AU2004216667B2 - Transformation Of Images For Image Capture And Viewing - Google Patents

Description

-. ,n rIn 64 4 472 3358 P.05/3:3 Regulation 3.2 AUSTRALIA PATENTS ACT, 1990 COMPLETE SPECIFICATION FOR A STANDARD PATENT ORIGINAL Name of Appcant: BRUCE PETER PARKER Actual Inventor: BRUCE PETER PARKER Address for service A J PARK, Level 11, 60 Marcus Clarkt Street, Canberra ACT in Australia: 2601, Australia Invention ride: Trisfonaton OfImages For Image Capture And Viewing The flowing tment is a full description of this inwutio.including fte best method of performing it Imown to me. 2577J.DOC 1nA niaciitano 6'+ 4 4Td JJ P. 06/33 2 TRANSFORMATION OF IMAGES FOR IMAGE CAPTURE AND VIEWING FIELD OF THE INVENTION 5 The present invention relates to signs, uch as advertising signs, that are applied to the surfaces where televised events are held, such as sports grounds. BACKGROUND TO THE INVENTION 10 Signs are commonly placed in and arowd televised events, such as sports events. Generally signs display promotional content of sponsors and customers who pay to have their advertisements at the ground. This provides them with the benefit of exposure to a wide viewer audience during televising of the event. 15 Advertising signs II are often painted directly on to the event surface 10, such as a sports field, as shown in Figure Ia. This provides tle advertiser with significant exposure, due to the surface featuring predominantly in any 'evlsing. The drawback of surface advertising signs is that often the signs 11 appear out of proportion to the audience when watching the event on television 13, as shown in FIgure lb. This is due to the geometric relationship 20 between the camera 12 and the event surface 10. This dIstortion is partially overcome by the inverse perspective transform technique described in AU 656609. The technique involves mathematically transforming an image of the sign into a deliberately pre-distorted m based on, among other things, the 25 relationship between camera t a predetnmlned location and the surface at tho mid point where the pro-distorted Image is to be applied The deliberately pre-distorted image is then applied to the surface When a camera at the predetermined position captures an image of the pre-distorted sign on the surface, it asin "proporion" to the viewer, and can eve create the illusion of the sign "sanding up" of the playing surface. 30 3 However, the technique disclosed in AU 656609 produces the pre-distorted surface sign by mathematically transforming the desired image from a virtual "viewing" plane, at a point in front of the camera, to the surface. Not all parts of the virtual plane are equi-distant from the camera lens, and therefore the viewer can still perceive distortions in the televised 5 image. In this specification where reference has been made to patent specifications, other external documents, or other sources of information, this is generally for the purpose of providing a context for discussing the features of the invention. Unless specifically stated otherwise, 10 reference to such external documents is not to be construed as an admission that such documents, or such sources of information, in any jurisdiction, are prior art, or form part of the common general knowledge in the art. SUMMARY OF THE INVENTION 15 It is an object of the present invention to provide an improved transformation method and/or computer system and/or computer program for creating pre-distorted images for application to event surfaces of televised events, or to at least provide the public with a useful choice. 20 In one aspect, the present invention broadly consists in a method of transforming an image into a pre-distorted image for application to an event surface comprising the steps of: transforming a plurality of fragments of the image, as they would appear on a virtual spherical surface, into transformed fragments of the image as they should appear on the 25 event surface, based on the geometric relationship between a camera, the virtual spherical surface and the event surface; and constructing the pre-distorted image from the transformed fragments. The transformed fragments are used to construct he pre-distorted image, in a suitable 30 manner. The pre-distorted image, when applied to the surface, televised and viewed, is perceived as the original image lying on the virtual spherical surface. The pre-distorted 4 image appears substantially in the correct proportions. That is, it appears on screen much like the original image. Typically, the image will be of an advertising nature, although the image can be any 5 graphical work for application to an event surface and captured by a camera for televising or dissemination for viewing by any other means. The event surface can be any surface on which a televised event takes place, such as a sports field, indoor/outdoor court, track, stage. snooker table, table tennis table, swimming pool, ice rink or the like. The event surface could also be an advertising hoarding. 10 The virtual spherical surface is the surface upon which the image applied to the event surface will be perceived as appearing by viewers. Preferably, the virtual spherical surface encompasses the camera, with the camera at the centre. The radius of the spherical surface is determined from the horizontal and vertical distances between the camera lens and the 15 base of where the pro-distorted image will be applied to the event surface. The virtual spherical surface intersects the event surface at the base of the pre-distorted image. The transform is preferably carried out by dividing the original image into rows of horizontal fragments, and transforming the fragments row by row. Preferably, the length or 20 height of the fragment row is transformed by determining the projection of the length or height onto the event surface. More preferably, the projection is calculated mathematically, using: 1 1 ASI=..() 1000 H H2+D 25 where H is the height of the camera above the event surface, D is the horizontal distance of the camera from the base of where the pre-distorted image is to be applied to the event surface, and ASI is the length of the adjacent side of a fragment in the row. Preferably, the width RW of the top of the pre-distorted image is transformed by using: 5 RW =+KW- v4(A+L) ...(2) A 2 where A is the distance from the camera to the base of where the pre-distorted image will be applied to the event surface. W is the width of the base of the original image, L is the 5 length of the pre-distorted image, and FOV is the diameter of the field of view of the camera at a distance of 1 metre, when the pre-distorted image is included in a general camera shot of the event surface. The FOV can be measured or alternatively calculated using H W FOV= (3) 10 where 1-1 is less than or equal to D. Where the height H of the camera is greater than the horizontal distance D of the camera from the base of the pre-distorted image, then 15 W D FOV =(4) 2 HI Preferably, the original image is obtained and then converted into an electronic representation. The electronic representation of the original image is transformed into an 20 electronic representation of the pre-distorted image for application to an event surface using a computer system or similar. A printout, stencil, template or other representation of the pre-distorted image can be generated by the computer system to provide a means by which the pre-distorted image can be applied to the event surface. The, stencil template or the like can be supplied to a third party who carries out the application of the image to the event 25 surface. In another aspect, the present invention broadly consists in a computer system for transforming an image into a pre-distorted image for application to an event surface comprising a processor arranged to: obtain the image; transform a plurality of fragments of 6 the image, as they would appear on a virtual spherical surface, into transformed fragments of the image as they should appear on the event surface, based on the geometric relationship between a camera, the virtual spherical surface and the event surface; construct the pre-distorted image from the transfonned fragments; and output a representation of the 5 pre-distorted image. In another aspect, the present invention broadly consists in a computer program for transforming an image into a pre-distorted image for application to an event surface, the computer program performing the steps of: transforming a plurality of fragments of the 10 image, as they would appear on a virtual spherical surface, into transformed fragments of the image as they should appear on the event surface, based on the geometric relationship between a camera, the virtual spherical surface and the event surface; and constructing the pre-distorted image from the transformed fragments. 15 In another aspect, the present invention broadly consists in a method of depicting an image comprising the steps of; obtaining an image to be televised or otherwise disseminated for viewing; transforming a plurality of fragments of the image, as they would appear (3n a virtual spherical surface, into transformed fragments of the image as they should appear on the event surface, based on the geometric relationship between a camera, the virtual 20 spherical surface and the event surface; constructing the pre-distorted image from the transformed fragments; applying the pre-distorted image to the event surface; capturing an image of the event surface with the applied pre-distorted image using the camera; and televising or otherwise disseminating the captured image for viewing. 25 The term "comprising" as used in this specification and claims means "consisting at least in part of". When interpreting each statement in this specification and claims that includes the term "comprising", features other than that or those prefaced by the term may also be present. Related terms such as "comprise" and "comprises" are to be interpreted in the same manner. 30 04-DCT-2664 15:45 A 5 PARK 64 4 472 3358 P.11/33 7 BRIEF DESCRJPflON OF THE Preferred embodiments of the inventi 11 be described with reference to the accompanying drawings, of which: Figures la and lb show how an irage i to a sports field become diastorted when televised, Figures2a and 2b show in schematic form. n mple of an image prior to transformation, and a pr-distorted image after tn a using a method according to a preferred 10 embodiment of the inventiom Figure 3 shows in elevation, the relations a camera, virtual spherical surface and event surface on which a prsaistorted is to be applied, Figure 4 shows in plan, the relationship the camera, event surface and position of a pno-istorted image, and 15 Figures Sa and Sb show the original and dirted images divided into corrcqpmding fragments. DETARrLn DESCRIPTION OF THE P FEMBODDVIMENS 20 According to a preferred embodiment of the ivetion, Figures 2a and 2b respectively show a schematic example of an original image 20 sponsor might want to have trnsformed into a deliberately pr-distorted or St 22, and applied to a surftaceof a televised event. As can be seen, the pre-di (tranformed) image 22 has a bigger width at the top, longer sides and curved t base edges. For example, the original 25 image 20 could be an advertising sign or si Ia. For illustrative purposes, the original image 20 is only shown as a frame and a siep c pi, however It should be appreciated that the image may be a graphical work dcpic motoal material of the sponsor. The pre-distortedimage 22 can be applied to the forexample a sports ficld.suchthat when pichres of the field are televised or disseminated for viewing, the pro 30 distorted image 22 will appear in concct ppins to the viewer. That is, the televised £lM n un casi ivnnflca - -- - - .JT,)c Pr 12/M image will appear similar to the original image 20 shown in Figure 2. In effect the relationship of the camera with the field "undistortd" the applied pre-distorted image 22, such that it appears in proportion when viewed. When viewed in correct proportions, the original image 20 may also have the llusiou of "standing up" from the field. 5 The original image 20 shown in Figure 2a has a width W along the base 21 and a height Hi along its edge 23. The original image 20 could be a graphical work of any kind, for example an advertising sign, team logo, event logo, general sign or the like. The camera will typically be a motion capture device, such as a television camera, although the 10 technique will also work for still captures. The event surface can be any surhce on which a televised event takes place, such as a sports field, indoor/outdoor coWr, track, stage, snooker table, table tennis table, swinging pool. ice rink or the like. The event surface could also be an advertising boarding at an event. This list Is not exhaustive of the possible applications. 15 In a preferred embodiment, an original image 20 will be obtained, transformed into a pre distorted image 22 and a stencil, template or printout of the pr-distorted image 22 to be applied to the subce will be provided. Prferably. the original image 20Is obtained and then converted Into an electronic representation. The electronic representation of the 20 original image 20 is transformed into an elet representation of the pro-distorted image 22 for application to an event surface using a computer system or similar. The transform can be implemented in a computer program opting on the computer system. A printout, stencil, template or other representation of the pre-distorted image can be generated by the computer system to provide a means by whi4h the pre-distorted image can be applied to the 25 event surface. Grid lines can be overlaid to assist in applying the pre-distorted image. The stencil, template or the like can be supplied to a thitd party who carries out the application of the pre-diutorted image to the event surfc. The event surface can then be televised. Figure 3 shows, in schematic form, an ovation view of a nominal arrangement of a 30 camera 30 In relation to the position on the vnt surface 31 where the pre-distorted image -"MI15-46 "A0 H U' 64 4 472 3358 P. 13/33 9 22 is to be applied. The camera 30 is ned at a heightH above the event surface 31, at a horizontal distance D from the mid t M of where the base 21 of the original image 20is to be positioned. The transform is out bused on knowledge of where the camera30Ois in relation to wherethe image 22 willbe applied torthesurface 5 31. Referring to Figures 3 and 4,a pitt embodiment of the transform method will first be described with reference to the base sides of the original image 20, to clearly demonstrate the general concept. The image 20 is considered to lie on a virtual sphere 33 that encompasses the camera 3 where the camera 30 is at the centre of the sphere 33. Only a portion of the sphere 3 is shown in Figure 3, corresponding to the 10 portion that is in the camera view, Obviou y, only the side of the sphere 33 is indicated, however it will be appreciated that the defines a virtual spherical surface around the camera 30. The radius R of th sphem 33, 1 by any one of projection lines 34a-34f between the camera 30 and the sphere 33, is calculated using Pythagoras' theorem, namely R21H 2

+D

2 . The virtual sphere 33 the event surface 31 at the base of the pre 15 distorted image 22. To generate the pre-distorted image 22, original image 20 is transformed from its staard flat form asshowninlFigure 2ai aforrmas it wouldappearifit weretobe applied to the inner surface of the virtual 33. This is, in effect, how the image that 20 will actually appear to viewers watching the evised version of the predistorted image 22 applied to the ground. Preferably, this form is carried out using a mathematical technique, known to those skilled in thin logy. Alteruatively, it may be assumed that the original image 20, is already lying on inner surface of the virtual sphere 33, in which case this transformation ned not take lace. The relationship between the camera 30 25 and the event surface 31 effectively trnf the pro-distorted image 22 painted en the field into an undistorted image appearing on virtual spherical surface 33. It should be noted that the image is not actually applied to spherical surface. As shown in Figure 3, the height or length of the pre-distorted image 22 is generted 30 fon a projection of the height of the ori image 20, lying on the virtual spherical eV%&SV I n L..CM .. nr'W-- . . .

"" C'- 1"*4" "C Rn J Ir 64 4 472 3358 P.14/33 surface 33, on to the event surface 31. C section lines 34a-34f enantng from the camera 30, through the original image 20 surface of the virtual sphere 33, onto the event surface 31. indicate the projection. Ta ewerpceives the inverse of this projected pe-distorted image 22 applied to the ev s namely the original image 20 on the 5 virtual spherical surface 33. In a prffat, the projection is determined mathematical by projecting a plurality of dualportions of original image 20 onto the event surface in turn To do so, the 020 Is processed in incremental portions of a suitable resolution along its height. F c portion, the height of the original image 20 increment is convted, by pwjection equivalent height increment, taned AS! 10 (adjacent side irse), that that porton of the original image 20 should take upon application the event surface 30. where: 1o00( H J 15 where H is the height of the camera 30 aboy event surfte 31, andD is the distance of the camera from the base of the pre-distorted in 22 once applied to the event surface 31. The multiplier 1000 corresponds to the e of incruemets, or resolution, used in processing. This can be readily increased or to the required resolution. 20 For example, refering to Figure 3, the ori - a.20 on the spherical surface is broken into 5 iscrmental portions for illustrative . The ASI or height of each portion when projected from the virtual spherical su e33 to the event surface 31 is calculated using (1), resulting in the converted heights ga) -ASI(e) as shown. The total converted height FH is also shown, which correlates to L in Figure 2b. It will be appreciated 25 that a much greater number of incrments typically be used. This projection can take place on both edges of the originaL 20, to determine the height of the pro distortedimage 22 to be applied to thesurfac 31 -6A- 1 -"" InZt4 4 4Y ?3358 P. 15/33 11 Mcx, the width dimensions of the pre-ik image 22 ae detained. Figure 4 shows, in schematic fon, a plan view of the 30 in relation the position on the event surface 31 where, tho pre-distred image 22 is to applied. The positioning of the pre-distorted 5 image 22 is mlso shown, although it should noted this is for illustrative pupose, and at this stage has not been fully determined. A the distance between the camera and the base of the pre-distorted image 22 on the event ace 31. Lines 41a, 41b indicate the field of view of the camera 30. W is the width of base of the pre-distorted image 22, which is equal to the width ofthe base 21 of the oril image 20. RW is thewidth of the top of the 10 pre-distorted image 22 applied to the event 31, where: RW= W FO))IA+) ... (2) and where FOV = diameter of the field of w of the camera from Im distance, when the 15 pr-distorted image 22 is included in a ge amera shot of the event surface 31. The FOV can be measured or alternatively cal using: FOVr= ...(3) D 2 20 where His less thanor equal to D. Where the height H of the camera 30 is than the horizontal distance D of the camera 30 hum the base of the pre-distorted image then 25 FOV=E-. ...(4) 2 H That is. the field of view FOV is equal to the ba~e width W of the pro-distorted image 22 divided by 2, and this, in turn, is all divided by th4 result of the camera height H divided by 12 the horizontal distance D of the camera 30 from the base of the pre-distorted Image 22. Both W and RW are curved with the radius of the sphere which is A +Im. The ratio between the width W and height Iji of the original image 20 may be determined. 5 During traWsformstiOn, this ratio can be maintained in the pr-istorted image 22 to ensure that the observer is presented with a true and ecurate image in the correct proportions. This can be achieved by an adjustment fler the initial tnsformation is carried out. Rasiing calculations can be implemented based on one of three variables, base width W, height of the image that is apparent to the viewer Hi, and actual length L of the pre 10 distorted image applied to the surface 31. For example, an advertiser may have a width allowance of X metres. The width of the preAistorted image 22 can therefore be esAied, in correct proportion to the apparent height Hi, according to the ratio. Alternatively, the height of the preo-distorted image 22, can be resized to retain the ratio. In many circumstaneme, an adverdse may have an ullowasce of X metres in length on the surface to 15 apply the pro-distorted image. From this, the apparent height Hi is calculated and then the base width is calculated, using the ratio. At this point the calculated AS!s and the lateral dimensions could be used to produce a visual printout or display of how the border of the pre-distorted image 22 should appear on the event surface 31, along with dimension, 20 The above description describes in a eonceptzal sense, the nature of the transform process, and provides an indication of the shape of th perimeter pre-distorted image 22. In practice, the entire original image 20 rcquirsc trpnsformation to produce the pre-distorted image 22. To do so, the original image 20 is nominaly divided into horizontal fragments 50, such as shown in Figure Sa, and these am transformed row by row. Preferably, each 25 fragment is of an equal height and width, determined by the size of the original image 20 and the dimensions of the pre-distorted image 22. In Figure Sa, only 25 fragments are shown for clarity, although in practice mady more will be utilised, depending on the required resolution of the final pre-distorted image 22. The pre-distorted image 22 can be constructed from fragments 51 bhnsformed from fagments 50 of the original image 20. 30 The ransfbrm is done row by row, 55-59 using equations (1) and (2). The height of each CflASf) tMaS fl2OA b0.....k.lA n Ab .r . . - - - - 13 fragment of the original image 20 is detennjned by the division ratio used in equation (1), (1000 - shown in equation (1) although t$is can be altered). In effect each fragment height relates to the Increment previously described with reference to equation (1). Each horizontal row of frngments, 55-59, in the original image 20 has its height transformed into 5 a conaponding ASI height in the pro-distufted image 22, using equation (1). The rear width, RW1-RW5, of each row 55-59 in the pre-distorted image 22 is determined from the width W of the row in the original image 20, using equation (2). It will be appreciated that RW5 is the sane as RW in Figure 2b. Fror* this, the corresponding transformed row, 55 59, can be constructed from the eight and rpar width, with the individual fragments being 10 divided evenly into the overall width. The appropriate lengthrwidth ratio adjustment is then made. Oce all rows 55-59 have been transfrmed and constructed, the transfonned image 22 can be constructed from these rows. Preferably, the process is canied out using software on a computer system or alternatively a 15 computer system with a processor which is ranged to carry out the process may be utilised. Th original image 20 can be imported and displayed by the software in Bitmap or JPEG formats. The image 20 is then treated by the software as a bitmap. This means that each pixel on the screen is broken down to tle Twip level. Thee are 100 Twips to a Picel. If the original image is 5 meters high, then it could be broken into 5,000 discrete aqual 20 horizontal components (based on the 1,000 iberenents used in the calculations). The ASI value for each of these 5,000 values may be recorded and correlates with the original image 20. Each of the ASI values increases with the distance from the base of the image. Every horizontal component from the original image 20 has a corresponding horizontal component in the pre-distorted image 22. The height is fixed in the original image 20, but 25 varies (ASI) in the pro-distorted image 22. By re-assignig the Twips fim the original image 20 to that required by the pro-distorted image 22, the "stretch" is achieved. For example, in the original image 20 there may be 100 Twips per pixel. The corresponding horizontal "line" in the transformed image ipay be 10 times the height, so 10 Twips are assigned per Pixel. In actual fact, there reans 100 Twips per pixel, but the other 90 are 30 averaged values of the 10 used. For the wdtb, each 'line is calculated" based on the 14 relevant width fonuula. In this case the Twips ae spread evenly across the width. As there are so many calculations and lines, the pre-4istozted image 22 appears to have been strtched. But it has been stretched in small incrments, in the correct amounts, at the right location, rather than just stretching the image evenly over the whole transformed image 5 area. It will be appreciated that the software and/or computer system can be arranged to receive, either automatically or manually, all the inpit dta required to transform the original inage into the prdstorted image. For example, thp vertical height of the camera above the event 10 surface and the horizontal distance of the capner from the base of where the pre-distorted image will be applied to the event suace can be entered into the software and/or computer system manually. The fbregoing dcouription of the invention ivdlu4es prefhnWd focus thereof Modifications 15 may be made thereto without departing froqa thb scope of the invention as defined in the accompanying claims.

Claims (36)

1. A method of transforming an image into a pre-distorted image for application to an event surface comprising the steps of: 5 transforming a plurality of fragments of the image, as they would appear on a virtual spherical surface, into transformed fragments of the image as they should appear on the event surface, based on the geometric relationship between a camera, the virtual spherical surface and the event surface; and constructing the pre-distorted image from the transformed fragments. 10

2. A method of depicting an image comprising the steps of: obtaining an image to be televised or otherwise disseminated for viewing; transforming a plurality of fragments of the image, as they would appear on a virtual spherical surface, into transformed fragments of the image as they should appear on 15 the event surface, based on the geometric relationship between a camera, the virtual spherical surface and the event surface; constructing the pre-distorted image from the transfonned fragments; applying the pre-distorted image to the event surface; capturing an image of the event surface with the applied pre-distorted image using 20 the camera; and televising or otherwise disseminating the captured image for viewing.

3. A method according to claim 1, wherein the pr-distorted image, when applied to the event surface, captured by the camera and viewed, is perceived as the image lying on 25 the virtual spherical surface.

4. A method according to claim 2, wherein the pre-distorted image is perceived as the image lying on the virtual spherical surface when the captured image is viewed. 16

5. A method according to any one of the preceding claims, wherein the step of transforming the fragments of the image comprises dividing the image into rows of fragments and transforming the fragments row by row. 5

6. A method according to any one of the preceding claims, wherein the step of transforming the fragments comprises firstly transforming the length of the fragments and then transforming the width of the fragments along the transformed length.

7. A method according to any one of the preceding claims, wherein the step of 10 transforming the fragments of the image comprises transforming the length of the fragments based on the vertical height of the camera above the event surface and the horizontal distance of the camera from the base of where the pre-distorted image will be applied to the event surface. 15

8. A method according to any one of the preceding claims, wherein the step of transforming the fragments of the image comprises transforming the width of the fragments based on the distance of the camera from the base of where the pre-distorted image will be applied to the event surface, the width of the base of the image, the length of the pre-distorted image, and the field of view of the camera when the pre-distorted 20 image is included in a general camera shot of the event surface.

9. A method according to claim 8, comprising the step of calculating the field of view of the camera based on the vertical height of the camera above the event surface, the horizontal distance of the camera from the base of where the pre-distorted image will be 25 applied to the event surface, and the width of the base of the image.

10. A method according to claim 8, comprising the step of measuring the field of view of the camera. 17

11. A method according to any one of the preceding claims, wherein the fragments of the image are the same size.

12. A method according to any one of the preceding claims, wherein the resolution of the 5 pre-distorted image is dependent on the number of fragments of the image.

13. A method according to any one of the preceding claims, wherein the virtual spherical surface encompasses the camera, the lens of the camera beirg at the center of the virtual spherical surface. 10

14. A method according to claim 13. wherein the radius of the virtual spherical surface is based on the horizontal and vertical distances between the lens of the camera and the base of where the pre-distorted image will be applied to the event surface.

15 15. A method according to any one of the preceding claims, wherein the virtual spherical surface intersects the event surface at the base of where the pre-distorted image will be applied to the event surface.

16. A method according to any one of the preceding claims, wherein the image is a 20 graphical work containing advertising or promotional material.

17. A method according to any one of [he preceding claims, wherein the event surface is a sports field, indoor/outdoor court, track, stage, snooker table, table tennis table, swimming pool, ice rink or advertising hoarding. 25

18. A method according to any one of the preceding claims, wherein the image is electronic and the fragments of the image are formed from one or more pixels of the electronic image. 18

19. A computer system for transforming an image into a pre-distorted image for application to an event surface comprising a processor arranged to: obtain the image; transform a plurality of fragments of the image, as they would appear on a virtual 5 spherical surface, into transformed fragments of the image as they should appear on the event surface, based on the geometric relationship between a camera, the virtual spherical surface and the event surface; construct the pre-distorted image from the transformed fragments; and output a representation of the pre-distorted image. 10

20. A computer system according to claim 19, wherein the pre-distorted image, when applied to the event surface, captured by tie camera arid viewed, is perceived as the image lying on the virtual spherical surface. 15

21, A computer system according to claim 19 or claim 20, wherein the processor is arranged to divide the image into rows of fragments and transform the fragments row by row.

22 A computer system according to any one of claims 19-21, wherein the processor is 20 arranged to firstly transform the length of the fragments and then transform the width of the fragments along the transformed length.

23. A computer system according to any one of claims 19-22, wherein the processor is arranged to: transform the length of the fragments based on the vertical height of the 25 camera above the event surface and the horizontal distance of the camera from the base of where the pre-distorted image will be applied to the event surface; and transform the width of the fragments based on the distance of the camera from the base of where the pre-distorted image will be applied to the event surface, the width of the base of the image, the length of the pre-distorted image, and the field of view of the camera when 30 the pro-distorted irnage is included in a general camera shot of the event surface. 19

24. A computer system according to claim 23, wherein the processor is arranged to calculate the field of view of the camera based on the vertical height of the camera 5 above the event surface, the horizontal distance of he camera from the base of where the pre-distorted image will be applied to the event surface, and the width of the base of the image.

25. A computer system according to any one of claims 19-24, wherein the processor is 10 arranged to divide the image into fragments of the same size.

26. A computer system according to any one of the claims 19-25, wherein the representation of the pre-distorted image is an electronic image, printout, template or stencil. 15

27. A computer program for transforming an image into a pre-distorted image for application to an event surface, the computer program performing the steps of: transforming a plurality of fragments of the image, as they would appear on a virtual spherical surface, into transformed fragments of the image as they should appear on 20 the event surface, based on the geometric relationship between a camera, the virtual spherical surface and the event surface; and constructing the pre-distorted image from the transformed fragments.

28. A computer program according to claim 27, wherein the pre-distorted image, when 25 applied to the event surface, captured by the camera and viewed, is perceived as the image lying on the virtual spherical surface.

29. A computer program according to claim 27 or claim 28, wherein the step of transforming the fragments of the image comprises dividing the image into rows of 30 fragments and transforming the fragments row by row. 20

30. A computer program according to any one of claims 27-29, wherein the step of transforming the fragments comprises firstly transforning the length of the fragments and then transforming the width of the fragments along the transformed length. -5

31 A computer program according to any one of claims 27-30, wherein the step of transforming the fragments of the inage comprises: transforming the length of the fragments based on the vertical height of the camera above the event surface and the horizontal distance of the camera from the base of where the pro-distorted image will be 10 applied to the event surface; and transforming the width of the fragments based on the distance of the camera from the base of where the pre-distorted image will be applied to the event surface, the width of the base of the image, the length of the pre-distorted image, and the field of view of the camera when the pre-distorted image is included in a general camera shot of the event surface. 15

32. A computer program according to claim 31, further performing the step of calculating the field of view of the camera based on the vertical height of the camera above the event surface. the horizontal distance of the camera from the base of where the pre distorted image will he applied to the event surface, and the width of the base of the 20 image.

33. A method of transforming an image into a pre-distorted image for application to an event surface substantially as herein described with reference to the accompanying Figures 2a-5b. 25

34. A computer system for transforming an image into a pre-distorted image for application to an event surface substantially as herein described with reference to the accompanying Figures 2a-5b. 21

35. A computer program for transforming an image into a pre-distorted image for application to an event surface substantially as herein described with reference to the accompanying Figures 2a-5b. 5

36. A method of depicting an image substantially as herein described with reference to the accompanying Figures 2a-5b.