ES2358691T3 - METHODS AND SYSTEMS FOR IDENTIFYING THE POSITIONS OF LAUNCHING DROP GOLF BALLS. - Google Patents

Abstract

A method for identifying, with respect to a descending ball, the position between a plurality of throwing positions (52, 53) from which the ball has been thrown, characterized in that the method comprises measuring the descent parameters of the descending ball to obtain at least the time of descent of the descending ball and a measure dependent on its angle of descent in elevation, calculate an estimate of the duration (Edur) of flight of the descending ball based on the measure of the angle of descent, measure , with respect to each position (52,53) of throwing, the time interval (Dur1, Dur 2) elapsed between the throwing of a ball from that position and the time of descent of the descending ball, compare the time interval ( Dur1, Dur2) measured with respect to each launching position (52, 53) with the calculated estimate of the duration (Edur) of the flight of the descending ball to determine for which of the positions ions (52, 53) of launching the respective interval is a very close coincidence with the estimate (Edur) calculated, and identifying the position from which the descending ball was thrown with the position (52, 53) of launching for which it exists the very near coincidence.

Description

This method refers to methods and systems to identify the throwing positions of descending golf balls. The methods and systems of the invention have application, in the context of golf strokes performed in golf driving ranges and in other facilities of this sport.

According to the invention, a method, and in another aspect a system, is provided in one aspect, to identify with respect to a descending golf ball the position between a plurality of throwing positions from which the ball, characterized in that it comprises a stage of, or means for, measuring the descending parameters of the descending ball to obtain at least the time of descent of the descending ball and a measure dependent on its elevation angle, a stage of, or means for calculating an estimate of the duration of the flight of the descending ball as a function of the measure of the angle of descent, a stage of, or means for, measuring, with respect to each throwing position the interval of time elapsed between throwing a ball from that position and the time of descent of the descending ball, a stage of, or means for, comparing the measured time interval with respect to each lan position with the calculated estimate of the duration of the flight of the descending ball to determine for which of the throwing positions the respective interval has a very close coincidence with the calculated estimate, and the stage of, or means for, identifying the position from the one that has thrown the ball that descends with the throwing position so that there is a very close coincidence.

A measure of the angle of descent in elevation of the ball that descends from the measurements of the component velocities of its descending trajectory could be obtained. .

References to the "descent" of a golf ball are intended to refer to the end portion of the bearing path of the ball as distinct from any part of its subsequent rebounds or its rolling path. Likewise, the term "a descending golf ball" is understood to refer to a golf ball that flies near the end of its range path and preferably within at least 10% of said trajectory. The end of the range path is the point where the ball hits with a “target” · at a certain distance from the initial throwing position. The launching position could be, for example, one of several fields for game reserve from among a golf driving range.

The present invention is applicable to all golf strokes, but is especially applicable to strokes in which the net aerodynamic force (i.e. the sum vector of lifting and dragging forces) is greater than 1.0 meters per second per second (m / s2) but more especially greater than 10 m / s2.

An object of the present invention is to provide methods and systems for identifying golf strokes that are not based on flight prediction or track monitoring and that do not require golf balls that are specially marked or electronically labeled or balls with similar markings.

Flight prediction methods require expensive equipment to mediate the launch of the ball capable of measuring the spin components of a ball, and sometimes require special balls. Flight prediction is also subject to significant non-systematic errors caused by high intensity winds or by random variations in the aerodynamic properties of the balls, which change due to the degradation of their surface. Flight prediction methods can be refined by measuring the landing positions of the balls and flight durations. However, systems that rely solely on the measurement of landing positions and times are inadequate, because most of the information about the flight history of a ball is destroyed on landing. Systems that identify individual golf strokes using ball flight prediction and the measurement of landing positions are described in US-B-6179720 and US-A-2007/0167247.

Path tracking systems (such as video tracking systems or radar tracking systems) are very expensive, require extensive means of data processing and may not work reliably when there are many concurrent balls in the space of flight. Typically, one or more video camera or radar tracking devices are locked on the flight path of a ball at a time and follow that ball throughout its flight from the initial impact to at least the end point of landing. This means that cameras or radars must capture case data continuously for several seconds, while, in the present invention, ball path measurements are only required for a few milliseconds just after the initial impact and just before landing on the furthest part of the field. Document GB-A-2294403 describes a method of identifying golf strokes by video tracking combined with swing measurements. WO 2006/002640 is known as a system that determines the launching position of a golf ball by using the horizontal and vertical angle

The identification of individual balls with labeling means such as embedded chips with radiofrequency identification (hereinafter RFID) is expensive, and the mechanical means of collecting balls to decode the RFID data is cumbersome and unreliable. JP-A-822431 and US-B-6607123 describe systems that identify individual golf strokes using RFID tagging.

The measurement of the descent parameters, and the launch parameters, could use electromechanical, electroacoustic, electro magnetic, electro-optical, microwave radar Doppler, ultrasonic Doppler, high speed video or any other technology that provides an electronic measurement Suitable for signal processing.

The measurement of the descent parameters could typically be performed as the ball approaches a “target” located outside the driving range. The objective could be a circular, oval or rectangular or similarly marked area or it could be a panoramic green area for golf, which resembles a real golf course green. Various alternatives of objective designs could be used. For example, embankments or structures that resemble large archery targets, large water sources shelters where the balls land and disappear without bouncing out, and other modalities of diverse structures. Alternatively, the measurement could be made when the ball that descends is about to land anywhere in the furthest part of the field or in a substantial part of it. Preferably, but not limited to, the descent parameters are measured within a short distance of the end of a ball range path (for example, in the final 5%, or more preferably in the final 2% of the distance of scope).

Limiting the extent of the measurement range of the descent reduces the cost of the measuring device and improves the reliability of the measurement. In a preferred embodiment, the descent parameters are measured within less than 2 meters of the landing surface or ground level and with the apparatus itself is less than 2 meters above the ground level. The restriction of the height of the detection apparatus in the farthest part of the field is much less important in the landscape of the farthest part of the field, which is highly desirable.

A preferred means of measuring the descent parameters in each objective uses at least four "detection planes". These detection planes comprise fan beams with a very small angular field of vision normal to the detection plane and wide fields of vision in the planes of the detection planes. Means are provided to detect the angular position (within the fan beam) and the instants of time when the balls cross the detection planes. Typically, the detection planes are arranged in two coplanar pairs that are parallel and deviated, but pointing in different directions, so that the two-dimensional position of a ball can be calculated by triangulation when it crosses the plane of the common visual field The difference between the moments when the ball crosses the detection planes, and the distance of deviation separation between the planes that contain the centers of the fields of vision, gives a measure of the velocity vectors of the ball and its instantaneous position in space . Preferably, but without limitation, the fields of vision of all the detection plane sensors are horizontal, or almost horizontal, and are located very close to the surface of the farthest part of the field.

The measurements of the parameters of the thrown balls are conveniently carried out by a means that is low cost, and has reliability and a broad "stroke acceptance" In this context, the term "stroke acceptance" means the capacity of the measuring medium to measure all types of hits when the ball leaves the starting point in a wide variety of speeds and directions. Preferably, the initial descent parameters are measured over a small length of an initial ball path (for example, not more than 5 meters, but more preferably less than 2 meters), such that the balls that are coming from Adjacent exit zones do not interfere with the measurement. However, any means of measuring the initial parameters of descent of a golf ball in the presence of other golf balls in flight could be employed. The signal processing means require data that identifies each exit zone of the driving range and records the distance, the direction of the azimuth and the height of each exit point with respect to each goal of the golf course.

You can measure the impact time of the balls in the exit zones by detecting the impact sound using one or several microphones. If required, an approximate measure of the velocity and direction of the ball could be obtained from the signals of several microphones, possibly configured as a phase grouping. Alternatively, the impact time can be obtained by optical means, for example by detecting the passage of the ball through one or more detection planes. The aforementioned optical means could be configured to measure the impact time and optionally at least one of the parameters of the throwing speed of the ball, azimuthal angle of the throwing and lifting angle of the throwing of each hit ball.

While in one embodiment of the invention standard golf balls can be used, it can be advantageous to use balls with a retro-reflective surface. There are two advantages of retro-reflective golf balls. First, the balls are easier to detect by means of optical sensors, especially in long ranges. Secondly, at night or in conditions of poor ambient lighting, the balls are much more visible to the human eye when they are illuminated with a light source very close to the golf player. This allows a very low energy and lighting intensity to be used, which is convenient from an environmental point of view.

The main drawbacks of retro-reflective golf balls include the additional cost of their manufacture and the possible degradation of the retro-reflective surface with frequent use. Therefore, it is an object of the invention to provide methods and systems for identifying strokes and golf that are compatible with standard unmodified golf balls, and also methods and systems that are designed to work with retroreflective golf balls.

It is also an object of the invention to provide methods and systems that are substantially immune to differences in the aerodynamic properties of the different types and conditions of golf balls. Although the diameters and weights of the different types of golf balls are all very similar, their dimple patterns vary significantly, and this results in substantial differences in the length of the range and in the duration of the flight under identical launch conditions. The above has been illustrated in a communication describing the research of the United States Golf Association (hereinafter USGA) on predicting the flight of golf balls (Fifth fence, SJ, 2002. A model generally applicable to aerodynamic behavior of golf balls Published in Science and Golf TV, Ed. E, Thain, 346-348 London: Routledge). In this communication, Quintavalla uses a mixture of types of balls that show variations in transport length and flight duration due only to differences in dimple patterns, with range lengths that vary by 25 meters (with an average length of range of 238 meters and flight durations that vary in 1.0 seconds (with an average flight duration of 6.3 seconds): Other dimple patterns not included in this investigation could greatly exceed these differences, and the effect of degradation of the surface will be even more critical as the degradation becomes severe.

Consequently, it is clear that drag and drag aerodynamic forces affect golf balls that have different surface characteristics differently. However, the upward and drag forces for short approach shots (for example, up to 25 meters) have a much lower intensity than the force of gravity, so that the trajectory is very close to a parabola, with the angle and the descent speed equal to the angle and launch speed (on a horizontal playing surface). A "bird's eye view" of a short chip hit shows that the ball always flies in a substantially straight direction without deviating significantly to the right or left (that is, without taking effect to the right or left). Again in this case, this is because the side effect and side winds have a negligible effect on the flight of golf balls at low speeds. It has been found that, if the speed of descent of the ball is measured, it can be said when the ball has been hit, assuming that the ball has been thrown in a given horizontal plane (for example, at ground level). In addition, if the direction of its azimuth and its velocity is measured, the position of the initial impact can be accurately calculated. Therefore, it is possible to identify who made the blow regardless of the pattern of dimples or the state of the surface of the ball. This illustrates the principle of the present invention as applicable to short chip beats.

However, the driving ranges are sometimes staggered, so it cannot be assumed that a ball has been thrown from ground level. For the case in which a short step hit has been thrown very steeply, errors in the measurement of its descent parameters make it unreliable to estimate whether the ball has been hit from one more step or from a lower step . For this reason, it is preferable to measure the impact time for all hits. This additional information allows reliable identification of where and when short-range golf shots have been taken. The impact time also provides a record of how many balls have been played at each starting point. This monitors the use of the balls by the user and helps prevent theft of the balls.

When launch speed and flight distance increase, the upward and drag forces increase. The forward speed is reduced rapidly during the flight, which makes the descent part of the path much shorter than the ascent part. Hence, the angle of elevation in descent is greater than the angle of elevation of launch and that the rate of descent is less than the speed of launch. Flight simulations for a wide range of blows and wind conditions show that, for a given flight length, the flight duration of any blow can be estimated very accurately just from the knowledge of the descent angle. When a ball lands on a target, optical, radar or similar sensors measure its component descent speeds just before it lands. Then, a central computer estimates the duration of the flight and seeks to couple it together with the interval elapsed since the impact time of each recently hit ball.

In many cases, the process of coupling an estimated impact time and actual impact time in a conjugate manner is all that is needed to identify a golf hit. However, in general, more than one estimated impact time and actual impact time must be considered to establish reliable identification. Thus, in order to identify the initial starting position of a certain descending golf ball, it is sometimes necessary to try to conjugate a set of descent parameters (for a particular stroke) in conjuction with the initial launch parameters of one of a plurality of "possible hits" that occur that have almost simultaneous impact times.

To clarify the description herein, it is convenient to designate the parameters that refer to a "possible hit" with the subscript k, where k has a value of 1 to K, and K is the total number of "possible hits" determined by the identification process and is never greater than the number of players using the installation.

It has been found that the expected flight duration Edurk for the k-th possible strike can be estimated very accurately from the following equation:

Edurk = C1k + C2k x β (1)

In the previous equation, Edurk is the expected flight duration in seconds; C1k and C2k are known constants dependent on the range of distance from the k-th exit zone to the descent position; and β is the measured angle of descent of the ball in the position of descent, in degrees.

Other calculation modalities could be used. For example, the second member of equation (1) could include additional terms dependent on other descent parameters, and the average wind speed along the direction of the flight and the air density. Optionally, the calculation could be implemented with a query table or other form of algorithm.

In one embodiment of the invention, the value of Edurk is determined by a calculation that depends exclusively on the angle of elevation of the descent β and the distance of the range. More preferably, Edurk depends on the angle of elevation of the descent, the distance of the range and at least one of the following parameters: absolute rate of descent, angle of azimuth of the descent, wind speed in the direction of flight of the ball, and air density.

In addition to Edurk, sometimes two other parameters are used in the present invention, which are the following: Edirk expected direction and Edecek expected deceleration ratio. While Edurk is estimated using only descent parameters or descent parameters combined with wind parameters, the expected Edirk direction and the expected deceleration ratio Edeck are calculated from both launch and descent parameters combined with the parameters of wind.

The expected direction or azimuth angle for the fifth possible Edirk stroke can be obtained from the following equation:

Edirk = αFk + C3k x (αFk - αLk) (2)

In the previous equation, C3k is a constant, αLk is the launching azimuth angle measured at the 50th starting point, and αFk is the azimuth angle of the position of the ball that descends with respect to its initial position in k -th place of departure. The Edirk, αLk and αFk angles have been conventionally measured in a dextrogyric direction from a fixed reference direction.

In one form of the invention, the value of C3k is preferably 1.0 for all values of k. More preferably, C3k depends on the range distance and at least one of the following parameters: intensity and direction of the side wind, range deviation, total speed and direction of the wind, elevation angle of the launch, launch speed, range duration, and air density.

In the absence of side effect and side wind, and when the starting point has been correctly coupled together, the term (αFk - αLk) is zero, that is, the ball travels along a straight line (in bird's eye view) from the starting point to the down position. When the side effect or the side wind causes a ball to deviate from the straight line, it has been found by simulation that Equation (2) can give a very accurate prediction of the azimuthal direction of descent even for very large variations in effect lateral and side wind.

The comparison of different Edirk values with the actual azimuthal direction in the descent (obtained from the measurement of the descent parameters) provides a means of identifying the correct starting position corresponding to a given descending ball.

EdeCk is a dimensionless parameter that is related to the total deceleration of a horizontal velocity of a ball, and is defined as the relationship between its horizontal velocity in the throw and its horizontal velocity in the descent. This relationship depends on a large scale on the duration and distance of the range. For short chip beats, the ratio is close to unity, since there is a negligible change in horizontal speed. For strokes with effect, the ratio is typically 3.0 or higher. EdeCk is preferably obtained from the following equation:

Edeck = C4k + C5k x Durk (3)

In the previous equation, C4k and C5k are constant and Durk is a “possible duration” value that corresponds to one of a number of exit positions. Therefore, Edeck and Durk each have a plurality of values, and comparing each Edeck with the values of the deceleration ratio obtained from the relationships between the actual launch and descent speeds, a means of identifying the position of correct output.

The values of C4k and C5k preferably depend on the range of reach and initial launch angle achieved for a stroke; but more preferably they depend on range distance, launch angle, wind speed, wind direction and air density.

Each of the Edurk, Edir, and Edeck parameters has a different error distribution, which can be determined by analyzing the actual measured data when they become available. Frequently, a correct expected coupling can be achieved in conjugate with the actual parameters with 100% certainty. This will usually be the case when there are only a few players using an installation according to the invention and the impact times of their strokes are separated by a few seconds. However, this will occur frequently in times of heavy occupation. The records of these blows "with 100% certainty" plus the prevailing wind and weather conditions records can be used to refine the beat identification algorithms and the accumulated error distribution data.

A preferred method of identifying the starting point corresponding to a given descending ball is to first base the identification exclusively on a conjugate coupling between one of the possible durations of Durk range (k = 1,2, ... k) and the corresponding expected durations Edurk. It is possible to make the number k depend on the distance of the range and the wind conditions, and be chosen in such a way that it includes all the blows that occurred within a time of obtainable duration. Alternatively, a default value can be set for k including all hits that have been taken during the previous 10 seconds. The possible durations of the Durk range are equal to the time differences (td - tk) where td is the time at the moment of measuring the descent and tk the impact time of a stroke at the k-th place of departure. In some cases there will be only a single very approximate coincidence, with other hits that do not match by 3-sigma or greater. In this case it is safe to assume that the unique match is correct.

In the event that two real impact times coincide in Edurk within 3 typical deviations (sigma) (or any other error limit), the coincidences are preferably made using at least two or three probabilities of event P (Durk), P (Dirk), and P (Deck) that are determined from the distributions of errors. P (Durk) is defined as the complement of the probability of the event in which the actual duration of range falls within ± (Edurk - Durk) around Edurk. Since Edurk is the most likely duration (obtained from analyzing a large sample of previous strokes), the value of the cumulative distribution function F (Edurk) is 0.5, while F (Durk) will take some value above or below 0.5 depending on whether it is greater or less than Edurk, respectively. Therefore, P (Durk) is defined in the following way:

(P (Durk) = 1 - 2 x | F (Durk) - 0.5 | (4)

When the values of Edurk and Durk are very close, the probability P (Durk) approaches unity, and reciprocally, when the values differ by 3 sigma or more, P (Durk) approaches zero.

P (Dirk) and P (Deck) are defined equivalently from the distribution of address errors and the distribution of errors in the deceleration ratio. It is assumed that the correct match is the stroke that has a maximum joint probability, which can be either P (Durk, Dirk) or P (Durk, Deck)) or, more preferably, (P (Durk, Dirk, Deck ).

It should be noted that there is a certain correlation between duration and deceleration, so it can be argued that Durk and Deck are not independent events. However, however, the errors between the estimated and actual parameters are extremely small, and, due to unsystematic errors of measurement and variations in unmeasured parameters, such as recovery and roughness of the ball. Thus, this is the case that the errors in Edurk are independent of the errors in Edeck, and therefore it is valid to say that the joint probability P (Durk; Deck) is equal to the product of the individual probabilities.

In a hypothetical example, if the errors in Edurk, Edirk, and Edeck have a normal distribution with 1-sigma values of 0.1 seconds, 0.7 degrees and 0.1 respectively, and:

Edur2 = 4.32 seconds Dur2 = 4.32 seconds

Edir2 = 81.5 degrees Dir 2 = 85.0 degrees

Edec2 = 2.2 Dec2 = 2.4

Then: P (Dur2) = 1 P (Dir2) = 0 P (Dec2) = 0.046

In the previous example, the expected duration of a descending ball coincided perfectly with the actual impact time of the second possible hit (k = 2), so P (Dur2) is equal to 1.0. However, P (Dir2) is zero, because the difference between the expected address and the actual address is much greater than 3-sigma. Thus, the joint probability that the second candidate hit is the correct match is zero, so this hit does not match.

In the following, methods and systems are described by way of example and with reference to the accompanying drawings, in which:

Figure 1 is a logical block diagram of a system according to the invention for use in the provision of identification means in a golf facility;

Figure 2 is a diagram showing the general profile of a golf stroke in side elevation;

Figure 3 is a graph of range distance as a function of recovery for a representative golf hit;

Figures 4 (a) and 4 (b) are graphical representations of flight duration T as a function of the angle of descent β for simulated golf strokes;

Figure 5 is a schematic plan view of a driving range showing the trajectory paths of two examples of golf shots;

Figure 6 is a graphical representation of azimuthal angle errors as a function of speed with side wind for simulated golf strokes;

Figure 7 is a graphical representation of deceleration ratios as a function of the duration of the range for simulated golf strokes;

Figure 8 is a plan elevation of a device for measuring the descent of the ball and the target according to the invention; Y

Figure 9 is a side elevation of the ball descent measurement apparatus and the objective of Figure 8.

Reference axes X, Y, Z have been shown for convenience for transport orientation when appropriate in some of the figures referred to in the description that follows. In this respect, the Z axis is vertical and points upwards, the Y axis is horizontal and points towards the ground (that is, along the general flight line of a golf stroke) and the X axis is perpendicular to the Y and Z axes and points in the direction from left to right, facing the ground.

The block diagram of Figure 1 schematizes the top level system for a golf facility according to an aspect of the present invention, where several players hit golf balls in the same general area and detection means have been provided to identify the Initial starting position of each hit.

Block 1 represents the primary input, which is a sequence of n balls hit at random moments tn from random output positions having coordinates xn, yn and zn. Blocks 2 and 3 represent secondary entrances that comprise various “disturbances” or unmeasured entries such as the speed of rotation ω of the ball, its axial precession of rotation τ, its roughness, the speed and direction of the wind in successive moments at along its flight path, and the density of the air ρ, which varies with atmospheric pressure and temperature.

Block 4 represents the force of gravity and the ascending and drag aerodynamic forces experienced by a ball throughout its flight, all of which determine the profile and duration of the flight. Block 5 represents the launch analyzers that measure the linear parameters of each ball after the impact where the linear parameters comprise the vector velocity, the position coordinates on the impact, and the time of the impact. Typically, launch analyzers do not measure rotation parameters, since this measurement is difficult and requires expensive equipment. The data of the launch analyzers is entered into a central computer 6.

The target sensors 7 also send data to the computer 6. The data measured by the target sensors 7 comprise the vector speeds and the position coordinates of the falling balls. Since not all balls belonging to the sequence 1 of ball impacts reach a target, the sample of the balls measured by the target sensors 7 is a subset m of the n ball impacts (m <n). Alternatively, the objective sensors can be extended throughout the furthest part of the field, so that all the balls that reach the furthest part of the field are measured. Optionally, data from one can be sent to the computer. or several anemometers.

The computer 6 processes the data of the various inputs to determine which output position (and therefore which player) corresponds to each of the m balls measured by the target sensors 7. When a player successfully poses a ball on a target, computer 6 sends a score indication to a reading device 9. This reading device 9 could be a central device that serves all the users of the installation. For example, once a player has finished a session in a practice position, he can collect printed information showing the score for each stroke. Additionally, each position can be provided with indicative means of individual scoring, which can be simple devices based on an audible tone, indicators with a light-emitting diode lamp (hereinafter LED) or a similar element, or more sophisticated means such as touchscreens.

In a preferred embodiment, the system is provided with two or more means of visual presentation of scores and other information. The flags installed in the furthest part of the field mark the centers of various objectives, and each flag can be provided with three differently colored LED beacons that are positioned and concentrated in such a way that they are easily visible to players in the exit posts. A simple scoring modality such as a score of one, two or three points could be used for strokes that land within a range of 10%, 5% or 2% of a target, respectively, corresponding (for example) to a Red, white or blue LED beacon light. The light of a beacon is preferably illuminated by pulses on and off, so that it can be better appreciated, and allow it to gradually turn off in a few seconds after the scoring has been completed. Flag-mounted beacons in the target areas would provide a means of indication, and a means of indication can be provided in parallel within each exit post. This second indication means could be an audible signal or corresponding colored LED indicators or similar elements that are only activated at an exit post when a punctuation hit has been struck from that position.

An advantage of a simple indicator, sound or LED, is that it can be very economical and robust, and therefore less vulnerable to theft or vandalism compared to more sophisticated devices. However, players often prefer a computer-generated visual presentation or touch screen installations or similar equipment, and such visual presentations could be permanently installed. Alternatively, instead of providing these visual presentations as permanently incorporated equipment, they could be advantageously provided as portable units. These portable units could be visual presentations designed on purpose that optionally rented players for the duration of their games or they could be a laptop owned by the user or a manual device such as a pocket computer or digital personal assistant (hereinafter PDA) or a "smartphone" or similar equipment. Modern laptops and PDAs are often equipped with a BluetoothTM short-range data communication device so that an appropriate software application can be installed on a user's personal PC so that it can communicate with the computer system. identification and punctuation of blows, which must also be equipped with BluetoothTM. Any standard wireless communication system could be installed, including broadcasting and infrared technology. In addition to the hit-by-hit score, the software application could provide many other services such as billing and payment on the fly, multi-player scoring, competition registration, blow analysis, individual performance history, etc.

Normally, players who are on the field are provided with an electronically readable "game card" or an equivalent device that contains a user identification code. All the starting points (at least some) belonging to the installation are provided with a game card reader device. When a game card is read at a given starting point, the throwing sensors and reading means are activated in relation to association with the determined starting point. The player can choose to use only the basic reading indicators or use a laptop. The user's laptop must be programmed to link to the specific control data and instructions for the identification code contained in their game card. If preferred, some users of the facility can hit balls outside the starting points without using a game card, but in this case the data link that is associated in association with their starting points is not activated. .

The target sensors 7 could also be configured to measure the bounce of the balls that land on a target. These measures are not used for impact identification, since rebounding depends on large-scale terrain conditions that vary randomly, such as surface irregularities, impact absorption, etc. However, the additional data is very useful for golfers, since they provide direct information in relation to the actual boat of a golf ball after its course of reach and from it the probable extension of the final race (that is, bouncing and rolling) of the ball. In addition, the prediction of the subsequent boat and rolling for a given terrain can be fairly accurately calibrated.

Figure 2 is a diagram showing the general profile of a golf stroke in side elevation. The trajectory of the scope 20 comprises the flight path of the ball from its initial launching point 21 to the point of its first rebound 22. Note that the trajectory of the scope is asymmetrical with the angle ε of initial launching elevation less than the angle of descent β (both angles with respect to the horizontal). This symmetry (β> ε) is almost always true for real golf shots, but for short chip shots (for example 25 meters) the path is almost symmetrical.

Figure 3 shows a graphical representation of the percentages of range distances as a function of recovery for a typical hit with effect with a launch angle of 12 degrees and a launch speed of 57 m / s in calm wind conditions. The values of the length of the range and the duration of the range were calculated using standard formulas and models for coefficients of upward and drag force that are known to approximately simulate the performance of real golf balls. The maximum range is shown to occur when the recovery is approximately 4500 rpm, but the range distance is reduced by less than 5% for a wide range of recoveries from below 3000 rpm to above 7000 rpm. The flight time T and the descent angle β are shown at three points on the graph, namely 2000 rpm, 4500 rpm and 8000 rpm. This shows that by increasing the recovery, so is the flight time T and the angle of descent β.

There is a strong correlation between the duration of the flight and the angle of descent. This follows from the fact that a ball that remains in the air for longer, compared to another ball that reaches the same distance, must rise higher in the air, and therefore will have a steeper descent. In the absence of wind and provided that all the balls of an installation have aerodynamic properties that have a very close coincidence, the flight time T of any ball can be determined with great precision simply by measuring the angle of descent β. In general, wind and surface roughness variations (which affect aerodynamic properties) will all affect T.

The roughness of a surface can be both intentional (such as the pattern of dimples of a golf ball molded on the surface during its manufacture) and accidental (for example, the roughness caused by cuts, abrasion, surface contamination and the like). The relationship between surface roughness and aerodynamic properties is very complex, but for simplicity, it will be assumed herein that an increase in surface roughness is accompanied by an increase in the intensity of the drag and drag aerodynamic forces. The aerodynamic forces acting on a ball are normally proportional to the square of the speed that the ball carries through the surrounding air

One aspect of the present invention is based on the realization that, in the first order, wind, recovery and surface roughness all have a similar effect on the flight of the ball. The winds against have the effect of increasing the upward and drag forces on a ball, because they increase its speed with respect to the surrounding air, while the winds in favor have the opposite effect. Thus, the common effect of a high recovery, intense winds against and high surface roughness is an increase in the duration of flight T and an increase in the angle of descent β for a certain distance of reach. Reciprocally, a recovery low, some winds in favor of low intensity and a small surface roughness are comparable to short flight durations and small angles of descent.

Figures 4 (a) and 4 (b) are graphs of flight duration T as a function of the angle of descent β, with data obtained by simulation.

In graph 4 (a), all the hits reached 91.4 meters (100 yards) with a wind against which varied in a random way from 3.5 m / s to 8.5 m / s with an average value of 6 m / s The launch angles varied randomly from 27.0 to 33.4 degrees, and the initial recovery also varied randomly with an average value of 2710 revolutions per minute (hereinafter rpm). For each hit, the throwing speed was adjusted to give 91.4 meters of range with the average throwing speed being 36.7 m / s.

The data for graph 4 (b) were obtained using the same random values of launch angles, but aided by the variation of winds in favor with an average of 6 m / s. The average launch speed required to achieve a range of 91.4 meters with reduced winds up to 33.6 m / s, with a corresponding lower average recovery of 2500 RPM. The graphs show that the winds against it significantly increase the duration of the flight and also cause a greater dispersion in the data compared to the winds in favor.

From the graphs of Figures 4 (a) and 4 (b) it is clear that the relationship between T and β is almost linear. Therefore, an optimal line of fit across each data set would give a very accurate estimate of T as a function of β. Therefore, a simple equation can be constructed to calculate an estimate of the expected duration of the Edurk flight as a function of β, where k takes a value of 1 to K and K is the total number of “possible blows” determined by the process of ID.

Edurk = C1k + C2k x β (1)

Equation (1) either expressed in the previous form or in an alternative form that is not known, nevertheless provides very precise estimates of the duration of Edurk flight, regardless of the roughness of the ball, prevailing wind and m, recovery imparted .

The constants C1k and C2 k could simply be determined as a function of the distance of the range from the k-th starting point to the descent position, but, especially for large range distances, they are preferably determined from several parameters including range distance, wind speed, wind direction, launch angle, launch speed, and air density.

It is preferable to choose C1k and C2k values that are specific for a small range of β values and preferably for a small range of ground wind speed measured by at least one anemometer. The winds against produce the maximum errors (compared to the conditions of calm wind or winds in favor) because this increases the aerodynamic effects and increases the duration of the flight for a given range length. This is evident in the graphs of Figures 4 (a) and 4 (b), which correspond to winds against and winds in favor, respectively.

The wind, measured at a fixed point, is rarely stable, but usually fluctuates with typically gusts of 1.3 to 1.6 times the average value and with a duration of gusts as small as two or three seconds (it is say, less than half of a flight duration of a typical hit with effect). In addition, Taylor's frozen turbulence hypothesis teaches that wind turbulence is transported along the total average wind direction and at the total average wind speed. Thus, if the average wind speed is 5 m / s, a wind gust of intensity and duration given at point A will be reproduced exactly two seconds later at a point B that will be 10 meters in favor of the wind from point A. This "frozen turbulence" has important implications in relation to the effect of wind on the flight of a ball. The speeds of golf balls are much greater than the average speeds of winds experienced in a driving range. Accordingly, the ball flies rapidly through gusts of wind, such that increases and decreases in wind gust intensities along the flight path are averaged. Thus, with a gust wind, the instantaneous wind speed experienced by a ball changes several

5

10

fifteen

twenty

25

30

35

40

Four. Five

fifty

times throughout the flight, but the average wind speed for the first half of a ball's flight is usually almost the same as the average wind speed during the second half of the flight. This is especially the case for winds against gusts. Therefore, in order to simulate the effect of the wind against the flight of a ball, it is reasonable to assume that the balls experience a constant wind speed throughout their entire flight.

Using simulation, values for C1k and C2k can be obtained at different range distances. Table 1 shows the results of simulation studies where the winds against each other varied randomly from 3.5 to 8.5 meters per second and where the launch angles and recoveries were varied randomly to produce a dispersion. of values for β. The launch speeds were adjusted to give the set range distance.

Table 1

Range distance in meters (yards)

Duration Min / Max. (seconds) Β Min / Max. (degrees) C1k (seconds) C2k (seconds) Σ (ms)

50 ((47)

2.7 / 3.3 35.7 / 47.8 0.94 0.048 27

* 100 (94)

4.4 / 5.9 45.9 / 63.0 0.97  0.076 70

200 (188)

4.3 / 6.4 24.8 / 49.7 2.2  0.085 86

* Data in Figure 4 (a)

Note that, in general, the values of C1k and C2k vary with the dependence of k, but, to generate the data for Table 1, blows with exact range distances were simulated as shown, and prevailing winds with a average value of 6 meters per second; so, in these circumstances, C1k and C2k have fixed values for each of the three range distance simulations.

The last column of Table 1 gives the standard deviation of the errors estimated in milliseconds. The errors are very small and only require a measure of β and an approximate measure of the prevailing wind conditions. Knowledge of launch speeds, recoveries or launch angles is not required. To take advantage of these estimation accuracies, the value of β must be measured very accurately, since the value of the constant C2k is typically of the order of 40 to 90 milliseconds per degree. In this way, a measurement error of 1.0 degrees in β will itself generate an estimation error of 40 to 90 milliseconds. Therefore, it is preferable that the measurement of the descent parameters implies errors with a standard deviation of less than 1.0 degrees and more preferably less than 0.5 degrees for the measurement of the descent angle in descent of a ball.

Since errors in flight duration estimates are generally very small, the identification of the location of the impact of a descending ball can be determined in many cases simply from the calculation of this parameter. This makes possible a very fast identification.

Figure 5 is a plan view of a typical golf facility with exit posts arranged in a circular arc 50 and a distant target 51 located in the furthest part of the field. In this example, the objective could be 100 meters to 200 meters, or more, of the starting points, so it could happen that the ball took significant effects to the right and left. In a hypothetical scenario, two golfers in adjacent positions 52 and 53 hit simultaneously with throwing speeds, elevation angles and almost identical recoveries, so that their shots reach equal distances and land simultaneously, but the shot from the post of Exit 52 is hit with effect to the right, while the shot from the starting point 53 has no side effect and there is no side wind that influences either of the two flight paths.

The full line 54 shows the path in azimuth of the stroke from the starting point 52, and the full line 55 shows the path in azimuth of the stroke from the starting point 53. The stroke 54 descends within the objective 51 at point 56, while the blow 55 descends on the outside to the right of the target 57. However, if the shot 54 had been hit with a side effect on the left, as shown with the dashed line 58, it could have descended exactly on the same point (56) than shot 54.

A central computer (which has not been shown) has to identify the starting position from which the ball that originated in point 56 was thrown. In this case, both starting positions have been selected as probable candidates, having been assigned at exit point 52 a k = 2 and at exit position 53 a k = 1. However, the actual impact times Dur2 and Dur1 correspond to the estimated Edur2, but the differences measured in azimuth angles provide the means of selecting exit post 52 and reject exit post 53.

In the starting positions, the initial azimuth throwing angle of each shot is measured, and in each objective the azimuth angle of descent α of every ball that reaches a target is measured. All angles are measured with reference to a fixed direction and in a dextrógira or levógira rotation. Figure 5 shows the angles measured with reference X away in a dextrogyral rotation. For each point of impact and each position of descent, there is an additional angle to what will be called αF, which is the angle subtended by a hypothetical straight line drawn between the point of impact and the position of descent (shown for shot 54 in the Figure 5 as dashed line 59).

Since the shot 55 is straight, the initial launch angle, the final angle and the angle of descent are all the same. For shot 54, the effect on the right causes the azimuth path to curve in an approximately circular arc, so these angles are all different. The azimuth path is not exactly circular, especially if the side winds affect the flight path, but in a surprisingly large number of cases, it has been found out (from simulation) that (αF - αL), is the same with many approximation to (αdes - αF), which corresponds to a circular arc.

This relationship can be used to accurately estimate the Edirk azimuth direction of a ball from measurements of αFk and αLk. For a given hit, a close coincidence between the Edirk value obtained from a given exit post and αdes indicates a high probability that said hit will be correctly coupled in conjuction with the said determined exit post. Referring again to Figure 5, as the value of Edir2 as estimated for the starting point 52 coincides closely with the measured value of αdes, the central computer can correctly identify the starting point 52 as the origin of the stroke 54, while Edir1 does not correspond to the measured value of αdes.

In general, Edirk is obtained from the following equation:

Edirk = αFk + C3k x (αFk - αLk) (2)

It has been found, from the simulation, that making C3k equal to 1.0 generally gives good estimates for the majority of possible golf shots. More preferably, C3k depends on the range distance, and at least one of the following parameters: intensity and direction of the side wind, range deviation, total speed and direction of the wind, elevation angle at launch, speed of launch, range duration and air density.

Optionally, an anemometer 60 is positioned near the exit posts 50 and a second anemometer is placed at the far end of the furthest part of the field. The anemometers preferably use 2-axis sonic detection means capable of decomposing very short-term variations in wind speed and direction. Anemometer data, which could be recorded at a sampling rate of 10 samples per second or greater, is used to calculate the approximate wind speed and direction along the farthest part of the field by an interpolation and extrapolation procedure. . The anemometers are preferably mounted on posts at a height of 10 meters or so or at the expected average height of the golf shots. In other arrangements, only a 2-axis anemometer is used to obtain an approximate estimate of the current wind direction and intensity, or several anemometers, which could be of the 2 or 3-axis types, are used to obtain more accurate estimates. . Special anemometers, built on purpose, can be provided when the length of the range for sonic measurement extends several meters instead of a few centimeters as usual in commercially available devices.

Figure 6 is a graphical representation of errors in Edirk as a function of the side wind speed for simulated blows. Each hit of the sample reached 220 meters (240 yards) and was launched with random elevation angles and random recovery. Additionally, each stroke had a large side effect component (random) that was sufficient to result in an average deviation of 27 meters in calm wind conditions. The side wind varied from - 4 m / s to + 12 m / s (where the negative side wind helped deviate from the side effect and vice versa). An optimal adjustment line 62 passes through the data points and shows the general trend of errors. In low-intensity side winds, errors are negligible, but there is a gradual increase in errors as the intensities of the side wind increase. This gradual increase is less than 0.25 degrees per m / s of side wind. It has been found that the standard deviation of the errors with respect to the optimum adjustment line 62 is less than 0.25 degrees. This is almost one hundredth of the average angular deviation caused simply by the imparted side effect. Therefore, it is evident that equation (2) combined with side-wind corrections (if measured) can provide an extremely accurate estimate of Edirk, even when there is a side effect and a high intensity side wind.

To ensure that there is very little chance of mismatch, it is preferable to include, as a possible match, any starting point where the difference between the actual impact time and the duration of the estimated Edurk range is less than 3- sigma or even less than 4- sigma. This will sometimes result in two

or more exit posts are included as possible matches even when the 3- sigma deviation has a value of one tenth of a second or so. So, Edirk is used to make a selection between the two mentioned

or more starting points. An even greater certainty of a correct match can be provided by including a third parameter in the matching process that establishes a match

of the speed of descent of a blow with a probable launch speed. It has been found that the relationship between the horizontal velocity of a ball in the pitch and its horizontal velocity in the descent (which will be referred to herein as the "deceleration ratio") can be accurately estimated even when the ball has decelerated considerably. along the length of its reach.

Figure 7 is a graphic representation of simulated golf shots showing the variation in acceleration ratio as a function of the duration of the range. All strokes in Figure 7 were subjected to random values of launch angle, recovery and wind, with launch speeds set to give an equal range distance of 220 meters (240 yards). The variations in the initial launch parameters were: speed 69.4 to 77.2 m / s, recovery 1010 to 3140 RPM and launch angle at elevation 12 to 14 degrees. Additionally, the wind along the direction of the flight varied in the range of ± 1.3 m / s (± 3 miles per hour). These data represent very long blows, typical of championship players or very good amateurs.

From Figure 7 it can be seen that there is a good correlation between deceleration ratio and flight duration, which varied from 5.2 to 6.8 seconds. Line 70 is the optimal linear fit through the data. The equation for this line, shown below, gives the optimal value of an estimate for the Edeck deceleration ratio for the k-th “possible starting position” based on the possible flight duration ”Durk.

Edeck = C4k + C5k x Durk (3)

The constants C4k and C5k depend mainly on the launch elevation angle, distance of the range obtained, wind speed, wind direction, and air density.

Despite the rather wide range of launch conditions, it has been found that the distribution of errors between Edeck and the actual data in Figure 7 has a standard deviation of only 3.2%. For the vast majority of hits, Edeck is even closer to the real one. This is especially true for hits with significantly lower launch speeds, since the deceleration ratio then approaches the unit and the errors become negligible. Thus, as long as reasonably accurate measures of horizontal launch speed, launch elevation angle and horizontal descent speed are available, the differences between Edeck and possible deceleration ratio matches at various exit positions provide A very reliable means of identifying golf shots. Edeck accuracy is further improved if measures of wind speed and direction are available.

The blow identification process is based on evaluating the probability that the actual impact time and additionally (if required) the actual launch direction or the actual horizontal launch speed, of one of a plurality of blows from the Exit positions, corresponds to the parameters of descent of a ball. To estimate this probability, the data of a sample of previous strokes were first analyzed to obtain Edurk, Edirk and Edeck, which are the values of the three relevant parameters that have the highest probability density. The distribution of the results above and below Edurk, Edirk and Edeck then gives the values for the three corresponding standard deviations of errors from which the probabilities can be estimated. It is not practical to determine Edurk, Edirk and Edeck and their typical deviations for any possible combination of descent parameters, wind parameters and launch parameters, but an analysis of the data can generate a model in such a way that the values for any specific set of conditions can be determined from interpolation values of query tables or other calculation means.

With the very small differences between Edurk, Edirk and Edeck and the real parameters for “blows with very close possibility”, the error distributions have the possibility of being symmetrical and normal, so that the probabilities can be obtained using standard formulas.

For example, the probability that the impact time of the k-th hit of a set of possible hits is the real time when a falling ball is hit is given by:

P (Durk) = 1 - 2 x [F (Durk) - 0.5] (4)

Where F (Durk) is the cumulative probability function for Durk. The probabilities for direction and deceleration are obtained with a similar method.

Referring now to Figures 8 and 9, a flagpole 80 marks the center of a circular target 81, which is located in the center of a golf driving range and typically at a distance of 50 to 250 meters from a row of Several starting points (which have not been shown). Several of these objectives could be arranged according to the length of the field and could vary in size and shape. The target limit could be marked so that it was visible to players at the starting points, or objective 81 could be designed as a circular zone on the furthest part of the field with a fixed radius from flagpole 80 .

Two sensor units 82 and 83 are rigidly fixed slightly above the ground level by means of supports 90 and positioned to the right of the target and preferably near the limit of the far right part of the furthest area of the field where it is not expected Let the golf balls land. A reflector structure 84 runs along the length of the field (or of a shorter length, if required) on the opposite side of the lens on the far left side where golf balls are not normally expected to land.

The positions of the sensor units and the reflector structure could be reversed or, if preferred, sensor units and reflector structures could be installed on both sides of the lens.

Each of the sensor units comprises an upper light emitter 91 and a lower light emitter 92 mounted above and below a light receiver 93 that acts together. Each of the light receivers for sensor units 82 and 83 has angular horizontal fields of view, designated by dashed lines 85 and 86 respectively, which are sufficient to encompass through objective 81 (for example 45 degrees

or so) but with the directions of the deflected axes as shown. The light emitters for the sensor units 82 and 83 have horizontal light radiation fields that extend at least beyond the field of view of their joint action light receivers. Both light emitters 91 and 92 in each sensor unit have vertical light radiation fields that diverge only slightly, but enough to illuminate two separate retro-reflector bands 94 and 95 that are fixed to the front side of the reflector structure. . The vertical separation HR of the retroreflective bands 94 and 95 and the vertical separation HE of each pair of light emitters 91 and 92 are preferably equal and, without limitation, in the range of 10 to 50 centimeters.

Preferably, the light emitters are cyclically activated such that both upper light emitters 91 are ON and both lower light emitters 92 are OFF in a half cycle and vice versa in alternating half cycles. During each half cycle, the light reflected from a pair of light emitters is reflected from the retro-reflective bands 94 and 95 and is received by the light receivers 93, which preferably have large openings and a high gain optical condensation system. The light receivers 93 focus the aforementioned light reflected on one or more linear light sensor clusters (which have not been shown) within the light receptor subsystem.

The retro-reflective bands 94 and 95 are of a uniform vertical width in the range of 20 to 30 millimeters, for example, 25 millimeters, a value that is slightly smaller than the diameter of a golf ball, and the vertical opening of the Light emitters are designed to have the same height as the retro-reflective bands 94 and 95, for example, 25 millimeters.

When the light emitters 91 are ON, a typically infrared fan beam illuminates a section of the retro-reflective bands 94 and 95, and this light beam contains two light paths designated by dashed lines 96 and 97. When a golf ball enters the light path96, part of the light coming from the retro-reflector94 is interrupted, and this is detected in both sensor units 82 and 83. In each light receiver, at least one pixel of the grouping of Light sensors will detect the interruption of light. Note that the aforementioned "at least one pixel" will still receive light from the retro-reflector 94 through path 97, whereby the signal in said at least one pixel changes by - 6 db and this value is sufficient to detect a Reliable way the presence of a golf ball within the path of light. The angular position of the golf ball that interrupts the light with respect to both sensor units 82 and 83 can be determined by the respective position of the detection pixels. The instantaneous height of the golf ball that interrupts the light is determined by the height of the luminous path 96. Therefore, by knowing the angular positions of the golf ball with respect to the two sensor units and their height, their height can be determined Instant position in three dimensions. As the ball passes through the four light paths 96 to 99, its velocity vectors can be measured. Note that the intermediate light paths 97 and 98 are not required to measure the velocity vectors of the ball, but they are nonetheless present, since it is not practical to focus the fan beam of the light emitter 91 on only the retro -reflector 94 but not on 93, and similarly it is not practical to focus the fan beam of the light emitter 92 only on the retro-reflector 95

The flagpole 80 may be provided with striking indicator means 87 comprising three beacons of different color that are switched ON, in accordance with the proximity to the flagpole of the detected golf balls. This provides a means of granting "Scores" depending on the accuracy and distance of the different golf shots.

A modified version of the sensor units and retro-reflective bands of Figures 8 and 9 could be used to measure launch parameters at the starting points. In this case, the light of the fan beams and the fields of view of the light sensors are preferably oriented in a vertical plane.

Claims (15)

one.

A method for identifying, with respect to a descending ball, the position between a plurality of throwing positions (52, 53) from which the ball has been thrown, characterized in that the method comprises measuring the descent parameters of the descending ball to obtain at least the time of descent of the descending ball and a measure dependent on its angle of descent in elevation, calculate an estimate of the duration (Edur) of flight of the descending ball based on the measure of the angle of descent, measure , with respect to each position (52,53) of throwing, the time interval (Dur1, Dur 2) elapsed between the throwing of a ball from that position and the time of descent of the descending ball, compare the time interval ( Dur1, Dur2) measured with respect to each launching position (52, 53) with the calculated estimate of the duration (Edur) of the flight of the descending ball to determine for which of the positions tion (52, 53) of throwing the respective interval is a very close coincidence with the estimate (Edur) calculated, and identify the position from which the descending ball was thrown with the position (52, 53) of throwing for which it exists the very near coincidence.

2.

A method according to claim 1, wherein the measured descent parameters include measures of component descending velocities of the descending ball to give the angle of descent in elevation.

3.

 A method according to claim 1 or claim 2, wherein the estimated flight duration is calculated according to a substantially linear function of the angle of descent.

Four.

A method according to claim 3, wherein the substantially linear function is the sum of at least two terms, the first of which is a constant and the second of which is the product of a constant by the measure dependent on the angle of decline.

5.

A method according to claim 4, wherein the first and second terms both depend on at least one of the parameters of: range, wind speed, wind direction, launch angle, launch speed, air density , speed of descent of the descending ball, and the angle of descent in azimuth of the descending ball.

6.

A method according to any one of claims 1 to 5, wherein a measure of the azimuth descent angle of the descending ball is obtained from the measured descent parameters, the step of identifying the position from which the ball has been thrown descending includes the stage whose purpose is said identification to discriminate between the throwing positions of the balls thrown for which there is a very close coincidence, whose discrimination is based on the extent to which with respect to each of these last balls there is a coincidence between the measure of the angle of descent in azimuth of the descending ball and an estimate calculated for it, and in which this estimate is calculated as a function of a measure of the angle of launching in azimuth of the ball in question and a measure of marking the descending ball from the throwing position of the ball in question.

7.

A method according to claim 6, wherein the calculated estimation of the azimuth descent angle of the descending ball is the sum of the measure of the marking of the descending ball from the throwing position of the ball in question and a dependent term of the difference between this marking and the measurement of the azimuth throwing angle of the ball thrown in question.

8.

A method according to any one of claims 1 to 7, wherein a measurement of the horizontal velocity of the descending ball is obtained from the measured descent parameters, the step of identifying the position from which the ball has been thrown descending includes the stage whose purpose is said identification to discriminate between the throwing positions of the balls thrown for which there is a very close match, whose discrimination is based on the extent to which each of these last balls exists a match between: (a) a calculated ratio of a measure of the horizontal velocity of that ball at the throw and the measure of the horizontal velocity of the descending ball, and (b) a function dependent on a value for the expected flight duration of the ball thrown in question from its throwing position.

9.

A method according to any one of claims 1 to 8, wherein the measurement of the descent parameters of the descending ball is made less than 2 meters above the ground level.

10.

 A system for identifying with respect to a descending golf ball the position between a plurality of launch positions (52, 53) from which the ball has been thrown, characterized in that the system comprises means for measuring parameters of ball descent descending in order to obtain at least the time of descent of the descending ball and a measure dependent on its angle of descent in elevation, means to calculate a duration estimate (Edur) of the flight of the descending ball as a function of the measure of the angle of descent, means for measuring with respect to each position (52, 53) of throwing the time interval (Dur1, Dur2) between the throwing of the ball from that position and the time of descent of the descending ball, means for compare the time interval (Dur1, Dur2) measured with respect to each launching position (52, 53) with the calculated estimate of the duration (Edur) of the ball flight

descending in order to determine for which of the positions (52, 53) of throwing the respective interval is in a very close coincidence with the calculated estimate (Edur), and means to identify the position from which the descending ball has been thrown with the launch position (52, 53) for which there is a very close match.

eleven.

A system according to claim 10, wherein the measured descent parameters include measurements of component descending velocities of the descending ball to give the angle of descent in elevation.

12.

 A system according to claim 10 or claim 11, wherein the estimated flight duration is calculated according to a substantially linear function that is the sum of at least two terms, the first of which is a constant and the second of which is the product of a constant by the measure dependent on the angle of descent, and in which the first and second terms both depend on at least one of the following parameters: range distance, wind speed, wind direction , throwing angle, throwing speed, air density, descent speed of the descending ball, and azimuth descent angle of the descending ball.

13.

A system according to any one of claims 10 to 12, wherein a measurement of the azimuth descent angle of the descending ball is obtained from measured descent parameters, the identification of the position from which the ball has been thrown descending includes, for the purpose of such identification, discriminating between the throwing positions of the balls thrown for which there is a very close match, whose discrimination is based on the extent to which with respect to each of these last balls there is a coincidence between the measure of the angle of descent in azimuth of the descending ball and an estimate calculated for it, and in which this estimate is calculated as a function of a measure of the launch angle n azimuth of the ball thrown in question and a measure of marking the descending ball from the throwing position of the ball in question.

14.

A system according to claim 13, wherein the calculated estimate of the azimuth descent angle of the descending ball is the sum of the measure of the marking of the descending ball from the throwing position of the ball in question and a term depending on the difference between this marking and the measurement of the azimuth throwing angle of the ball in question.

fifteen.

A system according to any one of claims 10 to 14, wherein a measurement of the horizontal velocity of the descending ball is obtained, from the measured descent parameters, the identification of the position from which the ball has been thrown descending includes, for the purpose of such identification, discriminating between the throwing positions of the balls thrown for which there is a very close match, whose discrimination is based on the extent to which each of these last balls exists a match very close between: (a) a calculated relationship between a measure of the horizontal velocity at the throw and a measure of the horizontal velocity of the descending ball, and (b) a function dependent on a value for the expected flight duration of the ball thrown in question from its throwing position.