GB2469863A

GB2469863A - Measuring surface profile of golf clubs, calibrating image capture device and apparatus for preparing a measurement specimen by taking a cast of a surface

Info

Publication number: GB2469863A
Application number: GB0907517A
Authority: GB
Inventors: Steven Otto; Peter Clements
Original assignee: R & A Rules Ltd
Current assignee: R & A Rules Ltd
Priority date: 2009-04-30
Filing date: 2009-04-30
Publication date: 2010-11-03
Also published as: GB0907517D0

Abstract

There is described a method and apparatus for testing a surface of a golf club to determine conformance or otherwise with at least one of the requirements of the Rules of Golf. The apparatus comprises a moulding dam (7) for producing a moulded sample / cast (20) having a surface corresponding to the surface to be measured, and a cutting apparatus (23) for cutting a specimen or "slice" from the moulded sample / cast. An image of a profile of the specimen / slice is captured and dimensions of features of the surface profile of the golf club determined from the captured image. The specification also discloses apparatus and a method for calibrating an image capture device in order to produce a dimensionally accurate trace of a profile of a surface from a captured pixellated image of the profile. The method of calibration involves capturing a pixellated image of a calibration grid, detecting pixel co-ordinates of predetermined points on the image, determining distance between selected pairs of points, determining local scale factors for a number of points on the captured image by comparing pixel distances associated with those points with known dimensions of the calibration grid; predicting scale factors for pixels of the captured image on the basis of the predetermined scale factors and applying the predicted scale factors to pixels of the image to convert pixel co-ordinate data to dimensional co-ordinate data. The step of predicting scale factors may comprise generating an nth degree polynomial equation to relate local scale factors to pixel co-ordinates and finding polynomial coefficients which best fit the measured scale factor values and their associated pixel co-ordinate data. The polynomial is then applied at each pixel location to determine a scale factor for that location.

Description

INTELLECTUAL

. .... PROPERTY OFFICE Application No. GB09075 17.7 RTM Date:26 May 2009 The following terms are registered trademarks and should be read as such wherever they occur in this document: Microset Intellectual Property Office is an operating name of the Patent Office www.ipo.gov.uk

I

GOLF CLUB TESTING

The game of golf is governed by The R&A, and by the United States Golf Association. These organisations s promulgate rules and regulations governing the playing of the game, and governing the equipment which is allowed to be used within the "Rules of Golf" (hereafter "the Rules") as published by The R&A of Beach House, Golf Place, St Andrews, Fife KY16 9JA.

In order to determine whether a golf club conforms to the Rules, a sample of the club is submitted to the governing bodies for testing. This testing includes accurate dimensional measurement of the club heads, to ensure that they satisfy the requirements of the Rules.

One feature of a club head which is strictly controlled is the generally horizontal parallel grooves which may be provided across the club face.

The shape and dimensional relationship of these

S

grooves is laid down in the Rules, and accurate measurement of the grooves must be carried out in order to determine whether a submitted sample of a particular make and model of club satisfies the dimensional requirements.

The inetrology equipment currently used for club face measurement is expensive and cumbersome, and its use is time-consuming, such that in practice accurate measurements are only undertaken at the premises of the governing body or of a manufacturer of clubs. A difficulty can arise if, for example, a golfer seeks to participate in a tournament but arrives at the tournament venue with a model of club which has not been previously tested and found to conform, or has been modified from its original manufacturer's design.

In such cases, the tournament official's choice is either to prohibit the. player from using that particular club in the tournament, or to make a "duration of competition" decision based on a measurement of the club and permit the player to use that club in that tournament. However, the manual

S

implements currently used for club measurement at tournament venues are not as accurate as the equipment used at the premises of the governing body, and disputes can arise if a club is refused for play on S the basis of measurements made with this manual equipment.

Furthermore, rule changes may result in the present portable equipment being incapable of providing sufficiently comprehensive measurements to permit even a "duration of competition" decision to be made.

There therefore exists a need for robust and portable yet accurate equipment for measuring golf club faces to determine whether or not the grooves conform to the Rules. Furthermore, there exists a need for a method of measuring golf club faces which can be performed either at a playing venue or by a manufacturer to give an accurate and reliable measurement of the golf club face.

The present invention seeks to provide a method for accurately measuring golf club faces using equipment which is portable yet maintains an accuracy level comparable to the equipment used at the premises of the governing body.

One aspect of the present invention seeks to provide a method for measuring a golf club face to determine conformance with the Rules, the method comprising the steps of: producing a cast of the surface to be measured; producing from the cast a specimen for measurement; capturing an image of a profile of the specimen; and determining the dimensions of features of the profile from the captured image.

The dimensions may then be compared with predetermined values set out in the Rules, and conformance or otherwise may be determined on the basis of the comparison.

Preferably, the cast is cut to produce a specimen or tslice' of predetermined width and having a profile corresponding to the profile of the surface to be measured. The cutting is preferably by two blades, operated sequentially or simultaneously. Most preferably, the two blades are set parallel to one another.

A second aspect of the invention provides apparatus for preparing a measurement specimen for measuring a grooved surface, the specimen having a surface corresponding to the surface to be measured.

According to this second aspect, apparatus for preparing a measurement specimen for measuring a surface comprises a casting dam and a sectioning tool, the casting dam comprising a frame having a front surface adapted to substantially correspond to a surface to be measured and a rear surface, a cavity of predetermined shape and dimensions and open to the front and rear surfaces of the frame, and a cap adapted to close the cavity at the rear surface of the frame, and the sectioning tool comprising a base having a recess corresponding in dimensions to the cavity and cutting means for cutting through a sample placed in the recess to produce a slice of a S predetermined thickness.

A third aspect of the invention provides a technique for calibrating a visual image capture device prior to conversion of captured image data from a specimen to dimension data of the specimen.

According to this third aspect, a method for calibrating a visual image capture device comprising the steps of: capturing a pixellated image of a calibration grid; detecting pixel coordinates of predetermined points on the image; determining pixel dimensions between selected pairs of points; determining local scale factors for a number of points on the captured image by comparing pixel distances associated with those points with known dimensions of the calibration grid; predicting scale factors for pixels of the captured image on the basis of the determined scale factors; and applying the predicted scale factors to pixels of the image to convert pixel coordinate data to dimensional coordinate data.

Embodiments of the invention will now be described in detail with reference to the accompanying drawings, in which: Figure 1 is a schematic view of an "iron" golf club head; Figure 2 is a cross-section of the golf club head, on the line Il-Il of Figure 1; Figures 3a, 3b and 3c show the parts of a moulding dam assembly for use in the method of the invention in perspective view;

S

Figure 4 is a perspective view of the moulding dam in position on a golf club face; S Figure 5 is a sectional view on the line V-V of Figure 4; Figure 6a shows the moulded lozenge when removed from the moulding dam; Figure 6b shows the moulded lozenge after removal of the moulding flash; Figures 7a and 7b are side views of a specimen cutting apparatus in an initial, loading, position and a final, cutting, position respectively; Figure 8 is an end view of the specimen cutting apparatus of Figure 7; Figure 8a is a view similar to Figure 8, of an alternative embodiment of the specimen cutting

S

apparatus; Figure 9 is a perspective view of a cut specimen; Figure lOa is a perspective view of a calibration screen; Figure lOb is a diametral sectional view of the calibration screen of Figure lOa; Figure 11 is a schematic view of the calibration screen in use in an optical imaging apparatus; Figure 12 is a schematic view of the cut specimen in position in the imaging apparatus for image capture; Figure 13 is a schematic drawing of the profile of a striking surface, to show the notation used; Figure 14 is a partial view of an image of a calibration grid, to show the notation used in the

description of the calibration method;

S

Figures 15 A to 15 D show steps in the measurement process of a profile trace; Figure 16 is a view, to an enlarged scale, of a part of a trace of a profile of a club head groove, to illustrate steps in the measurement process; Figure 17 shows a groove and a land area of a trace to a larger scale; and Figure 18 illustrates the determination of groove edge radius.

Referring now to the drawings, a method of measuring the grooves in a golf club face for conformance with the Rules will now be described.

Figures 1 and 2 show a front view and a sectional view of a golf club head 1, respectively. The golf club head has a toe end 2 and a heel end 3, at which is formed a hosel 4 for attaching the club head 1 to a

S ii

shaft (not shown) . Between the toe end 2 and the heel end 3 is a striking surface 5. The striking surface 5 is substantially flat, and is formed with a number of substantially parallel grooves 6 extending generally S horizontally across the striking surface 5.

The Rules set out constraints for the spacing between the grooves, the depths and widths of the grooves, the ratio of groove cross-sectional area divided by groove pitch, and the sharpness of the edges where the grooves meet the striking surface. Figure 13 illustrates the notation used in groove measurement.

The width W of each groove 6 is defined as the distance between the intersection points where a line at 300 to the striking surface 5 contacts the rounded edges of the groove. The spacing between grooves is determined, for some measurement purposes, to be the smaller of the distances Si and S2 between adjacent intersection points of two adjacent grooves, and the depth D of the groove is the perpendicular from an extension of the striking surface 5 to the lowest point in the groove cross-section.

In order to measure the profile of the grooves using the method and apparatus of the present invention, a moulding dam such as is illustrated in Figures 3a to S 3c is used to make a cast of a part of the striking surface 5 of the golf club head 1. A specimen is then prepared from the cast, and an image of the specimen is captured. Dimensional measurements are then made from the captured image, and these measurements are compared with the constraints set out in the Rules to determine whether the dimensions of the specimen, and therefore the dimensions of the club face grooves, conform to the rules.

Preparing the Specimen The moulding dam illustrated in Figure 3 comprises a generally rectangular frame 7 having two end pieces Ba and 8b and two side pieces 9a and 9b surrounding a central opening 10 of width w and length 1. At the intersection of each side piece and end piece, a recess 11 is formed in the upper surface of the frame 7. The recesses 11 reduce the height of the side

I

pieces 9a and 9b to a depth d. The length 1, depth d and width w of the central opening 10 are chosen to produce a moulded sample lozenge of a predetermined size.

At approximately the mid-points of the side pieces 9a and 9b, threaded holes are formed through the side pieces to receive grub screws 12, which extend through the side pieces and protrude beneath the frame 7.

Preferably, the grub screws 12 have conical ends, with their points protruding below the frame 7.

The moulding dam assembly also includes a cap having a generally flat undersurface 14 of length 1 and width w to fit within the frame 7. The undersurface 14 may be formed with a projection 14a (seen in Figure 3b) or with a transverse ridge (not shown) near one end of the undersurface 14, whose purpose will be described later. The cap is generally part-cylindrical in shape, the end faces 15 of the cap forming gripping surfaces for a user to hold the cap.

The third part of the moulding dam assembly is a release lever 16, shown in Figure 3c. The release lever will be described in detail later.

S In order to prepare a specimen, the frame 7 is placed with its front surface on to the face of the golf club head, as seen in Figures 4 and 5. The conical ends of the grub screws 12 engage one of the grooves 6 in the golf club head, to ensure that the longitudinal dimension 1 of the opening 10 is aligned substantially perpendicularly with the directions of the grooves 6.

The frame 7 may be held on to the golf club head 1 by manual pressure, for example on one or both of the end pieces Ba, Sb. Alternatively, a clamping arrangement (not shown) may be provided to engage a rear surface of the golf club head to retain the frame 7 in position. Other securing methods may be used, if suitable. For example, on metal clubs the dam may be held in place by magnetism.

The front surface of the frame 7 may be flat, or may be concave or convex in order more closely to follow the surface of a club having a curved face A settable moulding compound is then placed in the opening 10, in sufficient quantity to ensure that the opening is entirely filled with the compound, up to a depth above the recesses 11. Any suitable settable casting or replicating liquid may be used, but a preferred material is the synthetic rubber replicating compound commercially available under the name "Microset Grey 1O1RF".

When the opening or cavity 10 has been filled with liquid replicating compound, the cap 13 is aligned with the opening and pressed downward, to force the replicating compound into the grooves 6 of the striking surface 5. If the cap 13 is provided with a projection 14a, then the cap is preferably positioned so that the projection is adjacent the top of the striking surface of the golf club. The recesses 11 are shaped to allow some excess compound to escape, and the grub screws 12 may be positioned to provide a small clearance between the underside of the frame 7

S

and the striking surface 5 of the golf club head, so that casting compound is expelled between the frame 7 and the striking surface 5. Preferably, the grub screws 12 are adjusted to minimise the gap between the frame 7 and the club face, at the same time ensuring that the conical ends of the grub screws engage both edges of a groove so as positively to locate the frame relative to the club face. The cap, frame and golf club head are then held in position until the casting compound has set. Typically, for the compound identified above, this takes approximately 10 minutes.

When the casting compound has set, the frame 7 is removed from the golf club head, using the release lever 16. The release lever 16 comprises a cylindrical body 17, from one axial end of which extends an eccentrically-positioned pin 18. A lever arm 19 extends radially from the body 17 of the release lever. In order to release the frame 7 from the golf club head 1, the pin 18 is introduced into an opening 20 formed in a side piece 9a of the frame 7.

The lever arm 19 is then swung so that the body 17

I

rotates about the axis of the pin 18, and due to the eccentricity of the pin 18 the body 17 acts as a cam to push downward on the striking surface 5 of the golf club head, urging the frame 7 of the moulding dam away from the golf club head. Any adhesion between the replicating compound and the golf club head is overcome in a controlled manner, and the frame 7 with the solidified replicating compound is detached from the golf club head.

The cap 13 is then pulled away from the moulding dam, to leave a "lozenge" 20 of solidified replicating material within the opening 10 of the frame 7. The lozenge is ejected from the frame by downward pressure, resulting in the lozenge 20 as seen in Figure 6a. Excess moulding flash 21 is trimmed away, to arrive at the final lozenge seen in Figure 6b. The lozenge 20 has a generally rectangular body of length 1, width w and depth d, and on one face there are ridges 22 which correspond in size, shape and spacing to the grooves 6 of the striking face 5 of the golf club head.

If the cap 13 is provided with a projection 14a, then the lozenge will be formed with a divot 14b (seen in Figure 6a) positioned near one end, which will indicate the end of the lozenge cast from adjacent the top of the club face. If the cap 13 is provided with a transverse groove or projection, then a transverse projection 14b (seen in Figure 6b) or a transverse groove (not shown) will identify the end of the lozenge cast from the upper end of the club face.

The sample is finally prepared for measurement by cutting a longitudinal slice from the lozenge. In this embodiment, the lozenge 20 is cut in a slicing apparatus 23 seen in Figures 7 and 8.

Referring now to Figures 7 and 8, Figures 7a and 7b are side views of the slicing apparatus 23, at the beginning and end of its cutting stroke respectively.

Figure 8 is an end view of the slicing apparatus of Figure 7a.

The slicing apparatus 23 comprises a base block 24, and two pairs of swinging arms 25 and 26 extending upwardly from respective sides of the base block 24, the arms of each pair being pivotably mounted adjacent respective ends of the base block 24. A stepped slot 27 extends inwardly from one side face of the base block 24, the width and depth of the stepped slot 27 being such that the lozenge 20 can be introduced into the slot 27, with the length 1 of the lozenge extending across the slot, and pushed up against an end stop 28. The ends of the lozenge are received on steps 29 extending along the length of the slot 27.

The steps 29 of the slot provide a clearance between the lozenge and the slot, to ensure that the ridges 22 of the lozenge are not damaged during insertion of the lozenge into the slot. In an alternative embodiment, the steps 29 may be formed at the upper corners of the slot (as seen in Figures 7a and 7b), so that the flat face of the lozenge is supported by the lower wall of the slot, and the ridged surface of the lozenge is uppermost.

I

A cutting block 30 is attached to the upper ends of the swinging arms 25, 26. In the embodiment shown in Figure 8, the cutting block 30 is longitudinally divided into two cutting block portions 31 and 32.

Each cutting block portion has mounted thereto a blade 33, 34, extending towards the base block 24. A pair of slits 35 and 36 are formed in the base block 24, each slit being aligned with one of the blades 33 and 34. The slits 35 and 36 intersect with the slot 27, in planes spaced by predetermined amounts from the end stop 28 of the slot 27.

Referring now to Figures 7a and 7b, the cutting block portions 31 and 32 may be moved downwardly and to the right (as seen in the Figure) as the pairs of swinging arms 25 and 26 rotate clockwise. This movement of the cutting block portions 31 and 32 brings the blades 33 and 34 in an arcuate movement into the slits 35 and 36, eventually to the position shown in Figure 7b. In this position, the cutting edges 37 of the blades 33 and 34 will have passed through the slot 27, and will have cut through a lozenge 20 placed in the slot 27.

In the region between the blades 33 and 34, a t1slice" of the lozenge 20 will have been defined and separated from the remainder of the lozenge. The cutting blades 33 and 34 are positioned such that if a divot 14a is S formed in the lozenge, it is positioned between the cutting lines, and will be apparent on the face of the cut slice opposite to the ridges 22.

In the embodiment shown in Figure 8, the cutting block 30 is formed as two separate cutting block portions.

This allows the blades 33 and 34 to be brought down independently through a lozenge 20. In practice, a cleaner cut has been found to result if the blade 33 closest to the end stop 28 is brought down first, and subsequently the blade 34 is brought down to sever the part of the lozenge 20 to the side of the blade 33 remote from the end stop 28. This sequence of cuts also ensures that the part of the lozenge severed by the second cutting operation is not compressed or restrained during cutting, and the resulting "slice" corresponds in width to the distance T between the blades 33 and 34.

Figure Ba shows a modified type of cutting apparatus, in which the cutting block 30 is a single cutting block, on which the two blades 33 and 34 are mounted.

In this arrangement, the cutting edges 37 of the blades may be set at different heights from the cutting block 30, SO that as the two blades descend simultaneously, the blade 33 closest to the end stop 28 cuts through the lozenge 20 before the blade 34 contacts the lozenge. The cutting edges of the blades may be symmetrical with two inclined faces, or may be "chisel" edges with one inclined face and a vertical face.

Another feature of the cutting apparatus of Figure 8a is that the slot 27 is formed in the base block of 24 at an angle to the direction of movement of the blades 33 and 34. Preferably the slot 27 is formed in the base block 24 such that the angle a between the slot axis and a plane perpendicular to the cutting planes of the blades is between 00 and 100, preferably about and most preferably from 10 to 30* This results in

S

the "slice" cut from the lozenge having a parallelogram-shaped cross-section, the advantage of which will be explained below. The angled slot 27 may be used in combination with the divided cutting block illustrated in Figure 8.

After the blades have been advanced through the lozenge, the cutting block 30 is retracted to the position shown in Figure 7a and the cut pieces of the lozenge are removed from the slot 27. This may be done by inserting an ejecting pin (not shown) into an ejector bore 38 aligned with the slot 27 and extending through the end stop 28, to push the lozenge parts out of the mouth of slot 27.

Figure 9 illustrates the cut "slice" 39 of the lozenge 20, ready for the image capture operation. In order to capture the image, an optical system with very small depth of focus is used, in order to maximise the resolution of the optical system. The optical system is used to capture an image of the cut end face 40 of the slice 39. However, in order that accurate measurements may be taken from the captured image of the end face 40 of the slice 39, it is necessary first to calibrate the image capture apparatus so that an accurate transformation between pixel dimensions in the captured image and physical measurements of the sample may be determined.

Figure lOa illustrates, in perspective view, a calibration grid 41 for use with the optical image capture apparatus. Figure lOb is a diametral section of the calibration grid of Figure lOa.

Referring to Figure 10, the calibration grid 41 comprises an outer ring 42, which supports the edge of a circular transparent screen 43, on the upper surface of which is formed a pattern of indicia 44. In the embodiment shown, the indicia 44 comprise a regular pattern of circular dots. In a preferred embodiment, the dots are laid out in an orthogonal grid of ranks and files, with each dot having a diameter of from 0.5 to 2 mm and with a spacing of from 2 to 5 mm between centres of adjacent dots in each rank and file. The

S

pattern of dots preferably extends over the entire upper surface of the screen 43.

The diameter of the screen 43 is such that the end face 40 of the slice 39 of the lozenge can be imaged in a picture field which has been entirely covered by the screen 43. In other words, the diameter of the screen 43 is preferably greater than the length 1 of the cut slice 39.

The ring 42 is dimensioned so that the upper surface of the screen 43 on which the pattern of dots 44 is formed is spaced from the undersurface 45 of the ring by a distance T, equal to the thickness of the cut slice 39. The reason for this will be explained below.

Calibration of the Image Capture Device The correlation between the captured image and the physical measurements of the sample is determined by a calibration operation for the image capture device, which will now be described.

S

Calibration of the image capture apparatus is effected by placing the calibration grid 41 on top of the light box 46, such that the upper face of transparent screen 43 is positioned at the focal plane FP of the camera 48. This places the grid pattern of image elements or dots 44 exactly on the focal plane, so that a sharp image of the grid of dots can be captured by the camera.

Figure 14 schematically shows the image of the calibration mask grid which is captured by the camera 48. The image comprises a plurality of circular dots, arranged in vertical ranks and horizontal files (as shown in the Figure) . The image is captured by the image capture element of the camera, which is a rectangular array of pixels arranged in rows and columns, such that each pixel can be identified by coordinate information derived by counting pixels from an origin at one corner of the image capture element to derive the "x" and 11y1! coordinate values of each pixel.

S

The image data is first processed to detect the dark areas, which correspond to the dots 44 of the calibration grid 41. Each dark area is then examined to detect which of the dark areas are circles, and which are not. The dark areas which are not circles correspond to dots 44 which have not completely fallen within the field of view of the image, and these dark areas are discarded for the purposes of further image processing.

Each dark area of the image is then taken in turn, and the centroid of the dark area is found, this position corresponding to the centre of the circular dot 44.

The centroid position is not necessarily an integer number of pixels from the measurement origin Each dot 44 is then taken in turn as a subject dot Si, and the nearest dots Ni and N3 in the same rank and N2 and N4 in the same file are identified. Vectors Ri, R2, R3 and R4 are drawn from the centre of the dot Si to the centres of its four nearest dots Ni, N2, N3 and

S

N4, respectively. The four vectors are then added, and if the vector sum is zero or near zero, then it is determined that the dot in question is the centre of a "cross" made up of five dots.

From the pixel coordinates of the centres of each of the five dots of the "cross", the pixel lengths from the centre of the subject dot Si to each of the centres of the other four dots Nl-4 are determined. A value representative of these four lengths (in numbers of pixels) is then given the value R, and the value R for each dot represents the pixel distance in the image between the centre of that dot and the centres of its four adjacent dots. The value R may be the mean of the four lengths, or any other statistical representation of the four lengths.

Since the actual distance between the centres is known, from the physical dimensions of the calibration grid, the value R is a measure of the local scale of the image (i.e the physical dimensions represented by each image pixel) because the number R of pixels corresponds to the actual length between centres of the grid dots.

The result of this calculation is a set of values of R, each of which corresponds to the x, y coordinates of a pixel in the image. This can be expressed as a data function G, namely: R = G(x,y1) where x1 and y are the coordinates of a pixel in the image, and R1 is the value of the local scale R at that pixel.

If the function G is expressed as a polynomial, selected such that: = a(x)2 + b(y1) 2 + c(xj.yj) + d(x1) + e(y1) + f then the first three terms of the polynomial G represent circular distortions in the image, the fourth and fifth terms represent planar tilt of the

S

image, and the final term represents the average pixel size in the image.

Alternatively, a cubic or quartic polynomial in the S two variables x and y could be chosen.

By taking each of the values of R1 and its corresponding x and y and applying a suitable "best fit" algorithm, one can determine the coefficients of the polynomial, in this case a to f, for the set of calibration image data. By using these "best fit" coefficients, and applying the function to each pixel (not just those pixels where values of R are known), a scaling coefficient for each point on the pixel image is determined. The physical coordinates represented by the pixel point are derived by applying the scale factor to the pixel coordinates of the image at each pixel point, so that the image exactly corresponds in physical dimensions to the dimensions of the calibration grid at each pixel point.

Once the set of coefficients, in this case a to f, has

S

been determined for a particular image capture device, an image of an object produced by the image capture device can have the function applied to its pixel data in order to produce image data representing physical dimensions which exactly correspond to the physical dimensions of the object.

In the calibration method described, the calibration grid is an array of equally-spaced dots as seen in Figure 14, in which the dots are arranged in ranks and files and the array extends over the entire visual field of the image capture device. Scale values R for a number of points x,y of the image are calculated on the basis of distances between the centre of a dot and the centres of four dots adjacent to, and in the same rank and file as, the subject dot. It will however be understood that a local scale value R can be calculated by measuring distances from a subject dot S2 to four dots N5, N6, N7 and N8 diagonally adjacent the subject dot, as seen in Figure 14 where the distances R5, R6, R7 and R8 are measured relative to the subject dot S2. The local scale value R for subject dot S2 may then be determined from the sum, mean, median or another statistical value related to R5, R6, R7 and R8. The calculation of the local scale value R may, as appropriate, involve dividing the S measured lengths R5, R6, R7 and R8 by a scaling multiple such as i/2.

In an alternative calibration method, the scale value R for a point x,y of the image may be determined by measuring the distance from the centre of a dot Si to the centre of one neighbouring dot, for example Ni or N2. This however will give a local scale value R representative only of the local scale in one direction. More preferably, the scale value R for a point x,y of the image may be determined by measuring the distance from the centre of a dot Si to the centre of an adjacent dot Ni or N3 in the same rank, and the distance to the centre of an adjacent dot N2 or N4 in the same file. For each point x,y this will give a value of R which is representative of the local scale in both vertical and horizontal directions.

S

In a further alternative embodiment, the local scale value R may be determined on the basis of the distances from the centre of a dot S2 to the centres of eight or more adjacent dots N5 to N12. In such an arrangement, distances R5 to R8 measured in directions which are diagonal relative to the directions of the ranks and files may have a scaling factor 12 applied to them before they are combined with distances R9 to R12 made in directions parallel to the ranks and files to yield the scale value R. It is foreseen that the scale values R at respective points x,y may be calculated on a yet different basis, for example by determining the centre of a dot as the point x,y and measuring the diameter of the dot to yield a scale value R. The diameter of the dot may be measured in two or more directions, and the scale value R calculated as the sum, the mean or the median of the measured values. In one embodiment, the diameter of the dot may be measured in two orthogonal directions.

S

As an alternative to grid pattern is comprising orthogonal ranks and files, a hexagonal array of dots may be used and the selection of dots for dot-to-dot measurement adjusted accordingly.

S

It is foreseen, however, that a calibration grid having a single dot of known dimensions (for example using its diameter as R and the coordinates x,y of the centre of the dot), or two dots of known dimensions at a known spacing from one another, or a pattern of three dots of known dimensions at a known spacing from one another (for example in an L"shaped array) may be used as a calibration grid. Since many fewer values of R and its corresponding x1 and y will be gathered by such an arrangement, the accuracy of the calibration will be less than optimum. The accuracy of such a calibration grid may be improved by capturing multiple images of the calibration grid, and moving the calibration grid relative to the image

field between image capture operations so that the

successive images have the dots positioned in

different regions of the image field. The pixel

coordinates of the dots may then be determined, and amalgamated to produce a data set in which sets of pixel coordinates from many regions in the image field have associated local scale values. A grid having a S few dots, or only one dot, may thus be used to generate local scale data for many locations in the

image field.

In the described embodiments, the calibration grid is composed of circular dots. It is however foreseen that the calibration grid may be composed of image elements which are square, triangular, hexagonal or any other suitable geometric shape.

Capturing The Sample Image The optical measuring apparatus is shown schematically in Figures 11 and 12, and comprises a light box 46, the upper surface 47 of which is adapted to emit a diffused light, and an image capture apparatus 48 such as a digital camera.

The image capture apparatus 48 preferably has a very small depth of field centred about the focal plane FP, and is mounted above the light box 46 such that the focal plane FP of the camera is spaced from the upper surface 47 of the light box by the distance T. As was described above in relation to Figure lOb, the distance T is equal to the thickness of the cut sample slice 39, and the distance between the screen pattern 44 and the underside of the calibration grid 41.

To capture an image of an end face of the cut sample slice 39, the slice is placed on the light box 46 such that the surface of the slice bearing the ridges 22 is either vertical, or preferably slightly overhanging due to the oblique angle a at which the slice was cut.

If the cut slice has a divot mark 14b, then correct orientation of the cut slice on the light box can be ensured by arranging the divot mark to be at a predetermined end of the slice when the slice is placed on the light box with the ridged surface facing away from the operator. Alternatively, the divot mark 14b may be in the form of an arrow, which the operator arranges to point towards (or away from) the light box. In a further alternative the divot mark l4b may take the form of writing, which the operator arranges to be the right way up when the specimen is placed on the light box. In a yet further alternative (not illustrated) the cutting blades 33 and 34 may not be parallel but may be arranged to cut the sample in convergent planes, so that whichever cut surface is placed on the light box, the profiled surface of the sample will have an overhang.

Preferably no adjustment is made to the camera or light box following the calibration of the image capture equipment, since the end of the slice to be imaged is at the same spacing above the light box as was the calibration screen, and thus will be in the focal plane of the camera.

The image capture apparatus 48 may then be used to capture a precise image of the end face of the cut sample slice 39, which is positioned at the focal plane FP of the camera. The overhang beneath the edge of the upper face of the sample ensures that the profile of the sample is clearly defined on the captured image.

Measurement of the Sample Image S The image of the cut sample consists of a dark region corresponding to the outline of the end face 40 of the sample slice 39 seen in Figure 9. This image is converted to sample trace, which is a line representing the edge 40a of the sample slice, by edge detection software acting on the image. The sample trace 60 seen in Figure 16A thus comprises relatively straight regions 61 corresponding to the lands between the grooves, and curved regions 62 or "bulges" corresponding to the grooves in the golf club face.

The polynomial G, with the coefficients determined by the calibration process for the image capture device, is then applied to the pixel data of the image to produce a sample trace whose physical dimensions correspond to those of the sample slice 39.

The first step in the measurement process is to align

S

the sample trace with the coordinate axes to be used for measurement. This may be done by finding a straight line BFS of best fit to the sample trace.

This straight line BFS, seen in Fig 15A, will be S inclined at some angle B to the coordinate axes, and by rotating the trace through this angle the trace can be aligned to the measurement axes.

In cases where of the face of the golf club is not flat, the sample trace is then processed to straighten and align the regions of the trace between the grooves. This may be done by finding a curve BFC, conveniently a cubic curve, seen in Fig 16C, which best fits the sample trace, and then subtracting the curve BFC from the trace data to produce a flattened representation of the sample trace, seen in Fig 16D.

The next step is to find the position of the "centre of gravity" or centroid CG of the points on the flattened and rotated sample trace line. This centroid CG will be approximately mid-way along the length of the trace, and it will be slightly off the trace line, on the side where the line "bulges" away from straight.

For the measurement algorithm to operate correctly, the trace must be oriented so that the curved regions 62 extend downwardly from the flat areas 61 of the trace, as seen in Figure 16.

In order to determine which way the bulges extend from the straight part of the trace (i.e. "which way up" the trace is), an orientation line OL is constructed passing through the centroid CG and parallel to the now flattened and if necessary rotated trace, as seen in Fig 16D. A statistical representation of the distance h from the orientation line OL to the trace in one direction is then compared to the corresponding statistical representation of the distance d from the orientation line OL to the trace, in the other direction. The larger of these two measurements will be the measurement to the tip of the bulge, i.e. the measurement from the orientation line OL down into the grooves of the golf club head.

If the larger measurement is "above't the line OL, then the image is "flipped" about the line OL to achieve the correct orientation for the subsequent measuring process.

Distances from the orientation line OL to the trace in the now downward direction as seen in Figure 17 are then made from a plurality of points on the orientation line OL between adjacent intersections of the orientation line and the trace. The 95th percentile of these measurements, i.e. the value which 5% of these depth measurements are greater than, is taken to represent the nominal depth d of the groove.

The points on the trace which are spaced from the orientation line OL by d/2 (one half of the nominal depth d) are then designated as transition points TP marking the transition between a groove and a land area of the trace. The distance L between the transition points TP is then used to draw a "lid" CL across the groove, by taking points Qi and Q2 on the

S

trace which are each spaced by a distance AL (typically 1.5 to 2 times L) outside each transition point TP, and drawing a line CL between the two points Qi and Q2. The points Qi and Q2 may be arrived at by finding the centroid of a number of adjacent points on the trace, in order to smooth out local roughness of the trace at the points Q. The maximum distance between the line CL and the trace, measured perpendicularly to the line CL, gives the depth D of the groove.

The area between the trace and the line CL is then found, and this area is the cross-sectional area A of the groove. The area may be found by simply counting the pixels within the area bounded by the trace and the line CL, or some other integration method may be used.

The width W of the groove is determined by drawing two lines at 30° to the line CL, and which are tangents to the trace, as seen in Figure 13. The width W is the

S

distance between the points of contact between the two 300 lines, and is not necessarily the same as the distance L between the transition points TP. The distance between the points of contact and the line CL is also measured, and if this distance exceeds 0.003 inches or 0.0762mm, then points on the trace which are 0.003 inches or 0.0762mm from the line CL are found, and the distance between these points is measured and taken as the width W of the groove.

The pitch of the grooves is the distance from a point on one groove to a corresponding point on the next, i.e. the width W of a groove plus the separation distance S across the land area between the 30° tangent points. It the groove separation distances Si and S2 across land areas on either side of a groove are different, the smaller of the two distances is taken, and added to the width W to calculate the pitch of the grooves.

A further criterion relating to groove dimensions is the edge radius. To determine the edge radius, a first circle 100 of radius 0.010 inches or 0.254mm is constructed below and tangential to the land area 40a of the trace, as is seen in Figure 16. The circle is then moved towards the groove part of the trace, until the circle touches the trace at two points Ii and 12.

A second circle 102 is constructed concentric with the first circle 100, the second circle having a radius of 0.011 inches or 0.279mm. This second circle may, as shown, intersect the trace at points 13, 14, 15 and 16. If there are fewer than four intersects, then this edge meets the edge sharpness constraint. If the second circle intersects the trace at four points, and the two central intersection points 14 and 15 subtend an angle R of 100 or less at the centre of the circle, then the edge meets the edge sharpness constraint. If the angle R subtended by the two central intersection points 14 and IS is a greater than 10°, then the edge is non-conforming. Likewise, if a third concentric circle 103 is drawn with a radius of 0.0113 inches or 0.287mm, then if the trace goes outside this circle the edge is non-conforming. In the example shown in Figure 16, the circle 103 intersects the trace at only

S

two points, 17 and 18, and the edge does not fail on this constraint. More than two intersects between the circle 103 and the trace will indicate a non-conforming edge.

The circles may be constructed and appropriately placed adjacent the sample trace by a processor controlled by software, and the processor may locate and count intersects and to determine their positions and measure angles subtended at the centre of the circles, in order to determine whether the edge is a conforming edge or not. The software may then make a determination as to whether the club satisfies the constraints of the Rules or not, and may display that determination on a screen or as a paper printout or in some other form.

The measurements of dimensions from the sample trace, and a calculation of areas and ratios based on those measurements, may all be implemented in software, and the software may further cause the calculated and measured dimensions to be compared with respective limiting values as set out in the Rules, in order to arrive at a determination as to whether a sample trace represents the profile of a club which could conform to the current regulations if all other requirements of the Rules are met. The determination of conformance or nonconformance may then be displayed to a user, and/or a print of the measured dimensions, calculated values and the determination result may be provided to the user as a record

Claims

Claims: 1. A method for measuring a profile of a surface, the method comprising the steps of: producing a cast of the surface to be measured; S producing from the cast a specimen for measurement; capturing an image of a profile of the specimen; and determining the dimensions of features of the profile from the captured image.
2. A method according to claim 1, further comprising the steps of; comparing the determined dimensions with a predetermined set of dimensional requirements; determining conformance or otherwise with the requirements on the basis of the comparison.
3. A method according to claim 1 or claim 2, wherein the cast is cut to produce a specimen or "slice" of predetermined width and having a profile corresponding to the profile of the surface to be measured.
4. A method according to claim 3, wherein the cutting is performed by two spaced blades, cutting the cast sequentially or simultaneously.
5. A method according to claim 4, wherein the cutting of the cast is in planes set at an angle to the surface of the sample whose profile is to be measured.
6. A method according to claim 5, wherein the angle is from 0 to 10 degrees.
7. A method according to any of claims 3 to 6, wherein the cast is cut in two parallel planes.
8. Apparatus for preparing a measurement specimen for measuring a surface, the apparatus comprising: a casting dam; and a sectioning tool, wherein the casting dam comprises: a frame, having a front surface adapted to substantially correspond to a surface to be measured and a rear surface; a cavity of predetermined shape and dimensions and open to the front and rear surfaces of the frame; and a cap adapted to close the cavity at the rear surface of the frame; and the sectioning tool comprises: a base having a recess corresponding in dimensions to the cavity; and cutting means for cutting through a sample placed in the recess to produce a slice of predetermined thickness.
9. An apparatus according to claim 8, wherein the cutting means comprises a cutting blade movable relative to the base.
10. An apparatus according to claim 8 or claim 9, wherein the recess includes a stop surface for locating a sample placed in the recess.
11. An apparatus according to claim 10, wherein the stop is adapted to locate the sample in a direction perpendicular to the plane of the cutting blade.
12. An apparatus according to claim 10 or claim 11, wherein the cutting blade is movable relative to the stop in a direction perpendicular to the cutting direction between first and second cutting positions separated by the predetermined thickness of the slice.
13. An apparatus according to claim 10, wherein the stop is movable relative to base between first and second cutting positions separated in a direction perpendicular to cutting direction by the is predetermined thickness of the slice.
14. An apparatus according to claim 10, wherein the cutting blade is movable relative to the base between first and second cutting positions separated in a direction perpendicular to cutting direction by the predetermined thickness of the slice.
15. An apparatus according to claim 10 or claim 11, wherein the cutting means comprises first and second parallel cutting blades spaced by the slice thickness.s
16. An apparatus according to claim 15, wherein the first and second cutting blades are fixed relative one to the other.
17. An apparatus according to claim 10, wherein the first and second cutting blades are independently movable relative to the base in the cutting direction.
18. An apparatus according to any of claims 10 to 17, wherein the recess is arranged so that blade cuts sample at an oblique angle.
19. An apparatus according to claim 18, wherein the sample is cut at an angle of from 0 to 10° from a plane perpendicular to the surface of the sample corresponding to the surface to be measured.
20. An apparatus according to any of claims 8 to 19, wherein the dam includes locating means for positioning the dam relative to the surface to be measured.
21. An apparatus according to claim 20, wherein the locating means comprises an adjustable projection extending from the front surface of the frame.
22. An apparatus according to claim 21, wherein the locating means comprises two projections adapted to engage a linear feature of the surface to be measured, to align the cavity relative to the linear feature.
23. An apparatus according to any of claims 8 to 22, wherein the dam further comprises cam means engageable with the surface to be measured, for separating the dam from the surface.
24. An apparatus according to any of claims 8 to 23, wherein the darn includes a vent between the cap and the frame to allow egress of fluid from the cavity.
25. An apparatus according to any of claims 8 to 24, wherein the apparatus further comprises a calibration grid comprising a transparent screen with array of indicia formed thereon, and a support for spacing the S screen above a light box by a distance equal to thickness of a slice cut by the sectioning tool.
26. A casting dam for use in the apparatus of any of claims 8 to 25.
27. A sectioning tool for use in the apparatus of any of claims 8 to 25.
28. A method for calibrating a visual image capture device comprising the steps of: capturing a pixellated image of a calibration grid; detecting pixel coordinates of predetermined points on the image; determining pixel dimensions between selected pairs of points; determining local scale factors for a number of points on the captured image by comparing pixel distances associated with those points with known dimensions of the calibration grid; predicting scale factors for pixels of the captured image on the basis of the determined scale factors; and applying the predicted scale factors to pixels of the image to convert pixel coordinate data to dimensional coordinate data.
29. A method according to claim 28, wherein the calibration grid comprises an array of dots of known size and spacing.
30. A method according to claim 29, wherein the array of dots covers entire visual field of image capture device.
31. A method according to claim 29, wherein the array of dots covers part of visual field of image capture device.
32. A method according to claim 30, wherein the array of dots comprises single dot.
33. A method according to claim 31 or claim 32, wherein the step of capturing a pixellated image of a calibration grid comprises capturing plural images each with part of image covered by grid, and superimposing plural images.
34. A method according to any of claims 28 to 33, wherein the step of predicting scale factors comprises: generating an nth degree polynomial to relate local scale factors R1 to pixel coordinates x1 and yjj finding coefficients for the polynomial which best fit the determined scale factor values and their associated pixel coordinate data; applying the polynomial at each pixel location to determine a scale factor for that location.
35. A method according to claim 34, wherein the polynomial is a second, third or fourth degreeSpolynomial.
36. A method according to claim 34, wherein the polynomial is of the form a(x)2 + b(y) 2 + C(Xj.yj) + d(x) + e(y) + f where R1 is the scale factor and x1 and yj are pixel coordinates; and wherein the step of finding coefficients determines coefficients a to f.
37. A method for measuring a profile of a surface, substantially as herein described with reference to the accompanying drawings.
38. Apparatus for preparing a measurement specimen for measuring a surface, substantially as herein described with reference to Figures 3, 4, 5, 7, 8, 10, 11 or 12 of the accompanying drawings.
39. A method for calibrating a visual image capture device, substantially as herein described with reference to the accompanying drawings.
40. A data carrier bearing machine-readable instructions to enable a processing device to carry out the method of any of claims 28 to 37.