A six-axis motion capture system and motion capture method

Abstract

The present invention belongs to the field of motion capture and data processing, and discloses a six-axis motion capture system, which includes a motion capture unit, a control unit, a mobile terminal and a host computer; the motion capture unit and the control unit are connected through a hardware interface, and the control unit and the host computer and the mobile terminal are connected through a wireless network. The system has a simple structure, convenient management, and the system responds quickly, accurately and efficiently. The method for motion capture using the above system is: S1, the motion capture unit generates motion data according to the motion trajectory; S2, the mobile terminal sends a request, and the control unit obtains the data obtained by the motion capture unit; S3, the control unit transmits the motion data to the mobile terminal; S4, the mobile terminal visualizes the motion data to achieve 3D rendering and dynamic synchronization. The present invention can greatly help doctors to accurately and conveniently track and quantify the effects of rehabilitation patients in the rehabilitation treatment recovery process in real time through the mobile terminal.

Description

Technical Field

[0001] The present invention relates to the field of motion capture and data processing, and in particular to a six-axis motion capture system and a motion capture method. Background Art

[0002] With the development of computer hardware technology, the application field of motion capture and recognition technology is becoming more and more extensive. Traditional motion capture and recognition is mainly based on posture analysis of computer vision, that is, the movement process of the human body is captured by the camera at different angles at the same time, and relevant visual information is extracted from the video sequence. The motion recognition is realized by finding a suitable method to represent the visual information. In the above process, on the one hand, due to the individual differences between the human body, the diversity and complexity of the movements, how to fully and effectively express the movement posture is a major difficulty in the posture analysis based on vision, and in specific applications, it is also necessary to balance the computational cost and accuracy; on the other hand, the posture analysis method based on computer vision mainly relies on optical components and camera technology (such as PhaseSpace, NOKOV and Worldviz). The above analysis method has the following main defects:

[0003] (1) The overall optical system price is generally high, usually between $250,000 and $500,000, which cannot meet the price requirements of the general public. Most companies' equipment requires daily maintenance by professionals. In addition, the entire system, including workstations, software and data modification, will multiply the cost of use and hardware costs.

[0004] (2) The system is expensive. It can capture real-time motion, but the post-processing (including marker identification, tracking, and calculation of spatial coordinates) is very labor-intensive and is only suitable for scientific research applications.

[0005] Based on the above, the field has improved the motion capture recognition technology after long-term research, and proposed the posture analysis using inertial sensors. In this analysis method, the human body can wear a special sensor to obtain data, and then process it after the model is established by the relevant processing software, so as to achieve the purpose of motion recognition, and the motion capture recognition technology has been widely used in the field of rehabilitation medicine (such as InterSense). However, the posture analysis using inertial sensors cannot reflect everything, and the lag phenomenon of inertial capture data is quite obvious.

[0006] Based on the above analysis, it can be seen that the products used for motion capture and recognition still have the following problems:

[0007] (1) It is almost impossible to support low-cost motion capture technology based on mobile terminals to meet the needs of quantitative tracking after rehabilitation surgery. The equipment in optical components and camera technology requires daily maintenance by professionals, and the sales prices are generally high, which cannot meet the price requirements of the general public.

[0008] (2) The entire system, including workstations, software and data modification, will multiply the cost of use and hardware expenses.

[0009] (3) The system operation procedures are complex, the investment cost is high, and it is easily affected and interfered by external factors. This will result in the entire motion capture technology being used only in the fields of scientific research or film production. It is difficult to use it in the field of rehabilitation medicine to accurately assist doctors in rehabilitating patients.

[0010] Therefore, there is an urgent need to develop a new motion capture recognition system in this field to eliminate the defects and limitations of the prior art. Summary of the invention

[0011] The purpose of the present invention is to provide a six-axis motion capture system with a simple structure, convenient management, and quick, accurate and efficient system response.

[0012] The technical solution for achieving the purpose of the present invention is:

[0013] A six-axis motion capture system, comprising a motion capture unit, a control unit and a mobile terminal; the motion capture unit is used to collect motion data, the control unit is used to obtain the motion data generated by the motion capture unit and transmit it to the mobile terminal through a wireless network; the mobile terminal processes the motion data;

[0014] The motion capture unit and the control unit are connected via a hardware interface, and the control unit and the mobile terminal are connected via a wireless network; wherein, there are two motion capture units, namely a first motion capture unit and a second motion capture unit, and the two motion capture units are connected in parallel and then in series with the control unit.

[0015] Furthermore, it also includes a host computer; the host computer is used to measure the difference between the curves of the thigh and calf angles of a normal person when walking and the curves of the thigh and calf angles of a patient with deformed legs, and the host computer and the control unit are connected via a wireless network.

[0016] Another object of the present invention is to propose a six-axis motion capture method, which has a simple operation procedure and can greatly help doctors to accurately and conveniently track and quantify the effects of rehabilitation patients in the rehabilitation treatment recovery process in real time through mobile terminals.

[0017] The technical solution to achieve another object of the present invention is:

[0018] A motion capture method of a six-axis motion capture system comprises the following steps:

[0019] S1, the motion capture unit generates motion data according to the motion trajectory;

[0020] S2, the mobile terminal sends a request, and the control unit obtains the motion data generated by the motion capture unit;

[0021] S3, the control unit obtains the motion data of the motion capture unit and transmits it to the mobile terminal;

[0022] S4, the mobile terminal visualizes the motion data to achieve 3D rendering and dynamic synchronization.

[0023] The motion capture system of the present invention is mainly used in medical rehabilitation knee pads. The two motion capture units are divided into a thigh bracket and a calf bracket installed on a common knee pad bracket, and the control unit is arranged on the knee part of the knee pad. The control unit obtains the data of the motion capture unit at the request of the mobile terminal, and then transmits the data to the mobile terminal, which realizes the construction of the 3D model through software processing.

[0024] The system of the present invention also adopts a host computer. Before manufacturing or adjusting the medical rehabilitation knee brace, the host computer can obtain data from the motion capture unit, and calculate the obtained data with the normal data stored in the host computer database, accurately measuring the difference between the thigh and calf angle curves of a normal person when walking and the thigh and calf angle curves of patients with deformed legs, thereby assisting doctors in making diagnoses and helping doctors to better manufacture and adjust the medical rehabilitation knee brace.

[0025] The connection method of the system of the present invention is simple, and the data of the two motion capture units can be obtained synchronously, realizing the synchronous acquisition of the motion data of the thigh support and the calf support. Through the motion capture system of the present invention, the patient's motion state can be monitored and analyzed in fact, and the effect of rehabilitation patients in the rehabilitation treatment recovery process can be accurately and conveniently tracked and quantified through the mobile terminal in real time, and then the treatment plan can be adjusted in time to improve the rehabilitation treatment effect and shorten the rehabilitation cycle.

[0026] The beneficial effects of the present invention are:

[0027] (1) The present invention uses a mobile terminal to monitor the motion status of the system, which is convenient to carry and observe at any time, further improving the detection efficiency.

[0028] (2) The system of the present invention has a relatively simple structure, a small size, is easy to produce, has a low investment cost, and is not easily affected or disturbed by external factors.

[0029] (3) The present invention has a simple operation procedure, which can greatly help doctors to accurately and conveniently track and quantify the effects of rehabilitation patients in the rehabilitation treatment process in real time through mobile terminals.

[0030] (4) Before manufacturing or adjusting the medical rehabilitation knee brace, the present invention can obtain motion capture data through a host computer and measure the difference between the curves of the thigh and calf angles of a normal person when walking and the curves of the thigh and calf angles of a patient with deformed legs, thereby assisting doctors in making diagnoses and helping doctors to better manufacture and adjust the medical rehabilitation knee brace. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0032] Figure 1 A working principle diagram of the present invention;

[0033] Figure 2 This is a comparison chart of positive and abnormal data of the angles of the upper and lower legs;

[0034] Figure 3 It is a comparison chart of positive and abnormal data of relative angles of knee joints of both legs;

[0035] Figure 4 This is a comparison chart of the abnormal speed data of the upper and lower leg swings;

[0036] Figure 5 It is a comparison chart of positive and abnormal data of knee joint force.

[0037] Figure 6 A structural schematic diagram of the hardware connection method of the present invention;

[0038] Figure 7 Another working principle diagram of the hardware of the present invention;

[0039] Figure 8 A flow chart of the method of the present invention;

[0040] Fig. 9 A diagram for specifying the value of VALUE in the present invention;

[0041] Fig.10 It is a kind of motion trajectory and trend diagram obtained during the operation of the system of the present invention; DETAILED DESCRIPTION

[0042] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0043] like Figure 1 As shown, a six-axis motion capture system includes a motion capture unit, a control unit and a mobile terminal; the motion capture unit is used to collect motion data, the control unit is used to obtain the motion data generated by the motion capture unit, and transmit it to the mobile terminal through a wireless network; the mobile terminal processes the motion data;

[0044] The motion capture unit and the control unit are connected via a hardware interface, and the control unit and the mobile terminal are connected via a wireless network; wherein, there are two motion capture units, namely a first motion capture unit and a second motion capture unit, and the two motion capture units are connected in parallel and then in series with the control unit.

[0045] Furthermore, if Figure 1 As shown, it also includes a host computer; the host computer is used to measure the difference between the curves of the thigh and calf angles of a normal person when walking and the curves of the thigh and calf angles of a patient with deformed legs, and the host computer and the control unit are connected via a wireless network.

[0046] The motion capture system of the present invention is mainly used in medical rehabilitation knee pads. The two motion capture units are divided into a thigh bracket and a calf bracket installed on an ordinary knee pad bracket, and the control unit is arranged on the knee part of the knee pad. The control unit obtains the data of the motion capture unit at the request of the mobile terminal, and then transmits the data to the mobile terminal, and the mobile terminal realizes the construction of the 3D model through software processing. The system of the present invention also adopts a host computer. Before manufacturing or adjusting the medical rehabilitation knee pad, the host computer can obtain the data of the motion capture unit, and calculate the acquired data with the normal data stored in the host computer database, accurately measure the curve of the thigh and calf angle when a normal person walks and the difference between the thigh and calf angle curves of patients with deformed legs, thereby assisting doctors in making diagnoses and helping doctors to better manufacture and adjust medical rehabilitation knee pads. The connection mode of the system of the present invention is simple, and the data of the two motion capture units can be obtained synchronously, so as to realize the synchronous acquisition of the motion data of the thigh bracket and the calf bracket. Through the motion capture system of the present invention, the patient's motion state can be monitored and analyzed in fact, and the effect of rehabilitation patients in the rehabilitation treatment recovery process can be conveniently and accurately tracked and quantified through the mobile terminal in real time and conveniently, and then the treatment plan can be adjusted in time to achieve the effect of improving the rehabilitation treatment effect and shortening the rehabilitation cycle.

[0047] The motion capture unit in the present invention is a posture sensor, the control unit is a microcontroller, the mobile terminal is a mobile phone, and the host computer is a computer.

[0048] The posture sensor can use MPU6000, MPU6050 or MPU6055. The present invention preferably uses MPU6055, which has high data efficiency, low power consumption and small transmission signal jump.

[0049] The microcontroller can be ESP6088 or ESP32, and the present invention preferably uses ESP32. ESP32 is a WiFi chip that integrates an antenna switch, a radio frequency balun, a power amplifier, a low noise amplifier, a filter, and a power management module. The entire solution occupies the least printed circuit board area. The 2.4GHz Wi-Fi plus Bluetooth dual-mode chip uses TSMC's low-power 40nm technology, has the best power consumption performance and radio frequency performance, is safe and reliable, and is easy to expand to various applications.

[0050] The host computer is a computer, and the computer database contains the motion data of normal people, namely the normal values ​​ax1, ay1 and az1. The computer can obtain the patient's motion data collected by the posture sensor through the microcontroller, and obtain the abnormal ax2, ay2 and az2 after preprocessing. The computer then measures the positive and abnormal data to obtain the difference between the curve of the thigh and calf angle when a normal person walks and the curve of the thigh and calf angle of patients with deformed legs, so as to further help doctors design appropriate medical rehabilitation knee braces or adjust them.

[0051] Furthermore, the host computer includes a preprocessing unit and a data calculation unit.

[0052] The preprocessing unit preprocesses the data acquired by the control unit, and the preprocessing steps are:

[0053] S101, remove duplicate data

[0054] The deduplication process uses a period of time as a cycle, and the specific length of the period can be adjusted according to the actual situation.

[0055] S102, if the absolute value of the difference between the current value and the previous value is greater than 0.6, the value A before replacement is saved, otherwise it is discarded, and finally one data is retained;

[0056] S103, all values ​​were normalized by dividing by the maximum value;

[0057] The calculation contents of the data calculation unit include: comparison and calculation of positive and abnormal data of the angles of the upper and lower legs, comparison and calculation of positive and abnormal data of the relative angles of the knee joints of both legs, comparison and calculation of positive and abnormal data of the swing speeds of the upper and lower legs, and comparison and calculation of positive and abnormal data of the forces on the knee joints.

[0058] The 0.6 in step S102 is a threshold value set for filtering noise data after the data is captured by the motion capture chip. After a large number of tests, the data obtained at 0.6 has the most obvious features, and the fitted value is the smoothest.

[0059] Furthermore, the data calculation units are calculated using linear regression methods.

[0060] The steps for comparing and calculating the positive and abnormal data of the thigh and calf angles are as follows:

[0061] Y1=A*[ay1]+b1 Formula (1)

[0062] Y2=A*[ay2]+b1 Formula (2)

[0063] The comparison of abnormal data of the angles of the thighs and calves is shown in the figure below. Figure 2 shown.

[0064] The steps for comparing and calculating the positive and abnormal data of the relative angles of the knee joints of both legs are as follows:

[0065] Y3=A*[az1]+b2 Formula (3)

[0066] Y4=A*[az2]+b2 Formula (4)

[0067] The comparison of the positive and abnormal data of the relative angles of the knee joints of both legs is shown in the figure below. Figure 3 shown.

[0068] The steps for comparing and calculating the positive and abnormal speed data of the upper and lower leg swing are as follows:

[0069] Y5=A*[ay1]+b3 Formula (3)

[0070] Y6=A*[ay2]+b3 Formula (4)

[0071] The comparison of abnormal speed data of upper and lower leg swing is shown in the figure below: Figure 4 shown.

[0072] The steps for comparing and calculating the positive and abnormal data of knee joint force are as follows:

[0073] Normal knee force Kf1=Kw*ay1 Formula (5)

[0074] Abnormal knee force Kf2 = Kw*ay2 Formula (6)

[0075] Y7=A*[ay1]+b4 Formula (7)

[0076] Y8=A*[ay2]+b4 Formula (8)

[0077] The comparison of positive and abnormal knee joint force is shown in the figure below. Figure 5 shown.

[0078] In equations (1) to (8), A is a threshold value, and A = 0.6;

[0079] ax1, ay1 and az1 are normal values; ax2, ay2 and az2 are abnormal values; (normal values ​​or abnormal values ​​are all values ​​on the outside of the knee)

[0080] b1, b2, b3 and b4 are fixed parameters, and b1 = 0.3, b2 = 0.4, b3 = 0.5 and b4 = 0.75;

[0081] Kw is the weight of the knee; Kf1 is the normal force on the knee, and Kf2 is the abnormal force on the knee;

[0082] Y1 is the normal data value of the thigh and calf angles; Y2 is the abnormal data value of the thigh and calf angles;

[0083] Y3 is the normal data value of the relative angle of the knee joints of both legs; Y4 is the abnormal data value of the relative angle of the knee joints of both legs;

[0084] Y5 is the normal data value of the leg swing; Y6 is the abnormal data value of the leg swing;

[0085] Y7 is the normal data value of knee joint force; Y8 is the abnormal data value of knee joint force.

[0086] Before the step of comparing and calculating the positive and abnormal data of the knee joint force, there is also preparation work, which is divided into two steps, namely:

[0087] S201, calculate leg length ratio

[0088] (Height - Sitting Height) / Height

[0089] S202, calculate knee weight based on body weight and thigh-to-leg ratio

[0090] The normal size of the thigh is 0.34 times the height, and the normal proportion of the calf is 0.21 times the height. Therefore, in the present invention, the thigh proportion and the calf proportion are selected as 0.34 and 0.21 respectively.

[0091] Height is represented by Th, weight is represented by Tw, sitting height is represented by Sh, and leg length is represented by Ll, then:

[0092] Thigh length Bl = Th*0.34, thigh weight Bw = Tw*0.34;

[0093] Calf length Sl = Th*0.21, calf weight Sw = Tw*0.21;

[0094] Knee length Kl=Bl+Sl-Ll, knee weight Kw=Tw*Kl / Th.

[0095] In actual operation, depending on individual circumstances, height and weight need to be measured and manually entered into the computer for calculation.

[0096] In one embodiment of the present invention, the steps of comparing and calculating the positive and abnormal data of knee joint force are specifically as follows:

[0097] S201, calculate leg length ratio

[0098] (Height - Sitting Height) / Height

[0099] S202, calculate the knee weight based on body weight and the ratio of thigh to calf.

[0100] The thigh ratio and calf ratio are selected as 0.34 and 0.21 respectively.

[0101] Assume height Th = 180cm weight Tw = 90kg sitting height Sh = 108 leg length Ll = 72

[0102] Thigh length: Bl = 180*0.34 = 61.2 Weight Bw = 90*0.34 = 30.6

[0103] Calf length: Sl = 180 * 0.21 = 15.12 Weight Sw = 90 * 0.21 = 18.9

[0104] Knee length: Kl = 61.2 + 15.12 - 72 = 4.32 Weight Kw = 90 * 4.32 / 180 = 2.16

[0105] S303, calculate the force on the knee

[0106] The three-point acceleration obtained (i.e. the acceleration of the front, outside and inside of the knee)

[0107] Normal knee force Kf1=Kw*Ka=2.16*ay1

[0108] Abnormal knee force Kf2=Kw*Ka=2.16*ay2

[0109] Using linear regression method (where b4 = 0.75)

[0110] Y7=A*[kf1]+b4

[0111] Y8=A*[kf2]+b4

[0112] according to Figure 2-Figure 5 It can be seen that the curves of the thigh and calf angles of a normal person when walking are obviously different from those of the thigh and calf angles of a patient with deformed legs. The above precautions can assist doctors in making diagnoses and help us better manufacture medical rehabilitation knee braces.

[0113] The motion capture unit also includes a gyroscope and an acceleration sensor, which can generate angle information and acceleration information according to the motion trajectory of the motion capture unit.

[0114] The motion capture unit can obtain the inclination angles of the x, y, and z axes of the object to be measured (pitch angle Pitch, roll angle Roll, yaw angle Yaw). The microcontroller reads the six data (three-axis acceleration AD value, three-axis angular velocity AD value) of the motion capture unit and obtains the Pitch, Roll, and Yaw angles after posture fusion.

[0115] Furthermore, if Figure 6 As shown, the motion capture unit is a posture sensor, wherein:

[0116] The high level of the first motion capture unit is connected to the device address interface of the second motion capture unit;

[0117] The low level of the first motion capture unit is connected to the serial data interface of the second motion capture unit;

[0118] The clock interface of the first motion capture unit is connected to the clock interface of the second motion capture unit;

[0119] The serial data interface of the first motion capture unit is connected to the low level of the second motion capture unit;

[0120] The posture sensor of the present invention is preferably MPU6055, wherein the pins of the high level, low level, clock interface and serial data interface of the first motion capture unit are VCC, GND, SCL and SDA respectively; the device address interface, serial data interface, clock interface and low level pins of the second motion capture unit are AD0, SDA, SLC and CND respectively.

[0121] That is, the pins VCC, GND, SCL and SDA are connected to AD0, SDA, SLC and CND of the second motion capture unit respectively. The two SCLs (clock lines) in the present invention are connected to make the two MPU6055 use the same clock to achieve the purpose of clock synchronization.

[0122] Furthermore, if Figure 7 As shown, the motion capture unit is connected to the high level, low level and two protocol interfaces of the control unit.

[0123] The microcontroller is preferably ESP32, wherein the high level and low level pins of ESP32 are VCC and GND respectively, and the pins of the two protocol interfaces are IO21 and IO22 respectively. The purpose of connecting the VCC and GND of ESP32 to the two attitude sensors is to power and ground the two attitude sensors, while the two interfaces IO21 and IO22 act as SCL and SDA lines, and then control the high and low level changes of the SCL and SDA lines through artificially written software to simulate the I2C protocol. That is, the role of the two pins IO21 and IO22 of the present invention is to realize the communication of data reception and instruction sending. The present invention adopts the method of simulating the I2C protocol. Although the execution speed is not as fast as the hardware I2C and it takes a long time, the stability is worse than the hardware I2C.

[0124] like Figure 8 As shown, a motion capture method of a six-axis motion capture system comprises the following steps:

[0125] S1, the motion capture unit obtains motion data;

[0126] S2, the mobile terminal sends a request, and the control unit obtains motion data;

[0127] S3, the control unit obtains the motion data of the motion capture unit and transmits it to the mobile terminal;

[0128] S4, the mobile terminal visualizes the motion data to achieve 3D rendering and dynamic synchronization.

[0129] Furthermore, in step S1, the motion capture unit obtains the angle information and acceleration information of the motion trajectory through the gyroscope and the acceleration sensor. The motion capture unit can obtain the inclination angles (pitch angle Pitch, roll angle Roll, yaw angle Yaw) of the x, y, and z axes of the object to be measured. The microcontroller reads the six data (three-axis acceleration AD value, three-axis angular velocity AD value) of the motion capture unit and obtains the Pitch, Roll, and Yaw angles after posture fusion.

[0130] Further, in step S2, the control unit uses a multi-threaded method to dynamically and distributedly obtain the motion data generated by the motion capture unit. In the present invention, the control unit uses a polling mechanism to obtain the data of the motion capture unit, and the use of a polling mechanism can solve the problem of excessive pressure on the server. When creating a thread pool, the sub-threads for the first motion capture unit and the first motion capture unit are sub-thread A and sub-thread B, respectively, wherein sub-thread A and sub-thread B are high level and low level, respectively, in order to distinguish the two motion capture units.

[0131] Further, in step S2 and step S3, the mobile terminal and the control unit are transmitted via a wireless network. The wireless network can be broadcast using the low-power Bluetooth system (BLE) in ESP32, and each broadcast message parameter of Advertise Data.Builder is set in the broadcast packet and the scan packet respectively, where LEN and TYPE are default values ​​in the chip, and the VALUE value is specified and set according to the distributed acquisition mechanism. In the present invention, the specified setting of the VALUE value is specifically as follows Fig. 9 shown.

[0132] The order of specifying the setting of the VALUE value is immutable. The above setting can be used to quickly parse the original data and synchronize it with the corresponding points in the subsequent 3D model at one time. Step S4 can be triangulated using the KD-Tree algorithm to achieve the purpose of dynamic synchronization and secondary rendering. Among them, Kd-Tre (k-dimensionalTree) is a high-dimensional index tree data structure developed from BST (Binarysearchtree). It is also a binary tree in nature, but the data of each node is k-dimensional. It is often used for nearest neighbor search (NearestNeighbor) and approximate nearest neighbor search (ApproximateNearestNeighbor) in large-scale high-dimensional data space, such as K nearest neighbor search and matching of high-dimensional image feature vectors in image retrieval and recognition.

[0133] In the present invention, the process in step S4 is specifically as follows:

[0134] 1. Construct Kd-Tree algorithm:

[0135] (1) First, divide the space along the dividing line of the XX axis. Calculate the average value of the XX coordinates of all points, select the point closest to the average value as the dividing line, and divide the points in the space;

[0136] (2) The separated space is divided along the dividing line of the yy axis, and the division process is the same as above. After the division is completed, continue to divide along the xx axis, and so on, until there is only one point left in each space.

[0137] When constructing the Kd-Tree algorithm, the dimension should be divided so that the distribution of the values ​​of all data items in this dimension has as large a variance as possible, that is, the data is as dispersed as possible in this dimension. Such a division process corresponds to a binary tree, each division line corresponds to a branch, and each point corresponds to a leaf node.

[0138] 2. Search Algorithm

[0139] (1) According to the partition dimension and median information stored in the non-leaf nodes, start from the root node and search from top to bottom until you reach the leaf. The principle of traversal is to compare the query value and the median on the partition dimension. Let the query be Q and the currently traversed node be u. If Q>u, continue to traverse the right subtree of u. Otherwise, traverse the left subtree.

[0140] (2) After traversing to the leaf, calculate the distance between the data item with the smallest distance to the query Q in the leaf node and the query, denoted as D; then perform the "backtracking" operation, backtrack to the parent node of the current node, and determine whether the hypersphere with Q as the center and D as the radius intersects with the area represented by another branch of this parent node. If not, continue to backtrack to the previous layer; if so, continue to execute according to step (1). After exploring the bottom to the leaf node, if the distance between Q and the data item in this leaf node is smaller at this time, update D.

[0141] (3) Continue to perform the above steps (1) to (2) until the root node is traced back and both branches of the root node have been "explored".

[0142] 3. Denoising of disordered point cloud based on Kd-Tree:

[0143] (1) Generate Kd-Tree based on point cloud data and establish the topological relationship of point cloud;

[0144] (2) Find the neighborhood of any point;

[0145] (3) Calculate the distance between the point and each point in the neighborhood and take the average value;

[0146] (4) Determine whether the average value exceeds the threshold. If so, the point is determined to be a noise point and is removed.

[0147] After the system of the present invention is run, a motion trajectory trend graph is output, as shown in FIG. Fig.10 The two motion capture units of the present invention are respectively installed on the thigh support and the calf support of the common knee support. Fig.10 What is recorded mainly is the trajectory trend of rotation and displacement of the thigh and calf during movement.

[0148] The system of the present invention can realize accurate measurement of motion data, so as to facilitate analysis by doctors, thereby formulating reasonable rehabilitation strategies for patients and helping them recover as soon as possible.

[0149] The specific implementation methods described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific implementation method of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the scope of the present invention should be included in the protection scope of the present invention.

Claims

1. A six-axis motion capture system, characterized in that: It includes a motion capture unit, a control unit and a mobile terminal; the motion capture unit is used to collect motion data, the control unit is used to obtain the motion data generated by the motion capture unit, and transmit it to the mobile terminal through a wireless network; the mobile terminal processes the motion data; The motion capture unit and the control unit are connected via a hardware interface, and the control unit and the mobile terminal are connected via a wireless network; wherein the motion capture unit is two, divided into a first motion capture unit and a second motion capture unit, and the two motion capture units are connected in parallel and then in series with the control unit; The motion capture method of the above six-axis motion capture system comprises the following steps: S1, the motion capture unit generates motion data according to the motion trajectory; S2, the mobile terminal sends a request, and the control unit obtains the motion data generated by the motion capture unit; S3, the control unit obtains the motion data of the motion capture unit and transmits it to the mobile terminal; S4, the mobile terminal visualizes the motion data to achieve 3D rendering and dynamic synchronization; The process in the above step S4 is specifically as follows:

1. Construct Kd-Tree algorithm: (1) First, divide the space according to the dividing line of the XX axis; calculate the average value of the XX coordinates of all points, select the point closest to the average value as the dividing line, and divide the points in the space; (2) The separated space is divided along the dividing line of the yy axis, and the division process is the same as above; after the division is completed, continue to divide along the xx axis, and so on, until there is only one point left in each space and the division is completed; 2. Search Algorithm (1) According to the partition dimension and median information stored in the non-leaf nodes, start from the root node and search from top to bottom until you reach the leaf. The principle of traversal is to compare the query value and the median on the partition dimension. Let the query be Q and the currently traversed node be u. If Q>u, continue to traverse the right subtree of u. Otherwise, traverse the left subtree. (2) After traversing to the leaf, calculate the distance between the data item with the smallest distance to the query Q in the leaf node and the query, denoted as D; then perform the "backtracking" operation, backtrack to the parent node of the current node, and determine whether the hypersphere with Q as the center and D as the radius intersects with the area represented by another branch of this parent node; if not, continue to backtrack to the previous layer; if so, continue to execute according to step (1), and after exploring the bottom to the leaf node, if at this time Q has a smaller distance with the data item in this leaf node, update D; (3) Continue to perform the above steps (1) to (2) until the root node is traced back and both branches of the root node have been "explored"; 3. Denoising of disordered point cloud based on Kd-Tree: (1) Generate Kd-Tree based on point cloud data and establish the topological relationship of point cloud; (2) Find the neighborhood of any point; (3) Calculate the distance between the point and each point in the neighborhood and take the average value; (4) Determine whether the average value exceeds a threshold. If so, the point is determined to be a noise point and is removed; It also includes a host computer; the host computer is used to measure the difference between the curves of the thigh and calf angles of a normal person walking and the curves of the thigh and calf angles of a patient with deformed legs, and the host computer and the control unit are connected via a wireless network; The host computer includes a preprocessing unit and a data calculation unit. The preprocessing unit preprocesses the data acquired by the control unit, and the preprocessing steps are: S101, remove duplicate data; S102, if the absolute value of the difference between the current value and the previous value is greater than 0.6, then the value A before replacement is saved, otherwise it is discarded; S103, all values ​​were normalized by dividing by the maximum value; The data calculation unit's calculation contents include: comparison and calculation of positive and abnormal data of the angles of the upper and lower legs, comparison and calculation of positive and abnormal data of the relative angles of the knee joints of both legs, comparison and calculation of positive and abnormal data of the swing speed of the upper and lower legs, and comparison and calculation of positive and abnormal data of the force on the knee joints; Before the step of comparing and calculating the positive and abnormal data of the knee joint force, there is also preparation work, which is divided into two steps, namely: S201, calculate leg length ratio (Height - Sitting Height) / Height S202, calculate knee weight based on body weight and thigh-to-leg ratio Height is represented by Th, weight is represented by Tw, sitting height is represented by Sh, and leg length is represented by Ll, then: Thigh length Bl = Th*0.34, thigh weight Bw = Tw*0.34; Calf length Sl = Th*0.21, calf weight Sw = Tw*0.21; Knee length Kl=Bl+Sl-Ll, knee weight Kw=Tw*Kl / Th.

2. The six-axis motion capture system according to claim 1, characterized in that: The data calculation units are calculated using linear regression method. The steps for comparing and calculating the positive and abnormal data of the thigh and calf angles are as follows: Y1=A*[ay1]+b1 Formula (1) Y2=A*[ay2]+b1 Formula (2) The steps for comparing and calculating the positive and abnormal data of the relative angles of the knee joints of both legs are as follows: Y3=A*[az1]+b2 Formula (3) Y4=A*[az2]+b2 Formula (4) The steps for comparing and calculating the positive and abnormal speed data of the upper and lower leg swing are as follows: Y5=A*[ay1]+b3 Formula (3) Y6=A*[ay2]+b3 Formula (4) The steps for comparing and calculating the positive and abnormal data of knee joint force are as follows: Normal knee force Kf1 = Kw*ay1 (5) Abnormal knee force Kf2 = Kw*ay2 (6) Y7=A*[ay1]+b4 Formula (7) Y8=A*[ay2]+b4 Formula (8) In equations (1) to (8), A is a threshold value, and A = 0.6; ax1, ay1 and az1 are normal values; ax2, ay2 and az2 are abnormal values; b1, b2, b3 and b4 are fixed parameters, and b1 = 0.3, b2 = 0.4, b3 = 0.5 and b4 = 0.75; Kw is the weight of the knee; Kf1 is the normal force on the knee, and Kf2 is the abnormal force on the knee; Y1 is the normal data value of the angle of the thigh and calf; Y2 is the abnormal data value of the angle of the thigh and calf; Y3 is the normal data value of the relative angle of the knee joints of both legs; Y4 is the abnormal data value of the relative angle of the knee joints of both legs; Y5 is the normal data value of the leg swing; Y6 is the abnormal data value of the leg swing; Y7 is the normal data value of knee joint force; Y8 is the abnormal data value of knee joint force.

3. The six-axis motion capture system according to claim 1 or 2, characterized in that: The motion capture unit is a posture sensor, where: The high level of the first motion capture unit is connected to the device address interface of the second motion capture unit; The low level of the first motion capture unit is connected to the serial data interface of the second motion capture unit; The clock interface of the first motion capture unit is connected to the clock interface of the second motion capture unit; The serial data interface of the first motion capture unit is connected to the low level of the second motion capture unit.

4. The six-axis motion capture system according to claim 3, characterized in that: The motion capture unit is connected with the high level, low level and two protocol interfaces of the control unit.

5. The six-axis motion capture system according to claim 1 or 2, characterized in that: In step S1, the motion capture unit obtains the angle information and acceleration information of the motion trajectory through the gyroscope and the acceleration sensor.

6. The six-axis motion capture system according to claim 1 or 2, characterized in that: In step S2, the control unit dynamically and distributedly acquires the motion data generated by the motion capture unit in a multi-threaded manner.

7. The six-axis motion capture system according to claim 1 or 2, characterized in that: In step S2 and step S3, the mobile terminal and the control unit perform transmission via a wireless network.

Images