleopatrick Cherry Tracking Assembly
CN105435430A (zh) * 2014-09-26 2016-03-30 上海珏芯光电科技有限公司 空间运动轨迹自跟踪高尔夫球及其空间运动轨迹感测方法
CN204395358U (zh) * 2014-12-20 2015-06-17 西安思能网络科技有限公司 一种智能足球装置
CN105056530A (zh) * 2015-08-11 2015-11-18 王燕军 一种实时虚拟再现运动球类

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112929332A (zh) * 2021-01-19 2021-06-08 北京睿芯高通量科技有限公司 一种多轴物体运动检测***及检测方法
CN114796986A (zh) * 2022-06-08 2022-07-29 深圳市汇泰科电子有限公司 一种识别壶铃运动信息的方法
CN114796986B (zh) * 2022-06-08 2024-03-08 深圳市汇泰科电子有限公司 一种识别壶铃运动信息的方法

Also Published As

Publication number Publication date
CN107454970A (zh) 2017-12-08

Similar Documents

Publication Publication Date Title
CN106705968B (zh) 基于姿态识别和步长模型的室内惯性导航算法
WO2017185222A1 (zh) 一种基于球类运动的运动轨迹采集和分析的***与方法
AU2020273327B2 (en) Systems and methods of swimming analysis
US11166104B2 (en) Detecting use of a wearable device
US20240180445A1 (en) Systems and methods for determining swimming metrics
Zhou et al. Use it free: Instantly knowing your phone attitude
WO2018068771A1 (zh) 目标跟踪方法、***、电子设备和计算机存储介质
US20210068713A1 (en) Detecting swimming activities on a wearable device
US20120323521A1 (en) System and method for recognizing gestures
KR101872907B1 (ko) 듀얼 스마트 밴드를 이용한 모션 분석 장치 및 방법
CN110579207B (zh) 基于地磁信号结合电脑视觉的室内定位***及方法
CN104536558A (zh) 一种智能指环和控制智能设备的方法
CN107766930B (zh) 基于dtw群集模糊分群som神经元的等效rom距离计算方法
CN109724602A (zh) 一种基于硬件fpu的姿态解算***及其解算方法
CN111595344B (zh) 一种基于地图信息辅助的多姿态下行人行位推算方法
CN106108909A (zh) 一种人体姿态检测穿戴设备、***及控制方法
Qiu et al. MAGINS: Neural network inertial navigation system corrected by magnetic information
CN111435083A (zh) 行人航迹推算方法、导航方法及装置、手持终端及介质
CN109758737A (zh) 水中运动的姿势矫正装置及方法
WO2019047327A1 (zh) 一种智能球、***及方法
Lin et al. A Cnn-Speed-Based Gnss/Pdr Integrated System For Smartwatch
US20130006572A1 (en) Methods for detection of unique physical motions
US20230149789A1 (en) System and method for golf super tag multifunction golf swing capture and analysis device
Gu et al. Heading judgment for the waist-mounted MIMU using LSTM
Deng et al. An Indoor Intelligent Care System Based on the Fusion of Vision and UWB Heterogeneous Sensors

Legal Events

Date Code Title Description
NENP Non-entry into the national phase

Ref country code: DE

121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 16899736

Country of ref document: EP

Kind code of ref document: A1

122 Ep: pct application non-entry in european phase

Ref document number: 16899736

Country of ref document: EP

Kind code of ref document: A1