TIME PARAMETRIZED TRAJECTORY DETERMINATION

Field of the Invention

The invention relates to a method and a system for determining a trajectory of a moving object.

Background Art of the Invention

There are three main strategies of motion tracking and measuring acceleration, velocity and trajectory of an athlete in ski sports: namely high speed cameras, (differential) GPS tracking and inertial sensing based on accelerometers and possibly also on gyroscopes. Among these strategies only the inertial sensing system is portable and immediate enough for analysing ski sports. Furthermore, inertial sensing system is cost effective, which makes it a plausible analysing system for amateurs.

However, the accelerometers and gyroscopes suffer from latency, jitter and bias errors. Due to errors relating to time the reported output diverges from the actual values, and one observes drifting which eventually generates significant errors in the analysis in which the output of the sensing system is used. It is likely that the measurement errors mentioned above prevent engineers from developing portable high precision inertial sensing systems for tracking in sports.

An example of a ski sport demanding high precision tracking is ski jumping. It is essential, yet difficult to define the trajectory of the flying movement on the basis of the output generated by inaccurate devices. Accurate data and information would enable precise analysis so that the information could be used in training. The tracking system should also be fast enough so that takeoff could be analysed with sufficient precision using sufficient amount of data. The demand for precision is emphasized in sports where skis or snowboards are used and which involve jumping and flying movements, the trajectory of which is tracked using small and portable, yet imprecise accelerometers.

In ski jumping, the trajectory is affected by errors ranging to several meters. It should be mentioned that errors in the order of per mils cause errors of the trajectory to be several meters during a 50 s performance. The longer the flying movement, the greater the error. It is a problem in prior art systems that it is not possible to utilize the results. Data generated by using GPS systems is also inaccurate and the speed of the systems is not high enough for facilitating tracking that enables hundreds of measurements on the nose of the ski jump alone.

Document US 6959259 B2 discloses a tracking system in which sensor units are carried by the athlete or his/her equipment. Document US 6073086 discloses an accelerometer attached to sports gear. Documents US 6167356 and US 2002/0040601 A1 disclose a system based on a portable unit including sensors and moving with the athlete.

Summary of the Invention

The aim of the invention is to eliminate the drawback of prior art systems and to considerably improve the accuracy of prior art systems.

Formally, in the mathematical sense, tracking by means of inertial sensing is about generating a one-dimensional, time parametrized trajectory embedded in a global three-dimensional space.

It is characteristic to ski sports, especially to traditional cross-country skiing and ski jumping that the trajectory itself is partly known a priori, i.e. beforehand. Especially the trajectory of the skis in known because the skis follow a known path all the time or most of the time. However, in ski jumping, one of the most interesting parts of the performance is the flying movement. Thus, in ski sports inertial sensing may be limited to the act of generating those parts of the trajectory which are not known, e.g. the trajectory of the flying movement in ski jumping, and the time parametrization. Furthermore, by utilizing the trajectory data known as such, the present invention facilitates the act of creating a low-cost, high-precision inertial tracking system, i.e. a system for tracking acceleration, velocity and trajectory in ski sports.

The purpose of the invention is to yield complementary data which can be exploited to remove the measurement errors of inertial sensing. The principal idea is to "mark" a priori some positions along the trajectory, either a point in the trajectory or a path or line crossing the trajectory. The positions, which are fixed in the global coordinate system, are observed by a sensor system, and their position is fixed. Then, it is possible to detect when an athlete or his/her equipment passes these fixed positions and record the detected data synchronized with the inertial sensing system data. A part of the sensor system or a sensor may be embedded in the athlete's equipment for detecting the passing of a fixed location. If the data related to the fixed locations is combined with the chronometry data recorded by the inertial sensing system - consisting, for example, of the sampling frequency and the number of the time interval - every pre- marked intermediate fixed location in the trajectory will obtain a parameter in the time space, i.e. a time stamp.

The fixed locations are marked e.g. by making them as a source of an electromagnetic field or electromagnetic wave. The sensor embedded in the athlete's equipment detects the fixed locations and records the data. Optionally, the electromagnetic field may be affected by the passing by of the athlete, which may be detected and recorded. The fixed locations and the related equipment constitute a position sensing system, i.e. a system determining the time and the place.

It is a major advantage of the invention that the time parameters of the intermediate fixed locations enable removing measurement and computation errors of the inertial sensing system. That is, if the velocity and position data is integrated from the measured acceleration only, the cumulative error will weaken the accuracy of the data. Now, as the time parameters of the intermediate fixed locations are detected independently of the inertial sensing, it is possible to know what the result of the integration should be at a given time. This translates the problem of the trajectory tracking from extrapolation to interpolation and provides a reliable approach to remove the cumulative errors. Therefore, as a result, the accuracy of the inertial tracking system, i.e. the accuracy of acceleration measurements and the accurate

calculation of the velocity or position data, is significantly increased. Thereafter, a low-cost, high- precision inertial tracking system in ski sports becomes possible as a portable equipment.

Brief Description of the Drawings

Figs. 1 and 2 show simplified examples for marking fixed locations,

Fig. 3 shows the tracking system applied in ski jumping,

Fig. 4 shows an electromagnetic sensor embedded in an athlete's shoe, and

Fig. 5 shows the principle of the tracking system,

Fig. 6 shows the function of the tracking system during a ski jump,

Fig. 7 is a flow chart showing the method for inertial tracking and the use of fixed locations, and

Fig. 8 shows the principle of the calculation.

Detailed Description of the Invention

Fig. 7 is a flow chart showing the inertial tracking method and the use of fixed locations. To generate the tracking data related to the trajectory, the measured acceleration and thereafter the angular velocity data is calculated by integrating. Integration is a standard and well known procedure for generating velocity or position information on the basis of acceleration data. Tracking data is obtained by integrating twice the acceleration data given by the global coordinate system. The acceleration data is generated by means the inertial sensing system. The fixed locations of the system enable adding an additional calculation loop which is effectively used for removing errors from the measurements, e.g. by iteration or by means of any other mathematical methods known as such. Typically, the accelerometers and gyroscopes suffer from bias errors and drifting of the measurement reading, i.e.

roaming. The bias and roaming errors are probe-dependent, and the act of filtering out these errors depends on the accelerometer or the gyroscope in use. However, if there is a sufficiently large number of fixed locations in use, the tracking data related to the fixed locations may also be employed to estimate how the errors are generated during the measurement using the inertial sensing system.

Fig. 7 shows an embodiment of the method in which measured data, parametrized by time, is used (step 71), including 1 ) acceleration and angular velocity data, and 2) data related to the fixed locations, e.g. electromagnetic data generated by the position sensing system. The measured acceleration data is mapped to the global coordinate system and the effect of gravitation is subtracted (step 72). The time parameters related to the fixed locations are read (step 73). Then, the acceleration data is integrated to obtain velocity data, and further, the velocity data is integrated to obtain position data (step 74). Thereafter, the velocity and position data are matched with the data related to the fixed locations, i.e. the time parameter and the location of the fixed location in the global coordinate system (step 75). If the velocity and position data match the data related to the fixed locations, then the acceleration data, the velocity data and the position data will be accepted for further use (step 76). If the velocity and position data do not match the data related to the fixed locations (step 77), the errors of the acceleration data will be filtered out e.g. by iteration (step 78). Eventually, the velocity data and the position data will be accepted for further use (step 76).

Figs. 1 and 2 show an example for realizing the fixed locations 2 of the position sensing system. In the example, permanent magnets or thin wires carrying current are set across the ski track 3 for tracking the trajectory 1. The locations of the equipment, i.e. the fixed locations 2, are known in the global coordinate system.

Fig. 4 shows an electromagnetic sensor 4 used in the position sensing system. The sensor 4 is embedded within the athlete's equipment (the shoe 5 or the ski 7). Here, e.g. a magnetic sensor 4 is embedded within the shoe 5 to measure one or all of the three magnetic components of

a magnetic field 6. The magnetic sensor 4 is e.g. a Hall sensor or a small loop. The output of the sensor 4 is stored along with the inertial data received from the inertial sensing system, thus making it possible to recognize and track when the sensor 4 passes the fixed locations 2.

Fig. 3 shows an example of the inertial tracking system to be used in ski jumping where the trajectory 1 is not fully known a priori. In ski jumping the trajectory 1 b in the inrun 9 is well known, but the trajectory 1a of the flying movement and the trajectory 1c of the landing is not. Especially the point of landing 10 varies, even though the landing area 11 is known. Here, the trajectory 1 b to the end of the inrun 9 is marked as explained. When fixed locations 2 on the landing area 11 are marked by horizontal lines, the length of the jump can be specified from the inertial data andthe electromagnetic data. The fixed locations 2 are marked with permanent magnets or current carrying wires of the position tracking system. The magnets or wires cross the landing area 11 transversely in relation to the course of movement of an athlete. The position data, i.e. the electromagnetic data, is provided by the position tracking system. The moment the athlete hits the landing hill appears clearly in the data generated by the accelerometer 8 (i.e. the inertial sensing system) and thereafter the electromagnetic sensor 4 (i.e. the position tracking system) embedded in the athlete's shoe 5 or ski 7 detects all the fixed locations 2 situated after the landing point 10. The electromagnetic data yields the time parameter of these points of the trajectory 1 where the athlete has crossed the marked lines. The inrun 9 of the ski jump is marked with simi lar fixed locations 2 which the athlete passes before the takeoff point 12 and the flying movement. The time parameters related to the passings of the fixed locations 2 obtained from the inrun 9 and the landing hill can both be utilized to eliminate the cumulated inertial data errors related to the flying movement, thus making it possible to compute the trajectory 1 a parametrized by time. The length of the jump and the velocity at different points of the trajectory 1 are details which are then easily read from the trajectory parametrized by time.

Fig. 6 shows an example of data generated by an accelerometer and a magnetic sensor just before and after landing in ski jumping. The

horizontal axis represents time. The upper, solid line 13 represents the output of the accelerometer and the lower, dashed line 14 represents the output of the magnetic sensor. The first vertical line 15 at t ~ 1.05s shows the moment the athlete hits (see output at point 18) the landing hill (Fig.3, point 10). The second vertical line 16 at t - 1.4 shows the moment the athlete passes (see output at point 17) a fixed location marked with a permanent magnet or a wire.

A preferred embodiment of the invention for constructing the fixed locations is based on electromagnetism. A simple technique is to mark the locations with permanent magnets or current carrying wires. An alternative technique is to use a pair of thin wires having different potentials and then to detect the generated electric field. A third option is an antenna, such as a tiin wire, a long and narrow loop, or a thin radiating slot. The sensor carried by the athlete detects the fixed locations. Technology and sensors known as such may be used to implement the position sensing system.

The invention is used to eliminate errors of the inertia measurements carried out by inertial sensing system. The elimination is possible if some points of the trajectory with fixed spatial parameters are known a priori, i.e. the fixed locations. Thus, the remaining task is to select a sufficiently large number of fixed locations necessary for eliminating the errors to achieve the desired accuracy. In cross-country skiing, especially the traditional style, the skiing track or the trajectory is fully known and defined. However, inertial sensing combined with position sensing may be used. The inertial sensing is used to track the athlete between the fixed locations.

Fig. 8 shows an example how the errors of the acceleration data are filtered out when determining the corrected velocity and position data. The s-axis shows the fixed locations 81 (points s1 - s6) relating to the interpolation points 84 used in the calculation. The eaxis shows the errors 82 (points e1 - e6) related to the fixed locations 81. Points s1 - s6 define the position in which the athlete was at a certain moment of time, based on the data of the electromagnetic sensor, i.e. the position sensing system. The errors e1 - e6 can be determined by matching

this data with the velocity and position data calculated only on the basis of the acceleration data, provided by the inertial sensing system. Corrected velocity and position data is generated by subtracting the error from the calculated velocity and position data. An error related to a fixed location defines an interpolation point 84. By means of two or more interpolation points 84 it is possible to calculate an interpolation curve 83, based on known interpolation schemes. The curve 83 defines e.g. the error e7 between the two points 84 and between the fixed locations s3 and s4. The σror between the fixed locations 81 is also determined in this way, e.g. for the flight movement in ski jumping. By increasing the number of interpolation points it is possible to obtain more accurate data for the trajectory between the fixed locations 81. Furthermore, since the tracking does not necessarily start at a moment when the target object is completely immovable, the iteration is needed to find out the right initial conditions and the position for the point sθ.

In ski jumping, the events taking place during the flying movement must be recorded by the inertial system. The trajectory of the inrun is well specified, and the flight time is relatively short. Furthermore, the profile of the landing area is known beforehand and after landing two of the spatial parameters of the trajectory are known a priori as demonstrated in Fig. 4. The information related to the transverse location of the athlete related to the landing area is not known using the system of Fig. 1 , however, the information is not critical nor absolutely necessary.

Referring to Fig. 1 , in other ski sports, e.g. alpine and freestyle skiing, the spatial trajectory 1 is not precisely known and only two out of three spatial parameters are predictable at those parts of the track, where the athlete's ski is known to slide on snow or the athlete is moving sideways. In this case, marked locations 2 in the form of grid lines on the slope are needed. The grid lines may or may not be orthogonally situated. As an example, the poles of the slalom course may be marked with one or several rings of magnets or wires. However, the spatial parameters of the fixed locations 2 forming points, lines or a net must be known in the global coordinate system.

Referring to Fig. 5, the construction of the inertial tacking system 19 according to the invention may vary. The data generated by the inertial sensing system 20 including the accelerometers and the position sensing system 21 including the electromagnetic sensor is stored in the device 22, e.g. a data logger, carried by the athlete. The device 22 is controlled by a control system, comprising an output for transmitting the data to an analysing unit for completing the process described in Fig. 1. The data may be transferred e.g. wirelessly in real-time or downloaded from the device 22. Depending on the processing capacity of the device 22, some part of the analysis, the calculation and analysis of the information may be conducted in the device 22 carried by the athlete. Typically, the generated and loaded data is analysed in a remote device 23, connected to the data logger, typically a computer- controlled device comprising a programmable processor and programmed to perform the necessary steps of the process. Preferably, the spatial parameters related to the track or ski jumping are stored in the remote device 23 and used in the process. The remote device 23 includes necessary input devices and display devices for displaying the results of the analysis in an illustrative way or graphically. The remote device 23 is eg. a laptop or a mobile phone with adequate processing capacity. The laptop or the mobile phone may apply wireless data transfer methods for collecting the necessary information and may use e.g. radio waves or infrared. The invention may be implemented as a computer program product or a computer program stored on a computer-readable medium and configured to carry out the method according to the invention when uploaded in a device. The computer program comprising necessary instructions and program code may be run in the analysing unit, e.g. the remote device 23.

In another embodiment of the invention, the athlete may carry a permanent magnet which is detected by a wiring positioned in fixed locations. The wirings are connected to a device 24 analysing the output, and the point in time the athlete passes a fixed location may be deduced from the data. The clock of the device 24 must be synchronised with the clock of the inertial sensing system 20. The wirings and the permanent magnet constitute a position sensing

system 25 of different kind related to the embodiment shown in Figs. 4 and 5. Furthermore, the position sensing system 25 and the storing of location data requires a more complex system, separate from the inertial sensing system 20. However, the data generated by the inertial sensing system 20 and the position sensing system 25 is analysed in the remote device 23 as described earlier.

The results of the analysis and the corrected acceleration, velocity and position data may be stored for future use, e.g. for comparing the results of an athlete in training. The obtained data may also be used in other systems, e.g. sportscasting systems 26. As an example, the length of the flying movement, the velocity of the athlete at a predefined moment or the trajectory of the flying movement in a graphical way may be combined with sportscasting. The information added to the television signal or broadcast gives the public interesting information quickly and reliably during the event.

The present invention employs cevices known as such, however, the devices are modified for use in way which is evident for a person skilled in the art on the basis of the description above.

The invention may be applied in sports in general. The object to be followed and its trajectory to be traced may be an athlete, or an equipment of an athlete. In addition to ski sports, the invention may be applied in other type of sports also.

The invention is not limited to the above-mentioned examples but may vary according to the appended claims.