WO2019008361A1

WO2019008361A1 - Improvements to a ball tracking system, method and apparatus

Info

Publication number: WO2019008361A1
Application number: PCT/GB2018/051889
Authority: WO
Inventors: Steve AVERY
Original assignee: Avery Steve
Priority date: 2017-07-05
Filing date: 2018-07-04
Publication date: 2019-01-10
Also published as: GB2564394A; GB201710817D0

Abstract

A system for determining the location of impact of a ball on the boundary of a playing field or for determining the location where the ball crosses said boundary, the system comprises a primary detection mechanism for detecting the position and/or movement of a ball within said playing field; and a secondary detection mechanism which is configured to validate the impact of the ball on said boundary and/or validate where the ball crosses said boundary.

Description

Improvements to a Ball Tracking System. Method and Apparatus

Field of the Invention The invention relates to a system, method and apparatus for detecting a ball during a sporting occasion. In particular, the invention relates to a system, method and apparatus for accurately recording the location a ball leaves a playing area.

Background to the invention Modern sports are driven by volumes of data and statistical analysis. Nowhere is this drive for statistics more apparent than in cricket. Every ball of every game will be recorded and analysed to provide information on all aspects of the game. These statistics are used both in broadcasting to improve the viewing experience but are also invaluable tools for tactical analysis and coaching. One such statistic currently created, is an indication of the areas a particular batsman likes to hit the ball. To achieve this the ground is theoretically divided into segments, with each boundary hit through any given segment recorded as a running total. However, the current technology used to record the location of the batsmen's shots is predictive and has several limitations in terms of reliability. Current systems for recording the location of a batsman's shot, such as Hawkeye (trade mark), are only accurate to several millimetres. As a result, any statistics derived using this method of ball tracking can at best be an estimation. For example, if the ball is hit through what could be considered the division between 2 segments, it is difficult to determine accurately which segment it went through. In order to improve the accuracy of statistics available for both the viewing public and the players themselves, the current invention provides an improved method of recording the segments in which each boundary is scored.

Aspects of the invention seek to further improve on existing systems.

Summary of the Invention

In a first broad independent aspect, the invention provides a system for determining the location of impact of a ball on the boundary of a playing field or for determining the location where the ball crosses said boundary, the system comprising a primary detection mechanism for detecting the position and/or movement of a ball within said playing field; said primary detection mechanism optionally detecting the impact of the ball on said boundary and/or detecting where the ball crosses said boundary; and a secondary detection mechanism which is configured to validate the impact of the ball on said boundary and/or validate where the ball crosses said boundary. This configuration is particularly advantageous since it improves the accuracy of the determination of the exact location of the ball relative to the boundary. In certain embodiments, it removes the zones of uncertainty allowing the result to be reliably validated for further processing. The determination of a section against which the ball has impacted is improved allowing all aspects of the game to be improved which will also potentially yield commercial advantages although these are deemed to be secondary to the improvements in accuracy achieved in the various embodiments that follow.

In a subsidiary aspect, said primary detection mechanism comprises at least two cameras to detect the movements of a ball. In a subsidiary aspect, said secondary detection mechanism is integral with the boundary.

In a subsidiary aspect, said secondary detection mechanism is located adjacent to said boundary.

In a subsidiary aspect, said boundary is substantially triangular in cross-section and said secondary detection mechanism is located at least partially within the boundary. This allows improved protection of the secondary detection mechanism which could otherwise be damaged on impact.

In a subsidiary aspect, the boundary is segmented in a plurality of discrete sections; said secondary mechanism is configured to validate which section said ball primarily impacts or crosses.

In a further subsidiary aspect, the system further comprises a processor in communication with a plurality of cameras, capable of receiving signals from said cameras and predicting the flight path the ball will take; a data store; means for comparing the predicted flight path to a model of a playing field with pre-defined segments; and means for activating said secondary detection mechanism when a ball is predicted to pass within a zone of uncertainty defined between two discrete sections of a boundary. This embodiment is particularly advantageous since it would allow the secondary validation to operate in a low power mode which would be particularly efficient whilst achieving the beneficial accuracy outlined above. The zone of uncertainty may be defined as the width of a typical cricket or baseball ball on either side of a line of separation. The zone of separation may be no more than 10 centimetres either side of a line of separation between discrete sections of a boundary. One advantage of this secondary mechanism is the ability to determine the percentage of the ball which is on either side of the line of separation in order to determine where the majority of the ball has impacted.

In a further subsidiary aspect, said secondary detection mechanism comprises a sensor suitable for determining the exit location of a ball.

In a further subsidiary aspect, said secondary detection mechanism comprises a sensor which is housed within said boundary.

In a further subsidiary aspect, said secondary detection mechanism comprises a sensor which is housed within a zone of uncertainty defined between two discrete sections of a boundary (or in other words defined by adjacent end portions of a boundary section). This configuration deals with accuracy with problem areas of the boundary for more accurate determination of impact.

In a further subsidiary aspect, said secondary detection mechanism comprises one or more cameras.

In a further subsidiary aspect, said one or more cameras are angled upwards along lines of division between discrete sections of the boundary. This is particularly advantageous in certain embodiments as it would allow the determination of a percentage of the ball either side of a line of separation between adjacent sections of a boundary.

In a further subsidiary aspect, said secondary detection mechanism comprises one or more light emitters.

In a further subsidiary aspect, said secondary detection mechanism comprises a pressure sensor suitable for detecting contact between the ball and the boundary. This

configuration might allow further aspects of the impact to be determined such as the force of impact in addition to its location.

In a further subsidiary aspect, the secondary detection mechanism is configured to be dormant until activated by said primary detection mechanism.

In further independent aspect, the invention provides a playing field boundary comprising a secondary detection mechanism which is configured to validate the impact of the ball on said boundary and/or validate where the ball crosses said boundary.

In a subsidiary aspect, said boundary is substantially triangular in cross-section and said secondary detection mechanism is located at least partially within the boundary.

In a further subsidiary aspect, the boundary is segmented in a plurality of discrete sections; said secondary mechanism being configured to validate which section said ball primarily impacts or crosses.

In a further subsidiary aspect, said secondary detection mechanism comprises a sensor suitable for determining the exit location of a ball. In a further subsidiary aspect, said secondary detection mechanism comprises a sensor which is housed within said boundary.

In a further subsidiary aspect, said secondary detection mechanism comprises a sensor which is housed within a zone of uncertainty defined between two discrete sections of a boundary.

In a further subsidiary aspect, said one or more cameras are angled upwards along lines of division between discrete sections of the boundary.

In a further subsidiary aspect, said secondary detection mechanism comprises a pressure sensor suitable for detecting contact between the ball and the boundary.

In a further subsidiary aspect, the secondary detection mechanism is configured to be dormant until activated by a primary detection mechanism.

In a further independent aspect, the invention provides a method for determining the location of impact of a ball on the boundary of a playing field or for determining the location where the ball crosses said boundary, the method comprises the steps of detecting the position and/or movement of a ball within said playing field; and validating the impact of the ball on said boundary and/or validating where the ball crosses said boundary.

In a further subsidiary aspect, the boundary is segmented in a plurality of discrete sections; and the method comprises the further step of validating which section said ball primarily impacts or crosses.

In a further subsidiary aspect, the method comprises the further steps of providing a processor in communication with a plurality of cameras, receiving signals from said cameras, predicting the flight path the ball will take; comparing the predicted flight path to a model of a playing field with pre-defined segments; and activating a validation step when a ball is predicted to pass within a zone of uncertainty defined between two discrete sections of a boundary.

Brief Description of the Figures Figure 1 is a first schematic of camera locations in overhead view.

Figure 2 is a second schematic of camera locations in overhead view.

Figure 3 is a diagram of potential playing field segmentation with cricketing nomenclature provided.

Figure 4 is a flow diagram of the steps required to accurately predict the flight path of a ball.

Figure 5 is a first embodiment of a boundary rope in side view. Figure 6 is a second embodiment of a boundary rope in side view. Figure 7 is a third embodiment of a boundary rope in side view. Figure 8 is a fourth embodiment of a boundary rope in side view. Figure 9 is a fifth embodiment of a boundary rope in side view.

Detailed Description of the Figures

An embodiment defines a system for use during a sporting game such as cricket, to determine the location a ball exits from a pre-determined area, or playing field, within pre- set boundary segments. In a preferred embodiment, the boundary segments refer to literal portions of a boundary rope which are distinct from one another and provided adjacent from one another to collectively form the boundary of the playing field. In alternative embodiments, the boundary segments could refer to theoretical boundary division lines super imposed on the playing field. In certain embodiments, the ball may exit a playing field through a zone of uncertainty. The zone of uncertainty may be defined as the width of a typical cricket or baseball ball on either side of a line of separation. The zone of separation may be no more than 10 centimetres either side of a line of separation between discrete sections of a boundary. In other embodiments, the zone of uncertainty has a dimension which depends on the size of the ball itself. The secondary recording devices may assist in accurately determining what percentage of a ball impacts on each side of a line of separation between adjacent segments of a boundary.

To provide clarity, the components of the system which are common to different figures have retained identical numerical references throughout all figure descriptions.

In a first embodiment, the system includes a plurality of cameras, with a minimum of two cameras but ideally between three to six cameras. Figures 1 and 2 provide an overhead schematic of the optimal camera positions when using a system with three or six cameras. Using two cameras will provide sufficient information to triangulate a 3D location of the ball from the 2D images. However, using three or more cameras provides some

redundancy should the line of sight of one or more of the cameras be blocked, for example by fielders or wildlife.

The cameras are positioned evenly around the ground to have the greatest ability to triangulate the exact 3D location from the 2D images produced from each camera. The cameras will optimally be positioned in a plane above that of the playing field - such as the underside of the roof - to reduce the possibility that the line of sight of the ball will be blocked by players. Furthermore, being positioned at an increased height, reduces the possibility that the flight of the ball will take it out of the line of sight of the cameras which are in fixed locations.

Referring to the flow diagram of Figure 4, it is possible to track the steps required for the system to provide an accurate ball flight path. The cameras will be set up around the playing area based on the teachings of Figures 1 and 2, and calibrated for use 1. During the initiation of the processing procedure 2, the cameras feed data to a central processor. The central processor has access to a data storage which contains information appertaining to ball size, expected ball velocities and historically recorded ball flight paths. Using this information, the system is able to accurately identify the pixels of each 2D time frame image which represent the ball 3. For each time frame, the system will merge the 2D images of the ball from each camera 4, triangulating the data to produce a 3D location 5. This process requires a minimum of 2 images, however, the location accuracy increases as more 2D images from alternative angles are included in the triangulation. The 3D location must be mapped for at least 3 separate time frames, at which point, the processor will produce a flight path between those data points 6. Incorporated into the processor is a module for applying laws of physics to the predictive ball model. Using this module and the data from the historically recorded ball flight paths, the processor will extrapolate the initial flight path to predict the entire theoretical ball flight path 7. By using data from historically recorded ball flight paths - which will increase as the system is used - the system is able to improve its accuracy with each new ball flight measurement taken.

Producing a theoretical flight path whilst the ball is still in flight has a number of benefits for producing an accurate ball flight path model. Firstly, the processor has an indication of where the ball is expected to be during any time frame, making identifying the pixels representing the ball easier. Secondly, it allows the system to identify regions of the flight path that would be of particular importance to record. For example, any time the ball bounces on the ground, the large variety of influencing factors - including; the spin on the ball, moisture content and general architecture of the ground - make predicting the flight of the ball after this interaction particularly difficult. As such, the system will target several time frames during and immediately after any such interactions 8, to provide the greatest model of ball flight path. All time frames recorded will undergo the same processes of pixel analysis 3, image meshing 4, and 3D triangulation 5, as previously described. These subsequently determined 3D ball locations will be compared to the locations predicted by the initial flight path model 10. Where a disparity exists between the predicted ball locations and the recorded ball locations, the flight path model will be adjusted to incorporate the new data point 1 1. In the current embodiment, the cameras and processing means required to develop the predictive ball flight path constitutes the primary detection mechanism. This primary mechanism produces a prediction of which boundary segment the ball will leave the playing area through.

When the ball has finished its flight trajectory and the most accurate model of ball flight path produced, the location the ball exited the playing area will be recorded. This data (the runs scored and/or the number of boundaries) will be assigned to one of a plurality of pre- defined boundary segments based on the location of the ball as it exited from the playing area. The location of the ball as it exits the playing field, herein referred to as its exit location, is predicted based on the primary detection mechanism. Should the ball exit the playing field within a threshold distance of the division between segments, henceforth referred to as the zone of uncertainty, the system will activate a secondary detection mechanism to accurately record which segment the ball exited through. The secondary detection mechanism accurately records the location the ball leaves a playing field with particular emphasis on the exact boundary segment the ball exited the playing field. This information is compared to the information provided by the primary detection mechanism either validating that the ball did indeed exit through the predicted boundary segment or disagreeing with the predicted model and stating that it exited through an adjacent boundary segment. In an embodiment, the system records whether the ball hit the ground between being hit by the batsman and crossing the boundary rope. In such a way, the system can record both the number of boundaries scored in each segment and also the total number of runs sub- categorised by the number of 4s and number of 6s. When the system records that the ball has crossed the boundary rope within a zone of uncertainty, it activates a secondary method of determining which segment of the boundary the ball or the majority of the ball exited through. Provided are several potential secondary methods of determination, with the invention not limited by combination with anyone of the methods.

Figure 5 shows a boundary rope with a first embodiment of a secondary determination means incorporated. The boundary rope 20 as shown, is divided into adjacent segments 21 and 22 with dividing line 23. Either side of the dividing line 23 is a theoretical region referred to as the zone of uncertainty 24. The zone of uncertainty 24 could be altered manually or be a pre-set size based on the inherent error margin of the predictive methodologies. This zone of uncertainty may also be adjusted by the pre-determined size of a ball. In the displayed embodiment, the secondary means of determination is a light beam and sensor apparatus housed within the boundary rope 20. Two light beam generators or lasers 25 are housed either side of the dividing line 23, within the zone of uncertainty 24. Both light beam generators are connected to a battery source 27 and the central processor 30. The light beam generators 25, fire a beam of light at a sensor 26, which is also in electrical communication with the central processor 30. The light beam generators 25 are angled so that the light beams 28 produced are substantially vertical and run parallel to the theoretical divide between boundary segments - should the theoretical boundaries continue from the x axis into the y axis when they reach the boundary. The light beam generators 25 are located equidistant from the dividing line 23. In an embodiment, the light beams 28 are separated by 8mm at a location between the light beam generators 25 and the sensors 26. Alternatively, the apparatus could be adjusted so that the light beams 28 are closer together or further apart as necessary.

In use, as a ball 29 enters the zone of uncertainty 24, it will contact one or both of the light beams 28, depending on its flight path. As the ball 29 passes through a light beam 28 it prevents the light beam 28 from reaching the sensor 26 which is detected by the processor 30. It is also envisaged that the light beams might be broken in a sequence i.e. with a time delay between each recorded event. This would allow the determination that the majority of the ball might have been on one side of the line of separation in order to be able to confirm where the ball has exited the boundary with greater confidence than would otherwise be possible. If only one of the beams is broken, then the segment for which that beam corresponds will be attributed the boundary. However, if both beams are broken then it is the beam that is broken first which is considered to be the boundary segment through which the ball exited, and the boundary will be recorded as such. In an embodiment, the light beam generators 25 generate infra-red light, and correspondingly the light beams 28 are infra-red light beams. In the displayed embodiment, the light beam generators 25 are housed within the boundary rope 20 and the sensor 26 is located at a distal location such as the underside of the roof. In an alternative embodiment, the sensors

26 could be housed within the boundary rope 20 and the light beam generators 25 could be fixed at a distal location such as the underside of the roof. Equally, the battery sources

27 as displayed as being housed within the boundary rope 20, could be located externally of the boundary rope.

It should be understood that Figure 5 provides a representative system with reference to a single zone of uncertainty, however, this apparatus would be replicated at each zone of uncertainty between adjacent boundary segments around the entire boundary. Data from each zone of uncertainty may be transmitted to a central processor 30.

Figure 6 shows a boundary rope with a second embodiment of a secondary determination means incorporated. The boundary rope 20 as shown, is divided into adjacent segments 21 and 22 with dividing line 23. The zone of uncertainty 24 is highlighted as the region around the dividing line 23. In the displayed embodiment, the secondary means of determination is a camera housed within the boundary rope 20 at the dividing line 23. The camera 31 is in circuit with the battery 27 and the central processor 30. The camera 31 is angled so that its field of vision 35 is substantially vertical along the theoretical y axis of the divide between two boundary segments. When the system predicts that the flight path of the ball 29 will pass within the zone of uncertainty 24 the camera will capture data of the ball 29. The camera 31 displays an imaginary line 32 which represents the theoretical divide between the boundary segments. Using this line 32 as a reference point it is possible to determine the boundary segment through which the ball has exited the playing area. This determination mechanism would preferably be automated. In an automated embodiment, the side of the line with a greater area of ball 29 would be considered the boundary segment through which the ball 29 exited. In an embodiment, the information recorded by the camera 31 and processed by the processor 30, is stored on the data storage 33. The data storage 33 may be in circuit with the camera 31 or be a central data storage 33 in circuit with the processor 30. In a preferred embodiment, the data storage 33 records all information recorded by the cameras 31 located in all zones of uncertainty 24 around the boundary of the playing area.

Figure 7 shows a boundary rope with a third embodiment of a secondary determination means incorporated. In an embodiment, the boundary rope 20 is adapted to include a pressure gauge 34 capable of detecting when a ball 29 has struck the boundary rope 20. In a preferred embodiment, each segment of the boundary contains a separate pressure gauge 34, capable of detecting the impact of a ball. Each boundary segment, with individual pressure gauge 34 incorporated, is in electrical or wireless communication with the central processor 30. Under circumstances where 2 or more segments record impact, such as when the ball 29 strikes the boundary rope 20 within the zone of uncertainty 24, it will be the segment which records the contact first that will be recorded with the boundary. As with previous embodiments, the current embodiment can exist in circuit with or parallel to the data storage 33.

In an embodiment, the pressure gauge 34 is a piezo-resistive strain gauge, able to detect pressure on the boundary rope via a change in resistance of a material incorporated into the boundary rope. The material could be silicon, polysilicon thin film, bonded metal foil or any other suitable material. To maximise the sensitivity of the circuit and reduce errors the strain gauges are connected to form a wheatstone bridge circuit. In a preferred

embodiment, the change in resistance can be used as a quantitative measure of force, giving an indication of the force the ball had when it struck the rope. By using this value in combination with known quantities such as the weight of the ball and its shape, which are stored in a data store 33 in attachment with the processor 33, it is possible to determine the speed of the ball when it struck the boundary rope. This system could be used in conjunction with the previously described ball tracking system as a secondary means of determination or individually.

In an embodiment, the secondary determination means, are dormant until activated by the primary determination system which has predicted that the ball is likely to pass through the zone of uncertainty 24. In a further embodiment, the sensor responsible for each individual zone of uncertainty could be activated independently of each other as required. For example, when the ball tracking system first produces a predicted ball flight path 7, if the flight path suggests that the ball may pass within the zone of uncertainty 24, then the system would activate the relevant sensor. The threshold distance used to determine when a sensor would be activated would be larger than that of the zone of uncertainty 24, due to the reduced accuracy at that early stage of flight path prediction.

The system as currently disclosed could also generate a vast amount of statistical data of not only the locations that the batsman hits certain shots but also the type and quality of shots played through each segment. For example, from the flight path the system could provide information on the average trajectory the ball takes through each segment, indicating a batsman's propensity to hit shots through the air in certain regions. In addition, knowledge of the force a batsman generates through different areas by measuring the speed a ball is travelling through each segment or the force with which it strikes the boundary, gives a greater indication of whether a fielder in a nearby location would have been able to cut off the boundary. The accurate data obtained may also lend itself to the validation of a result which could be provided to an online or electronic betting platform to indicate with accuracy where a particular ball has impacted or crossed a boundary. Further steps may be envisaged such as the payment to a punter of a predetermined sum of money should the impact of the ball correspond to a correctly predicted section of the boundary. In this context, a user of the betting platform might interact with a user interface to select a section for the next predicted boundary section to be hit. The inputted data may then be compared in a subsequent step with the actual accurately determined position of the ball for the next awarded boundary leading to an accurate determination of whether the bet was successful or not. The enhanced validation process defined in the claims may therefore have multiple further secondary steps beyond those illustrated in this section of the application. It may for example have a step of recording the predicted impact of a ball or next position of a ball relative to a particular boundary section, determining the position of impact or of crossing the boundary and then comparing the predicted position of the ball with the actual position in order to determine whether or not the bet was successful.

In a preferred embodiment, the system can record the locations of all fielders within the playing field. This could be input manually or by object analysis in much the same way as the pixel analysis used to identify the location of the ball. This information could enable the shots to be further subcategorised based on presence or absence of a fielder within a certain location. For example, a fielder in the covers might necessitate a player to loft the ball through that segment at a higher frequency to when the fielder is not present at that location. In a second example, a player may be shown to hit the ball softer through the offside when there are slips in place, to reduce the chance of a nick carrying to the slips.

In a preferred embodiment, the system includes a display in communication with the processor 30 and the data storage 33, which enables the user to view all statistics produced by the system. In addition, the display enables the user to manually set features such as statistics to record and optionally change the size and location of the boundary segments. In a second independent aspect, an embodiment defines a system for use during a sporting game such as cricket, to determine the location a ball exits from a pre-determined area, or playing field, within pre-set boundary segments. Figure 8 displays an embodiment of a boundary rope 40, with a system for determining the location a ball exits a playing field incorporated. In the displayed embodiment, the boundary rope 40 is segmented into theoretical boundary segments 41 and 42

respectively. The number, size and location of the segments can be determined before the initiation of the system. Dealing first with an individual segment; at the peripheries of the segment 41 are cameras 43 and 44, angled so that their field of vision 51 and 52 respectively overlap. The cameras 43 and 44 are angled so that there is almost total coverage of the boundary segment 41. This system is optionally provided for each boundary segment. In use, when a ball 49 is hit through a boundary segment, the cameras are angled so that both cameras for an individual segment can only view a ball if it is hit through that segment.

In the example displayed in Figure 8, the ball 49 is within the field of view of cameras 43, 45 and 46. As the system will only record a boundary within a segment if both cameras for a segment record a ball within a pre-set time threshold of each other, segment 41 will not record a boundary and segment 42 will. Each camera is in electrical communication with a central processor 56 and a data storage 57. Optionally, the cameras could utilise pixel analysis to determine when a ball has been registered within their field of vision.

Alternatively, the ball location could be input manually by an operative viewing the images of each camera.

Figure 9 displays the embodiment of a boundary rope 40 as shown in Figure 9, in the event that the ball 49 passes within the zone of uncertainty. In the current embodiment, the zone of uncertainty is defined as the area between the field of vision 52 and 52 of cameras 44 and 45 respectively. A ball passing within this area would not be recorded by both cameras of any segment and as such would require a separate means of determination. A separate camera 50 is housed along the dividing line 47 within the zone of uncertainty 48. If no two cameras for a segment detect a ball within a threshold time, the ball is considered to have passed through a zone of uncertainty 48. At this point the processor 56 initiates a separate protocol. The camera 50 is angled and configured as described in Figure 7. The reference line 59 incorporated within the images taken by camera 50, provides a reference point for determining which boundary segment the ball passed through.

In an embodiment, the cameras have limited 3D field of vision to reduce the number of redundant ball recognition from cameras at separate locations around the boundary.

Claims

1. A system for determining the location of impact of a ball on the boundary of a playing field or for determining the location where the ball crosses said boundary, the system comprising a primary detection mechanism for detecting the position and/or movement of a ball within said playing field; and a secondary detection mechanism which is configured to validate the impact of the ball on said boundary and/or validate where the ball crosses said boundary.

2. A system according to claim 1, wherein said primary detection mechanism comprises at least two cameras to detect the movements of a ball.

3. A system according to any of the preceding claims, wherein said secondary detection mechanism is integral with the boundary.

4. A system according to any of the preceding claims, wherein said secondary detection mechanism is located adjacent to said boundary.

5. A system according to any of the preceding claims, wherein said boundary is

substantially triangular in cross-section and said secondary detection mechanism is located at least partially within the boundary.

6. A system according to any of the preceding claims, wherein the boundary is segmented in a plurality of discrete sections; said secondary mechanism is configured to validate which section said ball primarily impacts or crosses.

7. A system according to any of the preceding claims, further comprising a processor in communication with a plurality of cameras, capable of receiving signals from said cameras and predicting the flight path the ball will take; a data store; means for comparing the predicted flight path to a model of a playing field with pre-defined segments; and means for activating said secondary detection mechanism when a ball is predicted to pass within a zone of uncertainty defined between two discrete sections of a boundary.

8. A system according to any of the preceding claims, wherein said secondary detection mechanism comprises a sensor suitable for determining the exit location of a ball.

9. A system according to any of the preceding claims, wherein said secondary detection mechanism comprises a sensor which is housed within said boundary.

10. A system according to any of the preceding claims, wherein said secondary detection mechanism comprises a sensor which is housed within a zone of uncertainty defined between two discrete sections of a boundary.

11. A system according to any of the preceding claims, wherein said secondary detection mechanism comprises one or more cameras.

12. A system according to claim 11, wherein said one or more cameras are angled upwards along lines of division between discrete sections of the boundary.

13. A system according to any of the preceding claims, wherein said secondary detection mechanism comprises one or more light emitters.

14. A system according to any of the preceding claims, wherein said secondary detection mechanism comprises a pressure sensor suitable for detecting contact between the ball and the boundary.

15. A system according to any of the preceding claims, wherein the secondary detection mechanism is configured to be dormant until activated by said primary detection mechanism.

16. A playing field boundary comprising a secondary detection mechanism which is configured to validate the impact of the ball on said boundary and/or validate where the ball crosses said boundary.

17. A playing field boundary according to claim 16, wherein said boundary is substantially triangular in cross-section and said secondary detection mechanism is located at least partially within the boundary.

18. A playing field boundary according to any one of claims 16 to 17, which is segmented in a plurality of discrete sections; said secondary mechanism being configured to validate which section said ball primarily impacts or crosses.

19. A playing field boundary according to any one of claims 16 to 18, wherein said secondary detection mechanism comprises a sensor suitable for determining the exit location of a ball.

20. A playing field boundary according to any one of claims 16 to 19, wherein said secondary detection mechanism comprises a sensor which is housed within said boundary.

21. A playing field boundary according to any one of claims 16 to 20, wherein said secondary detection mechanism comprises a sensor which is housed within a zone of uncertainty defined between two discrete sections of a boundary.

22. A playing field boundary according to any one of claims 16 to 21, wherein said secondary detection mechanism comprises one or more cameras.

23. A playing field boundary according to any one of claims 16 to 22, wherein said one or more cameras are angled upwards along lines of division between discrete sections of the boundary.

24. A playing field boundary according to any one of claims 16 to 23, wherein said secondary detection mechanism comprises one or more light emitters.

25. A playing field boundary according to any one of claims 16 to 24, wherein said secondary detection mechanism comprises a pressure sensor suitable for detecting contact between the ball and the boundary.

26. A playing field boundary according to any one of claims 16 to 25, wherein the secondary detection mechanism is configured to be dormant until activated by a primary detection mechanism.

27. A method for determining the location of impact of a ball on the boundary of a playing field or for determining the location where the ball crosses said boundary, the method comprises the steps of detecting the position and/or movement of a ball within said playing field; and validating the impact of the ball on said boundary and/or validating where the ball crosses said boundary.

28. A method according to claim 27, wherein the boundary is segmented in a plurality of discrete sections; and the method comprises the further step of validating which section said ball primarily impacts or crosses.

29. A method according to either claim 27 or claim 28, comprising the further steps of providing a processor in communication with a plurality of cameras, receiving signals from said cameras, predicting the flight path the ball will take; comparing the predicted flight path to a model of a playing field with pre-defined segments; and activating a validation step when a ball is predicted to pass within a zone of uncertainty defined between two discrete sections of a boundary.