Motion Analysis Apparatus

The present invention relates to an apparatus for use in analysing the motion of a subject, in particular a force plate upon which a person stands. The apparatus has particular application in analysing human motion in sports, such as golf and baseball.

There are many situations in which the analysis of the motion of a person is desirable, for example in medicine for analysing the movement of a patient, and in a variety of training scenarios for providing feedback and instruction. One area in which motion analysis is used a great deal is in sports training and practice. In sports such as golf, the result of the shot depends considerably on the position and movement of the player before and during the swing. It is therefore not surprising that a number of systems have been developed that provide data relating to the movement of a player and/or provide analysis and feedback on that player's movement in relation to the swing. Many such systems work in conjunction with a golfing simulator for indoor practice.

Some known systems such as that disclosed in US 2007/0196800 A1 dynamically capture information concerning a golfer's body position and movement through a high resolution pressure mat under the golfer's feet. The pressure mat comprises an array of pressure sensors, and by analysing the pressure measurements from these sensors a number of features can be calculated such as the position of each foot, stance width, centre of gravity and weight distribution between the feet.

Another system disclosed in WO 2006/120658 uses two foot platforms, one for each foot, on which the player stands. Each foot platform has an arrangement of force sensors such as strain gauges, to measure the vertical and horizontal forces exerted by the player over time. A computer calculates various quantities such as the relative weight on each foot, weight transfer etc, from which characteristics such as pelvic rotation can be estimated. Two foot platforms are used so that the forces due to each foot can be distinguished. Without this information, quantities like relative weight could not be found.

The present inventors have recognised that such prior art systems have a number of disadvantages. The above system utilising a pressure mat can only provide data based on pressure measurements, and so for example the evaluation of horizontal and rotational motion, both of which are very important in the analysis of a golf swing, is not possible. Furthermore, even vertical motion is estimated indirectly from the pressure mat measurements, since there is no direct measurement of force. This introduces inaccuracies.

On the other hand, whilst direct forces can be measured using a system comprising two foot platforms, the player is not free to stand in a natural way, since he must place one foot on each platform in a certain way so that the measured forces are correctly attributed.

According to a first aspect, the present invention provides an apparatus for use in analysing the motion of a human subject comprising a platform upon which the subject stands, the platform comprising: force sensors arranged to measure force exerted on the platform by the subject in each of three dimensions; further comprising a foot position data collector arranged to collect data relating to the position of each foot of the subject on the platform; and an output arranged to output the force data and data relating to the position of each foot.

The platform is preferably in the form of a force plate and is sometimes referred to hereinafter as such.

The terms “motion of a human subject”, “motion information” and “motion characteristics” used herein include not only vector quantities that intrinsically involve a time component and thus “motion”, such as torque and rotational forces, but also those quantities which will change as the subject moves such as weight distribution and centre of pressure.

It will be appreciated that the meaning of a subject “standing” on a platform includes the subject moving about whilst situated on the platform, e.g. by leaning to one side, applying a torque through the feet etc. Furthermore the forces “exerted” by the subject encompasses all forces such as forces due to gravity, torque specifically applied by the subject etc.

Measuring force in each of three dimensions means for example measuring force in each of three perpendicular directions, commonly two perpendicular directions in an approximately horizontal plane, and a third direction perpendicular to the first two directions, in an approximately vertical plane.

To be able to calculate certain types of motion information, both measurements of the forces exerted by the subject and the position of the feet are required. Thus, since the force plate according to the invention collects data relating to both of these, it enables such types of motion information to be accurately calculated. Changes in balance and weight distribution over time, and body rotation, are types of motion information that require both force and foot position measurements, and thus the present invention enables them to be calculated.

Preferably, the data relating to the position of each foot is used in conjunction with data relating to the typical shape of a foot to determine the orientation of each foot (i.e. which direction each is facing). Such information is useful in calculating motion information.

The foot position data collector can take any suitable form. For example it may comprise visual/infrared markers on the feet and an associated camera. It may comprise capacitive proximity sensors. It may alternatively comprise an optical grid located just above the force plate surface, wherein each line in the grid detects whether a foot is blocking it. Technology used in touch-sensitive monitors may be used.

In one particularly preferred embodiment the foot position data collecting means comprises a pressure sensitive mat. Any suitable type of pressure mat may be used which collects pressure data relating to the position of the feet. In one preferred embodiment the pressure mat comprises a grid of force sensitive resistors (FSRs). Force sensitive resistors change resistivity when pressed down and are commercially available. In one embodiment, the mat comprises 1000 FSRs on a mat of 800 mm×500 mm, resulting in a sensor spacing of 20 mm. In another embodiment, a higher resolution is provided using a mat comprising 2048 FSRs resulting in a sensor spacing of approximately 14 to 15 mm. In a further embodiment the pressure mat comprises piezoelectric sensors which generate voltage when pressure is exerted on them.

In one embodiment it is necessary to determine only the outline position of the feet and not the amount of pressure exerted by the feet. Thus, the mat can be inexpensive and easy to implement compared with pressure mats used in the prior art because less sensors are needed and less processing, is required.

In an alternative embodiment, data collected by the pressure mat can be used to find the pressure distribution for each foot, which can be used to evaluate the balance of the subject. For example, it can be used to determine if the subject is standing flat on his feet or if any part of the foot (heel, toes, left/right side) is partially or fully lifted from the ground. This information can be visualised on a screen.

A conventional three dimensional cartesian co-ordinate system is used in this application, comprising three mutually perpendicular X, Y and Z axes. In the following it is generally considered that the X and Y axes are perpendicular to each other in the horizontal plane, and the Z axis is perpendicular to the X and Y axes in the vertical plane.

The force sensors are arranged to measure force in three dimensions. Preferably, for optimal measurements, force sensors are provided that align with each axis/direction. For example at least one X force sensor may be aligned with the X axis, at least one Y force sensor may be aligned with the Y axis, and at least one Z force sensor may be aligned with the Z axis. Alternatively, force sensors may not be aligned with the axes, but may simply be arranged to measure forces from which X, Y and Z components can be calculated.

The combination of three dimensional force information in particular, together with foot position data, provides considerable new possibilities in the analysis of human motion. For example, features relating to body rotation such as torque, momentum and rotational strength can be evaluated in all three dimensions which is simply not possible with prior art systems.

In the prior art as discussed above, two force plates (one for each foot) are needed in order to distinguish between the forces of each foot. In contrast, in the present invention, a single foot platform/force plate can be used for both feet, since a pressure sensitive mat is used to collect data relating to the foot positions. This pressure data enables the forces to be correctly attributed to each foot. Using only a single foot platform/force plate simplifies manufacture and allows the subject to stand naturally on the force plate rather than having to place one foot on each force plate in a contrived manner as necessary in the prior art. Importantly, using only one force plate reduces cost since only one set of sensors is required to analyse both feet.

The apparatus of the present invention also enables motion to be analysed much more accurately due to measurement of both force and pressure data.

In one particularly preferred embodiment the apparatus is arranged for use in analysing the motion of a subject during a golf swing, though it has a wider application both in sports (for example in analysing motion during a baseball swing) and more generally for example in analysing the motion of patients in a medical context.

The force sensors can be any suitable sensors that measure force such as strain gauge load transducers. As discussed above, to enable a wide range of motion information to be calculated, sensors are provided to measure forces in the X, Y and Z directions. Preferably the platform/force plate is rectangular. If it is considered to have a front and a back, and the front is taken as the side to which the toes are directed, the back being taken to be the opposite side (i.e. the side to which the heels are directed) the X direction is considered here as the direction perpendicular to the direction the feet face when placed on the force plate (this is shown in the example of FIG. 1). The Y direction is considered here as the direction parallel to the direction the feet face and the Z direction is the vertical direction (again shown in FIG. 1).

In a preferred embodiment, two X sensors are provided (although in other embodiments more or less are provided), which are strategically placed in order to enable the most accurate measurements of horizontal forces in the X direction.

Preferably one X sensor is placed towards the front and one towards the back of the force plate. Preferably both sensors are centred halfway along the force plate in the X direction. As will be seen in more detail below, the X sensor data is particularly useful in analysing body rotation around the Z axis. The torque exerted on the force plate by the subject turning themselves clockwise/anticlockwise while the feet remain stationary, i.e. the torque around the Z axis, is equal to the cross product of force exerted by the feet around the Z axis and distance from the point about which the force causes rotation, i.e. the centre of rotation of the feet. Since the force is measured at the position of the X sensors, each X sensor should be located at approximately the same distance from the centre of rotation as it can be expected the average subject's feet will be placed. Thus, each X sensor should be placed along the Y axis at approximately the same distance from the centre of rotation as each foot, in order for torque to be accurately estimated using the X sensors.

Although the centre of rotation will vary depending on the position of the feet, since the X sensors will generally need to be fixed in place on the force plate prior to use, it is necessary for this purpose to pre-estimate the centre of rotation. Here, the centre of rotation is assumed to be approximately the force plate centre, thus the distance along the Y axis from the force plate centre to each X sensor should be approximately equal to the distance from the centre of the force plate to each foot. This enables as close to a 1:1 mapping as possible between measured forces and the actual measured forces when evaluating rotation around the Z axis, described later. The phrase “spin kick” is sometimes used in this application to refer to Z-axis torque, particular when this is discussed in a golfing context. This “spin kick” describes the force travelling in a circular path, with a diameter equal to the stance width. If the subject stands perfectly centred on the force plate, this “circle of force” passes approximately straight through both X-sensors parallel to their measuring direction, i.e. the x-axis. Therefore preferably the subject stands centred on the force plate.

Thus it can be understood that in a preferred embodiment, the X sensors are arranged towards the front and back of the force place respectively at roughly the same distance from the centre as which the feet are located on each side. In the case of use of the force plate in analysing golf swings, the width of an average golf stance (distance between feet when hitting a golf ball) is about 50 cm, thus the distance from each foot to the centre of rotation is approximately 25 cm. Each X sensor is therefore positioned in the Y direction approximately 25 cm from the force plate centre. This enables the “spin kick” to be measured more accurately.

In one preferred embodiment one Y sensor to measure forces in the Y direction is provided towards the centre of the force plate; most preferably in the centre. In an alternative embodiment two or more Y sensors are provided. For example two Y sensors could be provided towards the left and right sides of the force plate in the same manner as the X sensors are provided towards the front and back. Thus preferably, each of two Y sensors would be arranged halfway along the force plate in the Y direction and approximately 25 cm in the X direction from the force plate centre.

In one preferred embodiment four Z sensors are provided to measure forces in the vertical Z direction. These force sensors are preferably spaced out around the force plate, in the case of a rectangular force plate they are most preferably located in each of the four corners.

The force plate preferably includes a base plate on which the force sensor(s) are mounted, a cover plate disposed over the base plate and the pressure mat disposed over the cover plate. Preferably the cover plate is arranged to move freely on the force sensors, such that its movement is only restricted by the sensors and thus the forces absorbed by the mounting structure are minimised. In the case of golf swing analysis, preferably an artificial turf mat is arranged on top of the pressure mat for the subject to stand on. This provides the subject with a more natural environment. Preferably, the force plate further comprises electronics arranged to control the force sensors and pressure mat, collect data from them and output this data. This data is then preferably input to a processing means, for example a computer, which is arranged to process the data and calculate information relating to the motion of the subject.

Accordingly, the invention also provides a system for analysing the motion of a human subject comprising a force plate as described above, a processing means arranged to determine some information relating to the motion of the subject using the measured forces in three dimensions and/or data relating to the position of each foot, and an output to output the motion information. Preferably the information is determined using both the measured forces and foot position data.

Generally, when the foot position data collecting means comprises a pressure sensitive mat, the processing means is arranged to estimate the outline of each foot on the force plate from the pressure mat data and use this together the measured force in three dimensions to calculate relevant motion information such as rotational force of the subject or balance. Details of types of motion information which are particularly useful to analyse in a golfing context are described below in Tables 1 and 2. The torque exerted by the feet around the Z axis (denoted elsewhere in a golfing context as “spin kick”) can be found using the X sensor force measurements and the estimated foot positions according to the method as described later in relation to FIGS. 5 and 6. The body rotation around the Y and X axes can be found using the X and Y sensor data respectively together with the vertical Z sensor data and the position of the feet. For this rotation the body is seen as a sphere with a diameter equal to the subject's height. A preferred method is described later in relation to FIGS. 7-10.

The processing means may be external to the force plate, integral with it, or partly external and partly integrated. For example, the force plate may include electronics arranged to estimate the outline of each foot from the pressure mat data and to output this together with the force measurements to an external processing means which then performs further calculations.

The processing means preferably comprises a computer having software which when executed by the computer causes the computer to determine the motion information. Preferably the software includes a user interface, most preferably a graphical user interface, by which a person can interact with the system. Preferably the output is a monitor which displays the calculated motion information, which in one preferred embodiment is a touch-screen monitor.

In another aspect the invention provides a software product carrying instructions which when carried out by a data processing apparatus will configure the data processing apparatus to determine information relating to the motion of a subject from measured forces exerted by the feet of the subject on a platform in each of three dimensions and data relating to the position of each foot of the subject.

The computer program product can be in the form of a physical carrier bearing software, or in the form of signals transmitted from a remote location. The computer program product may also be described as a software product.

In a further aspect the invention provides a method of manufacturing a software product which is in the form of a physical data carrier, comprising storing on the data carrier instructions which when carried out by a data processing apparatus will configure the data processing apparatus to determine information relating to the motion of a subject from measured forces exerted by the subject on a platform in each of three dimensions and data relating to the position of each foot of the subject.

In a still further aspect the invention provides a method of providing a computer program product to a remote location by means of transmitting data to a data processing apparatus at that remote location, the data comprising instructions which when carried out by the data processing apparatus will configure the data processing apparatus to determine information relating to the motion of a subject from measured force exerted by the subject on a platform in each of three dimensions and data relating to the position of each foot of the subject.

The data processing apparatus can be a computer.

Various features described previously and subsequently are applicable to such software products, for example in one preferred embodiment the information relating to the motion of the subject is torque or angular momentum around the X, Y and/or Z axes.

The system may include further components such as a video camera arranged to record video of the movement of the subject, which is then displayed along with calculated motion information on the monitor. More than one video camera may be provided in order to take video at different angles, in one preferred embodiment four video cameras are provided.

In one embodiment, the system is arranged to analyse the motion of a human subject during a golf swing. In this context the system has particular application as a golf training and practice tool. The software is preferably arranged to provide analysis of and feedback on the measured motion characteristics. The system may be used as a practice tool by a student alone (whose motion is analysed), or with an instructor.

In this context, other components may be included for providing information relevant to a golf swing, such as a sensor capable of measuring aspects of the motion of the golf club. Suitable sensors include doppler radar sensors, ultrasound positioning sensors, laser technology and high speed cameras. A sensor may also be included which is capable of measuring aspects of the ball motion such as ball speed and vertical launch angle. Suitable sensors include doppler radar, laser technology and high-speed cameras. In a preferred embodiment a TrackMan™ (http://trackmangolf.com/) launch monitor is used which evaluates the motion of the golf club and ball using doppler radar.

Sensors for measuring specific aspects of the body motion may also be included, for example a motion-tracking suit or motion capture device.

A trigger sensor may also be included which reacts to a specific event in the motion of the subject, such as the striking of the ball. For example, a microphone may be included and obtained sound information used to determine the strike of the ball from the distinct sound it makes. Alternatively, any ball or club motion sensor may be used, for example one utilising doppler radar.

The force plate may be integrated into a tee unit which tees the ball. A commercially available tee unit may be used, or alternatively a bespoke tee unit may be provided. Preferably the force plate is integrated into the tee unit so that the turf mat of the force plate is flush with the turf of the tee unit.

It will be understood that whilst these additional components are described in the context of golf swings, they may also be applicable in other sports and activities.

In a broader aspect, the present invention provides a system for analysing the motion of a human subject, comprising force sensors arranged to measure force exerted by the subject in each of three dimensions; a foot position data collecting means arranged to collect data relating to the position of each foot; and a processor arranged to determine information relating to the motion of the subject using the measured forces and/or the position data of each foot. Preferably the motion information is determined using both force and foot position data. The foot position data collecting means can be any suitable means for collecting data relating to the individual foot positions including orientation and outline. However in a preferred embodiment the foot position data collecting means is a pressure mat as described previously. Most preferably the foot position data collecting means is a pressure mat and is integrated with the force sensor into a force plate as previously described.

Viewed from another aspect, the present invention provides a golf practice system comprising: a force plate on which a subject stands whilst practising a golf swing comprising force sensors arranged to measure forces exerted by the subject in each of three dimensions (i.e. in the X, Y and Z directions) and a foot position data collecting means arranged to collect data relating to the position of the each foot of the subject; a processor arranged to receive the force data and data relating to the location of each foot from the force plate and to evaluate motion characteristics of the subject from both force and foot position data; and a display arranged to display the calculated motion characteristics.

The invention also extends to a method of analysing the motion of the human subject comprising the use of any of the above described apparatuses or systems.

In a further aspect the invention provides a method for use in analysing the motion of a human subject comprising: measuring force exerted on a platform in each of three dimensions by a subject standing on the platform; collecting data relating to the position of each foot of the subject on the platform; and outputting the measured force data and data relating to the position of each foot.

In yet another aspect the invention provides a method of analysing the motion of a human subject comprising: measuring force exerted on a platform in each of three dimensions by the subject standing on the platform; collecting data relating to the position of each foot of the subject on the platform; evaluating some information relating to the motion of the subject using both the measured forces and data relating to the position of each foot; and outputting the motion information.

In another aspect of the invention, force is measured in only one or two dimensions, e.g. in X, Y or Z directions, or two of these directions. This force information together with data relating to the position of the feet enables many useful motion characteristics to be measured. Thus, the invention also provides an apparatus for use in analysing the motion of a human subject comprising a platform upon which the subject stands, the platform comprising:

at least one force sensor arranged to measure forces exerted on the platform by the subject; a foot position data collector to collect data relating to the position of each foot of the subject on the platform; and an output arranged to output the force measurements and data relating to the position of each foot. The invention also provides a system for analysing the motion of a human subject, comprising at least one force sensor arranged to measure forces exerted by the subject; a foot position data collector arranged to collect data relating to the position of each foot of the subject; and a processor arranged to determine information relating to the motion of the subject using both measured forces and position data of each foot.

The invention further provides a method for use in analysing the motion of a human subject comprising: measuring forces exerted on a platform by a subject standing on the platform; collecting data relating to the position of each foot of the subject on the platform; and outputting the measured force and pressure data. The invention also provides a software product comprising instructions which when executed by a computer cause the computer to determine information relating to the motion of a subject from measured forces exerted by the subject on a platform and data relating to the position of each foot of the subject.

Preferably the foot position data collector comprises a pressure mat as previously described.

It will be understood that preferred features described in relation to certain aspects and embodiments may also be applicable to other aspects and embodiments and vice versa.

Preferred embodiments of the present invention will now be described by way of example only and with reference to the accompanying drawings, in which:

FIG. 1 illustrates a force plate according to an embodiment of the invention;

FIG. 2 illustrates a system comprising a force plate and a computer according to an embodiment of the invention;

FIG. 3 shows a user standing on a force plate according to an embodiment of the invention;

FIG. 4 is a flow diagram illustrating the processing performed in order to analyse the user's motion, according to an embodiment of the invention;

FIG. 5 illustrates a force plate according to an embodiment of the invention;

FIG. 6 illustrates the forces on an X sensor according to an embodiment of the invention;

FIG. 7
a is a front schematic view of a golfer in a golfing stance prior to starting a golf swing;

FIG. 7
b is a side schematic view of a golfer in a golfing stance prior to starting a golf swing;

FIG. 8 is a schematic view of a golfer performing a golf swing;

FIG. 9 illustrates the calculation of the horizontal torque component;

FIG. 10 illustrates the calculation of the vertical torque component;

FIG. 11 is a screen capture of the welcome screen displayed by example software according to an embodiment of the invention;

FIG. 12 is a screen capture of the main menu displayed by example software according to an embodiment of the invention;

FIG. 13 is a screen capture of the information displayed during swing analysis by example software according to an embodiment of the invention;

FIG. 14 is a screen capture of the information displayed during two swing comparison by example software according to an embodiment of the invention; and

FIG. 15 is a screen capture of the swing library displayed by example software according to an embodiment of the invention.

A force plate 1 comprises a base plate 2 for the mounting of force sensors 11a, 11b, 12, 13, and a signal conditioner 14; a cover plate 3 disposed over the base plate 2; a pressure mat 4 disposed over the cover plate; and a turf mat 5 arranged on the top of the pressure mat 4.

A total of seven force sensors are mounted on the base plate 2: two X sensors 11a and 11b arranged to measure horizontal forces in the X direction; one Y sensor 12 arranged to measure horizontal forces in the Y direction; and four Z sensors 13 one at each corner of the base plate 2 arranged to measure vertical forces. The sensors are all commercially available strain gauge load transducers. The sensors are connected to a commercially available signal conditioner 14 for strain gauge sensors which delivers an excitation voltage to the sensors, has a programmable sample rate and comprises an analog-to-digital converter. The X sensors are placed towards the front and back of the force plate, each approximately 24 cm from the force plate centre in the Y direction.

The pressure mat 4 comprises a grid of force sensing resistors (FSRs). The number of FSRs can be chosen according to the desired resolution. In this embodiment, the pressure mat 4 comprises 32 rows and 64 columns of FSRs=2048 FSRs in total, on a mat of 800 mm×500 mm, resulting in a sensor spacing of approximately 14-15 mm.

Two multiplexers 23A and 23B and microcontroller 24 are provided inside the force plate 1 (not shown in FIG. 1). The FSRs are connected to the multiplexers 23A and 2313 and the multiplexers are connected to the microcontroller 24.

A system 20 according to an embodiment of the invention is shown in FIGS. 2 and 3. The force plate 1 is integrated into a tee unit 27 for teeing the ball, and is connected to a computer 22 via a USB cable 30 (in an alternative embodiment wireless communication can be used). The outputs from the signal conditioner 14 (i.e. information from the force sensors) and the microcontroller 24 (i.e. information from the pressure mat) are both input to a USB hub, which outputs the signals via USB cable 30 to the computer 22. The microcontroller controls the sampling, the multiplexer and converts signals from the pressure mat from analogue to digital.

The computer comprises software that is arranged to evaluate a variety of different types of information (denoted herein as “attributes”) using these outputs, as will further be described below. The computer includes a touch-sensitive monitor (not shown) via which the information can be displayed and a person can interact with the system.

The system 20 further includes a microphone 25 and four video cameras 28A, 28B, 28C and 28D connected to the computer (only one video camera is illustrated in FIG. 2 for clarity). The video cameras are positioned to collect video data of the golf swing from different viewpoints. Camera 28A is positioned along the target line, viewing the shot direction; camera 28B is positioned face on, viewing the front of the subject; camera 28C is positioned behind the subject, viewing the back of the subject; and camera 28D is mounted in the ceiling looking down on the subject. A TrackMan™ (http://trackmangolf.com/) device 26 is also connected to the computer. This evaluates the motion of the golf club and ball using doppler radar (in alternative embodiments other known motion sensors/launch monitors are used).

The system is intended to be used either by a student alone or the student and his instructor. In the former case the student will operate the system (i.e. turn it on, interact with the computer etc), the system being an independent-learning/practice tool. In the latter case it is intended that the instructor will operate the system, the system being a teaching tool to assist the instructor in coaching his student. In the embodiment described herein, the computer software has various “modes” to suit different types of practice/teaching. The term “student” used herein means the person who stands on the force plate and hits the ball.

In use, a student (or his instructor) turns on the system 20 and initialises it for use using the touch-sensitive monitor. As illustrated in FIG. 3, the student stands on the top of the force plate 1, i.e. on the turf mat 5. The tee unit 27 tees the ball, and the user takes a swing. The seven sensors 11a, 11b, 12 and 13 measure the ground reaction forces of the student's feet 21 in the X, Y and vertical (Z) directions respectively and output the seven resulting analog signals to the signal conditioner 14. FIG. 4 is a flow diagram illustrating the information flow. The signal conditioner 14 converts the analog signals from the sensors into digital values. These are then output to the computer 22 via the USB hub 31.

The microcontroller 24 decides which grid point to be read from the pressure mat 4 and sends a control signal to the multiplexers 23A and 23B. It also provides a 5V output to multiplexer 23A. Multiplexer 23A sequentially applies 5 volts on the 32 rows and B sequentially reads the 64 columns of the pressure mat. This enables the pressure at a single grid point to be measured at a particular time, by choosing a particular row and column.

The multiplexer 23B multiplexes the analogue grid point readings (one for each of the 2048 FSRs) into one analog signal and outputs this to the microcontroller 24. The microcontroller 24 performs analogue-digital conversion and outputs the digital pressure information, 2048 digital values, to the computer 22 via the USB hub 31. Data collected by the TrackMan™ device 27 relating to the ball and golf club motion, and video data from the video cameras 28A, B, C and D are also output to the computer.

The computer software “listens” to the output of the microphone device 25 and assesses the time of ball strike (impact) when it “hears” a noise which meets characteristics predetermined to indicate a ball strike. This is used to estimate the start and end time of the golf swing, by assuming that a golf swing should last 5 seconds in total: 3 seconds before the ball strike and 2 seconds after. These estimated times allow the sensor data to be matched with the ball strike and enable video clips to be generated that are synchronised on the ball strike.

The computer 22 stores the received data. The computer is loaded with software which is programmed to take the outputs from the signal conditioner 14 and microcontroller 24 and calculate a variety of data (attributes) related to the motion of the user. In particular, the computer estimates the position of each foot by analysing the pressure readings and determining which areas of the mat 4 are pressed down by a foot. If two islands of areas are identified each having a minimum area, they are identified as footprints (in a different embodiment the microcontroller performs this computation itself and outputs it to the computer).

Table 1 provides details of attributes calculated using the output of the pressure mat 4, whilst Table 2 provides details of attributes calculated using the output of the force sensors 11a, 11b, 12, 13, in some cases in combination with the output of the pressure mat. Conventional “phase” terminology for golf swings is used as follows:

1st phase=address (standing still)

2nd phase=take away (move club away from ball)

3rd phase=down swing (acceleration towards ball)

4th phase=follow through (short period after impact)

5th phase=relax (deceleration of swing)

6th phase=final pose (standing still after completing a swing)

The term “Spin kick” is used to express the forces related to rotating the hip during the golf swing. This is found by estimating the torque exerted by the feet around the Z axis since this is strongly related to the force exerted by rotating the hip during a golf swing. A method of determining the spin kick magnitude is described in detail after the Tables.

TABLE 1

Measurement

Attribute
Unit & type
Description
Algorithm details

Left
Footprint
The position and
Estimate outline

Footprint
info
outline of the left
by measuring FSR

Continuous
foot, relative to the
outputs from pressure

(footprints
top surface of the
mat and determining

measured
force plate
which area of mat is

continuously

pressed down by foot.

during

If two islands of areas

swing)

can be identified each

having a minimum

area each, they are

identified as the left

and right footprints.

An estimation

algorithm can be used

to enhance accuracy

by comparing a typical

footprint with the

measured footprint. For

example some edges of

the islands can be

ignored if it can be

identified that the foot

probably does not

cover the entire area of

that sensor spot (which

will be the case for

many of the sensors

along the island edges).

Right
Footprint
Position and outline
As above for left

footprint
info
of the right foot,
footprint

Continuous
relative to the force

plate top surface.

Stance
Centimetres
How wide the feet
Calculate distance on

width
Single value
are placed apart,
the X-axis between the

measured from the
centre of each footprint

centre of each
during the address

footprint.
phase.

Stance
Centimetres
Deviation in
Calculate distance on

depth

position of left and
the Y-axis between the

right foot in the
centre of each footprint

heel-toe direction.
during the address

phase.

Stance
Degrees
An angle describing
Use trigonometry to

angle
Single value
how open or closed
find the angle between

the stance is in
the line formed between

relation to the shot
the centre of each

direction.
measured footprint,

A stance of zero
and the aiming line of

degrees is a
the shot direction. The

perfectly neutral
shot direction is

stance, with both
assumed to be parallel

feet parallel to the
with the tee unit, i.e.

force plate.
along the X axis of

the force plate.

TABLE 2

Measurement

Attribute
unit & type
Description
Algorithm details

Balance
%
Where the balance
Compare the centre

Continuous
between the heel
of pressure (COP)

and the toe is during
measured by the

the swing.
force plate to the

Ranges from 100%
footprint positions

on the heel to 100%
found from the

on the toe.
pressure mat, and

0% means the
determine where

balance is at the
the COP lies in

centre of the foot.
relation to the

heel and toe.

Only determine

sum of balance as

no need to

distinguish

between the feet.

Centre of
Position
Represents user's
Calculate COP

Pressure
Continuous
weight position in
using the forces

(COP)

the horizontal plane
measured by the

four Z sensors and

the dimensions of

the force plate.

Calculate in

standard way

using linear

algebra.

Weight
%
How the weight is
Compare COP

distribution
Continuous
distributed between
measured by the

the left and right
force plate to the

feet during the
footprint positions

swing.
found from the

Ranges from 100%
pressure mat, and

on left foot to 100%
determine where

on right foot.
the COP lies

between the left

and right feet.

Rotational
Newton
The sum of all
Sum measurements

strength
Single value
rotational forces
for a continuous

per time
during the golf
rotational attribute

interval
swing for a
such as spin kick

particular rotational
or total body

motion, e.g. spin
rotation for the

kick. This sum
duration of the

represents how
interval.

much power the

user generates in

the rotational

movement.

Forward/
Newton
The sum of forces
Sum absolute

backwards
Single value
exerted by shifting
value (so negative

shift
per time
the feet forwards or
values contribute

interval
backwards during
positively) of forces

the golf swing
exerted on the Y

sensor for the

duration of the

interval.

Sideways
Newton
The sum of forces
Sum absolute

shift
Single value
exerted by shifting
value (so negative

per time
the feet sideways
values contribute

interval
during the golf
positively) of forces

swing.
exerted on the X

sensors for the

duration of the

interval.

Overall
%
The degree of
All continuous

smoothness
Single value
smoothness in the
motion attributes

per time
golf swing motion.
v. time such as COP,

interval.
A rushed back
spin kick, shifts

Higher % =
swing or a swing
and rotational forces

better
plane that demands
can be used as a

smoothness
a lot of
base for

compensation with
evaluating the

hand-power will
smoothness of

produce a more
the golf swing.

jerky motion, and
By differentiating

this will be reflected
the chosen data

in the overall
with respect to time

smoothness
in the 2nd and 3rd

attribute.
order, the speed and

acceleration of the

rate of change can

be found at a given

time.

A number of

methods can be

used to measure

how smooth the

change in data is.

One method is to

find the average

absolute

acceleration for

the time interval

and rate a lower

average with a

higher percentage.

Balance at
%
Degree of balance
Evaluate stability

end of
Single value
maintained at the
of COP for 5th

swing

end of the swing.
phase and express

The lower the %, the
as %

higher the rate and

magnitude of

variation in

movement, i.e. if

losing balance or

ending swing with

sudden jerk.

Rhythm
Fraction
Compares the
Calculate rhythm

Single value
duration of the
R as

takeaway and the
R = (Tta/Tds)

down swing
Tta = duration of

phase 2

(takeaway).

Identify start of

swing motion

and top of back

swing by

evaluating motion

characteristics.

Tds = duration of

phase 3 (down

swing). Identify top

of back swing as

above and time of

impact using

microphone.

Vertical
Newton
Difference between
Swing strength =

swing
Single value
the maximum and
VFmax −

power

minimum vertical
VFmin

force. The min
Where VF is the

force is generated
net sum of the

during the down
four Z sensors at

swing and the max
a particular time

force is generated
after substracting

just before impact.
the person's weight

Tempo
milliseconds
Time from the start
Tempo =

Single value
of the takeaway to
(T_imp− T_start)

the impact
T_imp= time of

impact (via

microphone)

T_start= time of

takeaway (find

by analysing force

plate data to

identify when the

user is moving

according to the

start of a golf

swing.

Torque
Newton
Torque generated by
Described

around
metre
body around Y and
below

Y and
Continuous
X axes, caused by

X axes,

leaning in a

direction during the

golf swing

Angular
Newton
Angular momentum
Described

momentum
metre
generated by body
below

seconds
around Y & X axes,
Angular

Single value
caused by leaning in
momentum

per time
a direction for the
is torque

interval
duration of the time
integrated

interval.
over a time

interval.

Spin kick
milliseconds
Timing of the spin
Timing =

turn timing
Single value
kick turning point
(T_imp− T_sk)

(i.e. where user
T_sk= time of

shifts from
spin kick

accelerating to
turning point

decelerating) in
(found by

relation to impact
analysing

sensor data).

Maximum
Newton
The maximum
Find maximum

spin kick
Single value
measured spin kick
peak of spin

magnitude

during the swing
kick

Maximum
milliseconds
Timing of maximum
Timing =

spin kick
Single value
spin kick (i.e.
(T_imp− T_skm)

timing

maximum spin kick
T_skm= time of

acceleration towards
max spin kick

the ball) in relation
acceleration in

to impact
relation to ball

impact (found

by analysing

sensor data)

Minimum
As for
As for maximum
As for

spin kick
maximum
This shows when
maximum

magnitude

maximum

and timing

deceleration of hip

rotation happens and

how great the

deceleration is.

Top pause
milliseconds
The duration of the
Find negative

Single value
turning point
peak of total

between the back
vertical force

swing and the start
at the transition

of the down swing
between take

away and down

swing phases.

Top pause is

the duration of

the approximately

flat part of the

peak.

Weight of
Kg
The measured
Average total

user
Single value
weight of the user
vertical force

over time/

gravitational

acceleration

Where the measurement is given as a percentage (except for balance and weight distribution), 100% means the measurement is perfectly close to a pre-defined ideal, whilst 0% is outside a pre-defined reasonable limit. Thus the higher the percentage, the better that aspect of the swing.

A method of determining the “spin kick” according to a preferred embodiment of the invention is now described in relation to FIGS. 5 and 6. FIG. 5 illustrates the feet on the force plate 1. The torque exerted by the feet around the Z axis (in the XY plane) is:

Torque=rotational force F×distance to centre of rotation D

The centre of rotation CoR is assumed to be the midpoint between the feet, which can be determined in a straightforward manner from the foot positions found using the pressure mat 4. As the position (e.g. the x and y coordinates) of the X sensors 11a and 11b are known, the distances Dx and Dy in the x and y directions respectively from each X sensor to the centre of rotation (i.e. the x and y components of D) can be calculated. These distances are illustrated in FIG. 5 for sensor 11a. The direct distance D from X sensor 11a to the centre of rotation CoR can then be found for example by using Pythagoras:

D=√(Dx²+Dy²)

The rotational force F is then found. The approximate angle at which the force passes through the sensor is φ1 as can be seen in FIGS. 5 and 6. φ2 is geometrically related to φ2, since the force vector F is perpendicular to distance D. Therefore:

φ1=90°−φ2

- where: tan φ2=Dy/Dx
- thus: φ2=tan⁻¹(Dy/Dx) Dy and Dx known see above
  
  It can be seen from FIG. 6 that:

F≈Fx_a/cos φ1

Therefore F≈Fx_a/cos(90°−(Dy/Dx)))

- Where:
  - Dy and Dx are known see above
  - Fx_ais the force measured by the X sensor 11a
    
    Thus the torque measured using X sensor 11a is:

Torque=Fx_a/cos(90°−(tan⁻¹(Dy/Dx)))×√(Dx²+Dy²)

Corresponding calculations are then performed to find the torque for the other X sensor 11b, and the two values are summed. The “spin kick” is considered to be this total torque exerted on the X sensors. The spin kick will vary over the course of the swing and thus it is measured at various time intervals and a curve of spin kick v. time is obtained.

A method of determining the torque/angular momentum around the Y and X axes (in the XZ and YZ planes) according to a preferred embodiment of the invention is now described in relation to FIGS. 7 to 10.

The angular momentum of a subject can be approximated as the motion of leaning in a direction while standing stationary with the feet. Consider the person as a sphere with a diameter equal to the person's height. The angular momentum is the pace at which the sphere “rolls” when leaning in a direction. Angular momentum around the Y axis is caused by leaning left or right, while in the X axis it is caused by leaning forwards or backwards. FIGS. 7a and 7b illustrate a golfer in this co-ordinate system preparing to take a swing, whilst FIG. 8 illustrates the golfer during the swing in relation to the notional sphere.

A method for calculating the angular momentum around the Y and X axes will now be described. Let:

Fz₁(t), Fz₂(t), Fz₃(t), Fz₄(t)=the four Z sensors (measuring vertical forces), placed towards each corner of the rectangular force plate

Fz(t)=sum of forces measured by all four Z sensors at time t

Fx(t)=sum of forces measured by the two X-sensors at time t (note the X sensors are mounted so that they measure positively when clockwise torque is applied to the plate. This means X sensor 11a is positive in the positive X axis, and 11b is negative in the positive X axis).

Fy(t)=force measured by the single horizontal Y-sensor at time t.

- (note that the sensors are arranged to absorb the entire force, so the sum of the forces is the total force exerted. If more sensors were used, each sensor would absorb less force and so the sum would be substantially the same).
  
  COM(t)=Centre of Mass, the centre point of the body's mass in space. Typically this is at the centre of a person. When a person is standing in a golfing stance, the centre of the person is approximately at half the person's height, in the region of the person's bellybutton. Calculation of the COM is described later. The COM has x, y and z components.
  
  COP(t)=Centre of Pressure, the point on the force plate that constitutes the centre of the vertical forces. When standing still, this is typically at the midpoint between the feet, see FIGS. 7a and 7b. Calculation of the COP is described later.
  
  FZ_NET=net vertical force, when ignoring the constant force of gravity caused by the weight of the person, i.e. Fz₁(t)+Fz₂(t)+Fz₃(t)+Fz₄(t)−F_WEIGHT
  
  F_WEIGHT=vertical force caused by gravity and the weight of the person, i.e. Fz₁(t0)+Fz₂(t0)+Fz₃(t0)+Fa₄(t0) when the person is standing still.

Weight (Kg)=F_WEIGHT/g [g=acceleration due to gravity, approx. 9.81 ms²]

It is well known that:

Torque=rotational force F×perpendicular distance from the pivot to the line of action of the force (torque arm)

To find the torque around the Y axis, the X (horizontal) and Z (vertical) torque components need to be found. Let:

X torque component=τ_x(t)

Z torque component=τ_z(t)

Considering first the horizontal torque component, this is generated by exerting horizontal forces Fx with the feet which are measured by the X sensors. The centre of rotation COR can be approximated as the centre of mass COM as illustrated in FIG. 9. The perpendicular distance r_xfrom the pivot to the line of action of the force Fx is therefore the vertical distance from the COM to the plane on which the X sensors are positioned, r_x, (sometimes denoted the “torque arm”).

Thus:

τ_x(t)=Fx(t)×COM_z(t)

where: COM_z(t)=position of COM in Z direction

Regarding the vertical torque component, this is generated due to the force of gravity when the person leans to one side, resulting in a deviation between the COM and COP. As above, the centre of rotation can be assumed as being the COM of the person at half the person's height. When standing still (in other words when the mechanical system is in full static equilibrium), the horizontal position of COM and COP is approximately equal, as can be seen in FIGS. 7a and 7b. However when the person starts moving around (while the feet remain stationary, such as in a golf swing) as shown in FIG. 8 the COM and COP move, in particular the COP (i.e. the centre of the vertical forces measured by the Z sensors) moves to one side. As can be seen from FIG. 10, the perpendicular distance r_z(vertical torque arm) from the COM to the line of action of the vertical forces is equal to this horizontal deviation of the COP from the COM. Thus:

τ_z(t)=Fz(t)×[COP_x(t)−COM_x(t)]

where:

- COP_x(t)=position of COP along X axis
- COM_x(t)=position of COM along X axis

Together, these two constitute the total torque exerted by the body seen as a sphere. Thus, the torque τ_xz(t) around the Y axis (in the XZ plane) is:

τ_xz(t)=τ_x(t)=τ_z(t)

and therefore:

τ_xz(t)=Fz(t)×[COP_x(t)−COM_x(t)]+Fx(t)×COM_z(t)

The COM is estimated during the golf swing as follows. Apply a double integral in the time domain of the acceleration exerted onto the body in all three axes, from the start of the golf swing up until time=t. Assume that the person stands approximately still at the start of the golf swing, at time t=t0. The COM at t=t0 must be added to the double integral. At t=t0 it can be assumed that COMx and COMy=0. COMz can be assumed to be H/2, approximately half the person's height H (see FIG. 7a).

${COM}_{x} (t) = \int_{t 0}^{t} \int_{t 0}^{t} AccelX (t) \partial t \partial t = \int_{t 0}^{t} \int_{t 0}^{t} \frac{Fx (t)}{Weight} \partial t \partial t$

${COM}_{y} (t) = \int_{t 0}^{t} \int_{t 0}^{t} AccelY (t) \partial t \partial t = \int_{t 0}^{t} \int_{t 0}^{t} \frac{Fy (t)}{Weight} \partial t \partial t$

${COM}_{z} (t) = \int_{t 0}^{t} \int_{t 0}^{t} NetAccelZ (t) \partial t \partial t + {COM}_{z} (0) = \int_{t 0}^{t} \int_{t 0}^{t} \frac{{Fz}_{net} (t)}{Weight} \partial t \partial t + \frac{H}{2}$

Where:

- Fz=vertical force measured by the z vertical force sensors (result of body weight+body motion against the floor)
- Fz_net=net vertical force, where force of gravity caused by mass of body and acceleration due to gravity (mg) is subtracted from Fz (result of body motion against the floor).

When calculating the Z-component of COM, subtract the negative normal force from the ground which is a result of the positive force of gravity and the person's weight. This constant force should be included in the equation to obtain the net vertical force, and thus the net acceleration of the body mass. Then:

{right arrow over (COM)}(t)=[COM_x(t),COM_y(t),COM_z(t)]

The Centre of Pressure is the actual measured centre point of the four vertical force sensors. Because the dimensions of the force plate and the positions of the vertical sensors are known, the position of the COP can be derived by comparing each sensor position vector {right arrow over (r)} with the respective vertical force measured at the sensor at time=t, and finding the average position vector.

$\overset{}{C O P} (t) = \frac{\sum_{i = 1}^{4} {Fz}_{i} (t) \times \overset{}{ri}}{Fz (t)}$

The angular momentum around the Y and X axes respectively is found by integrating the torque over time for the respective axes.

Thus, the angular momentum L_xz(t) around the Y axis (in XZ plane) can be found according to the following equation:

$\begin{matrix} L_{xz} (t) = I \times ω_{xz} \\ = \int_{t 0}^{t} τ_{xz} (t) \partial t \\ = \int_{t 0}^{t} (Fz (t) \times (C O P_{x} (t) - C O M_{x} (t)) + Fx (t) \times C O M_{z} (t)) \partial t \\ = \int_{t 0}^{t} (\begin{matrix} Fz (t) \times (\begin{matrix} \frac{\sum_{i = 1}^{4} {Fz}_{i} (t) \times \overset{}{{ri}_{x}}}{Fz (t)} - \\ \int_{t 0}^{t} \int_{t 0}^{t} \frac{Fx (t)}{Weight} \partial t \partial t \end{matrix}) + \\ Fx (t) \times \int_{t 0}^{t} \int_{t 0}^{t} \frac{{Fz}_{net} (t)}{Weight} \partial t \partial t + \frac{H}{2} \end{matrix}) \partial t \end{matrix}$

Considering now the X axis (YZ plane), all references to the X horizontal direction in the above equations can simply be replaced with the Y direction. Thus:

Angular momentum L around the X axis:

$\begin{matrix} L_{yz} (t) = I \times ω_{yz} \\ = \int_{t 0}^{t} τ_{yz} (t) \partial t \\ = \int_{t 0}^{t} (Fz (t) \times (C O P_{y} (t) - C O M_{y} (t)) + Fy (t) \times C O M_{z} (t)) \partial t \\ = \int_{t 0}^{t} (\begin{matrix} Fz (t) \times (\begin{matrix} \frac{\sum_{i = 1}^{4} {Fz}_{i} (t) \times \overset{}{{ri}_{y}}}{Fz (t)} - \\ \int_{t 0}^{t} \int_{t 0}^{t} \frac{Fy (t)}{Weight} \partial t \partial t \end{matrix}) + \\ Fy (t) \times \int_{t 0}^{t} \int_{t 0}^{t} \frac{{Fz}_{net} (t)}{Weight} \partial t \partial t + \frac{H}{2} \end{matrix}) \partial t \end{matrix}$

The above described spin kick and torque/angular momentum around the Y and X axes are particularly useful attributes for analysing golf swing motion.

The computer 22 further comprises software that enables the student and/or his instructor to interact with the system and which displays various selected attributes on the monitor in a user-friendly manner to enable the student and/or his instructor to review performance. FIG. 11 is a screen capture of the welcome screen according to an embodiment of the system of the invention. In this embodiment the system is named “swing catalyst”. The student/instructor inputs a username and password in order to gain access to the system. They are then presented with the Main Menu as shown in FIG. 12. This embodiment has two different “modes” of operation: Analysis mode, wherein attributes are displayed in an advanced way, more suitable for use when an instructor is present; and Practice mode which provides simplified feedback, more suitable for use by a student alone.

FIG. 13 is a screen capture of the system in a single swing analysis mode. A video clip 40 captured by video camera 28A and a video clip 41 (image not shown) captured by video camera 28B are shown, showing the swing from different angles. This video can be played back using the controls 42. Measured data of the spin kick timing, temp and rhythm is illustrated in boxes 43, 44 and 45, and is shown as compared to the optimal performance to enable quantified evaluation of the swing. Data collected by the TrackMan™ 27 is displayed in box 46.

FIG. 14 is a screen capture of the system in two swing comparison analysis mode. As can be seen, data relating to two swings is shown to enable comparison between the two swings.

A screen capture of a swing “library” is shown in FIG. 15. This enables the student/instructor to navigate through previously recorded swings using the file structure on the left. The data displayed on the right is the data for the selected recorded swing, in this example the swing of FIG. 3.

As discussed above, the illustrated system is arranged to analyse motion during a golf swing and thus the computer is programmed to calculate attributes relevant to a golf swing, such as those listed in Tables 1 and 2. Of course many of these attributes (and the footprint data) would be relevant in other uses such as analysis of a baseball swing, although further attributes may also be desirable.

Motion Analysis Apparatus

Information

Publication Number

Date Filed

Date Published

Inventors

Original Assignees

CPC

US Classifications

International Classifications

Abstract

Description

Claims

Priority Claims (1)

PCT Information