EP0805955A1

EP0805955A1 - Method and apparatus for detecting an object

Info

Publication number: EP0805955A1
Application number: EP96900695A
Authority: EP
Inventors: Charles Markham
Original assignee: Industrial Innovation And Development Ltd
Current assignee: Industrial Innovation And Development Ltd
Priority date: 1995-01-25
Filing date: 1996-01-25
Publication date: 1997-11-12
Also published as: WO1996023199A1; AU4458096A

Abstract

In a method of determining the location of at least one point on an object boundary a beam of light (12) is directed at a detector (14) which provides a signal (16) related to the intensity of the spot (13) falling on it. The object (22) is located between the light source (10) and the detector (14), and the light beam (12) is deflectable, under the control of a signal (18) from a controller (20), over an angle alpha so that at some point the light beam crosses the boundary (24) of the object so that the detected intensity of the spot (13) falls from a maximum to a minimum. The signal (16) is fed back to the controller which adjusts the control signal (18) in an attempt to deflect the spot to a position wherein the signal (16) equals a reference value corresponding to the light intensity detected by the detector (14) when the beam is partially obscured by the edge (24). In another embodiment the spot (13) falls on a translucent screen viewed by the detector. The beam may be reflected from the object instead of being transmitted past it.

Description

METHOD AND APPARATUS FOR DETECTING AN OBJECT

This invention relates to a method and apparatus for determining the location of at least one point on an object boundary.

According to the invention there is provided a method of determining the location of at least one point on an object boundary, the method comprising the following steps not necessarily in the order stated:

(a) generating a spot of light deflectable along at least one direction,

(b) providing a control means for generating a first signal ("spot control signal") which controls the deflection of the spot along the one direction,

(c) providing a light detector to generate a second signal ("intensity signal") related to the intensity of the spot as measured at the detector,

(d) placing the object such that the boundary is located at a position where it is able to modify the intensity of the spot, as measured at the detector, at a point within the range of deflection of the spot, and

(e) feeding the intensity signal back to the control means, the control means being responsive to the intensity signal to adjust the spot control signal in an attempt to deflect the spot to a position wherein the intensity signal equals, or falls within a range including, a value corresponding to the modification of die spot intensity by the said boundary.

The invention also provides an apparatus for performing the above method.

In the foregoing and in the appended claims:

"Object" includes solid and fluid entities which may be opaque, translucent, transparent or reflective, or more than one of these in different regions.

"Boundary" includes both a physical boundary (edge) of an object and a line demarcating regions of the object of different transmissivity or reflectivity.

"Light" includes UV and IR wavelengths.

Embodiments of the invention will now be described, by way of example, with reference to the accompanying drawings, in which:

Figure 1 is a schematic diagram of a first embodiment of the invention,

Figure 2 is a schematic diagram of a second embodiment of the invention,

Figure 3 is a schematic diagram of a third embodiment of the invention,

Figure 4 is a schematic diagram of a fourth embodiment of the invention,

Figure 5 is a block diagram of a position-controlled light source for use in the embodiments of Figs. 1-4,

Figure 6 is a block diagram of a light detector for use in the embodiments of Figs. 1-4,

Figure 7 is a block diagram of a controller for use in the embodiments of Figs. 1-4,

Figure 8 is a block diagram of an alternative controller to that shown in Fig. 7 which provides for object scanning in mutually perpendicular X and Y directions,

Figure 9 is a block diagram of an alternative controller to that shown in Fig. 7 which provides for object scanning in radial and angular directions,

Figure 10 illustrates how certain objects may need to be subject to radial scanning about more than one origin,

Figure 11 is a block diagram of circuitry which uses the position signal derived from the embodiment of Figs. 1-4,

Figures 12 and 13 are images which may be derived from the circuitry of Fig. 11,

Figure 14 is a schematic diagram of a fifth embodiment of the invention,

Figure 15 is a schematic diagram of a sixth embodiment using two light beams for determining range information,

Figure 16 is a schematic diagram of a sixth embodiment using one light beam for determining range information, and

Figure 17 is a schematic diagram of a seventh embodiment for measuring the height of a surface.

Referring to the drawings, the embodiment shown in Fig. 1 includes a position-controlled light source 10 for generating a spot beam of light 12. The spot from the light beam 12 falls on a light detector 14 which generates an intensity signal 16 representative of the instantaneous intensity of the spot 13 as measured at the detector 14.

The light beam 12 is deflectable throughout an angular range alpha. For the moment the beam will be assumed to be deflectable only in a single plane, so that the spot 13 correspondingly moves in only a single direction on the detector 14, herein referred to as the Y direction. Modifications which permit two dimensional X-Y movement and movement in radial and angular directions will be described later. The deflection of the beam 12, and hence the movement of the spot 13, is determined by a spot control signal 18 generated by a controller 20.

In the absence of an object 22 interposed in the path of the light beam 12 between the light source 10 and the detector 14, the intensity of the spot 13 at the detector 14 for any position of the beam 12 in its range of angular deflection alpha is at a substantially constant maximum value, and hence the amplitude A of the intensity signal 16 is also at a substantially constant maximum value. Let us call this maximum amplitude Al .

If now an object 22 (which for the moment we will assume to be opaque) is placed so that it intercepts d e light beam 12 at some point within the range of deflection of the beam 12, the amplitude A of the intensity signal 16 will drop to a minimum value AO when the beam is totally obstructed by the object since the intensity of the spot 13, as measured at the detector 14, will be zero. In the absence of ambient light AO will be substantially zero, but the apparatus will work even in the presence of ambient light provided the level of the ambient light is substantially constant at the detector.

Alternatively, as will be described, modulation of the light beam and/or filters can be used to substantially eliminate the effect of ambient light, or the apparatus can be placed in a darkened environment.

For angular positions of the beam 12 where the beam is only partially intercepted by ώe object 22, i.e. the beam 12 "grazes" an edge 24 of the object, the amplitude A of the intensity signal 16 will be some value between AO and Al depending upon the degree of obstruction of the beam and, correspondingly, a function of the spot area falling on the detector 14. Therefore, a value of A between AO and Al will correspond to the beam being located at the edge 24 of the object 22. For example, a value of (Al + A0)/2 will correspond to half the cross-sectional area of the beam being obstructed and only half the full spot area falling on the detector 14 where a circular spot having an intensity profile symmetrical about the centre of the spot is used.

This gives a basis for deteπnining the location of the edge 24 of the object 22, or at least that point on the edge within the plane of deflection of the beam 12. This is achieved by feeding the intensity signal 16 back to the controller 20 and there comparing its instantaneous amplitude A with a reference value which is selected to be equal to a value of the amplitude A which occurs when the beam 12 is at the edge 24. Thus the reference value is selected to have a value between Al and AO and preferably (Al + A0)/2 which is the value which occurs when half the cross-section of the beam 12 is obscured by the edge 24. In response to the comparison the controller 20, for example using a PID (Proportional Integral Differential) loop, adjusts the spot control signal 18 in an attempt to deflect the beam 12, and hence the spot 13, to a position where the amplitude A of the intensity signal 16 equals the reference value or at least falls within a narrow range centered on the reference value, i.e. to a position where the beam is partially intercepted by die edge of the object. The attempt may be unsuccessful, for example if the object 22 is incorrectly placed, but assuming the object is properly placed so that at some point the edge 24 can intercept the beam 12, then the beam ends up positioned at that edge.

Further, provided the gain in the feedback path is sufficiently high, the apparatus can substantially instantaneously track any movement in the edge 24 of the object, the controller 20 constantly adjusting and re-adjusting the position of the beam 12 as the edge 24 of the object moves so as to maintain the beam at that edge. This can have application in, for example, monitoring the undulating surface of a moving fluid.

The apparatus shown in Fig. 1 generates a position signal 26. This is a signal representing the instantaneous angular position of the light beam 12 and, accordingly, represents the Y coordinate of the detected edge 24. As will be described, this position signal 26 is derived from the deflection mechanism of the position-controlled light source 10 or by another mechanism if necessary.

It will be recognised that the embodiment of Fig. 1 requires a relatively large area detector 14, to ensure that the spot 13 falls on the detector over the range of deflection of the light beam 12. Of course, the angle of deflection alpha need not be made greater than is necessary to embrace the expected location of the edge 24, and can be made smaller by bringing the light source 10, object 22 and detector 14 as close together as possible. However, in general practical considerations will still mandate a relatively large area detector.

This requirement is obviated by the embodiment shown in Fig. 2. In this case the spot 13 does not fall directly on the detector 14 but on a translucent ground glass screen 28 located in front of the detector 14. Now, provided the angle of view beta of the detector 14 is sufficient to include the range of deflection of the spot 13 on the screen 28, the detector itself can be of relatively small area. Clearly, as the beam 12 is deflected across the edge 24 of the object 22 the area of the spot 13 on the screen 28 will vary from its full area to zero and the amplitude A of the intensity signal 16 will vary accordingly. Otherwise, the embodiment of Fig. 2 operates in the same manner as that of Fig. 1.

In a modification (not shown) to the embodiment shown in Fig. 2, it is possible to place the detector 14 in a position to view the spot 13 from the same side of the screen 28 as the object 22, in which case the screen does not need to be translucent.

Another possibility for overcoming the need for a large area detector is shown in Fig. 3. In this case a translucent ground glass screen 28 is still used, but the object 22 is located between the screen 28 and the detector 14 rather than between the light source 10 and the screen 28. Clearly, as the beam 12 is deflected across the screen 28 there is no substantial change in the size or intensity of the spot 13 on the screen itself. However, the proportion of the spot area, and correspondingly its intensity, as seen on the screen 28 by the detector 14 will change at that point within the range of deflection of the beam 12 where the line of sight 29 between the spot 13 and the detector becomes obscured by the edge of the object 22. Therefore, as in the preceding embodiments, as the area of the spot 13, as seen by the detector 14, varies from its full area to zero the amplitude A of die intensity signal 16 will vary accordingly. In other respects the embodiment of Fig. 3 operates in the same manner as that of Fig. 1.

Although the preceding embodiments have detected die edge of the object by placing the object in a position to obscure the beam at a certain point in its range of deflection, it is alternatively possible to detect the edge of the object by reflection. An embodiment is shown in Fig. 4.

In this case a reflective object 22 is located to reflect light to the detector 14. When the light beam 12 is falling onto the surface of the object 22, as seen in Fig. 4, the spot 13 falls with full intensity on the detector and the amplitude of the intensity signal 16 is at a maximum Al . As the beam moves across the edge 24 of the object, the amplitude of die intensity signal 16 falls from the maximum Al to a minimum AO as an increasing proportion of the beam cross-section misses the object and is not reflected to the detector 14, and die spot 13 correspondingly decreases in size to zero at the detector. Therefore, by setting the reference value in the controller 20 to a value between AO and Al, and preferably (Al + A0)/2, the beam 12 is positioned on die edge as in the embodiment of Fig. 1.

Alternatively, if the object 22 in this embodiment is not reflective but acts like a screen so that a visible spot 13 is formed on the object itself rather than being reflected directly onto the detector 14, it is still possible to detect die edge 24. In diis case, by analogy with Fig. 2, the detector 14 measures the intensity of the spot as seen on the surface of the object, and generates an intensity signal whose amplitude decreases from Al to AO as the spot moves across the edge 24.

Figure 5 is a block diagram of a position-controlled light source 10 for use in the embodiments of Figs. 1-4.

The light source 30 proper is, in the present example, a laser diode but it may be any light source which provides a narrow light beam 12 whose angular deflection may be controlled. Light source conditioning circuits 32 may be used to provide conditioning of die light source, for example, modulation or intensity control, in certain applications. Also, in certain applications, beam conditioning means 34 may provide optical conditioning of the light beam 12 using, for example, focusing lenses, collimators, filters or spatial filters etc.

The beam deflector 36 provides deflection of the beam 12 along the Y direction under die control of d e spot control signal 18 from the controller 20. For a laser beam one can use galvoscanner deflectors, acousto-optic deflectors, rotating polygons, rotating prisms, resonant scanners, electro-optic deflectors, pencil scanners, etc. In the embodiments described above it was assumed tiiat the beam 12 was only deflected in a single Y direction in response to a single spot control signal 18, and in such a case deflection in only one dimension is needed. However, where (as will be described) d e beam is deflected in orthogonal X-Y directions two independently controllable deflectors or one deflector deflectable in X and Y directions is required.

A deflection signal generator 38 receives the spot control signal 18 from the controller 20 and provides a corresponding beam deflection signal to the beam deflector 36; this signal is an instruction to the deflector 36 to position the light beam at a specific point along the Y direction. Post-deflection conditioning means 40 may be provided for optical beam conditioning after the deflector 36 in some applications.

Finally, a position detector and output generator 42 provides the position signal 26, which is a signal representing the Y coordinate of the edge 24 of the object 22. This signal is derived from me instantaneous position of d e deflector 36 by any suitable optical, mechamcal or electronic technique for monitoring the movement of d e light beam deflector 36. Alternatively, the position signal 26 could be derived directly from the spot control signal 18, since the latter is also representative of the Y coordinate of the edge 24.

Figure 6 is a block diagram of a light detector 14 for use in die embodiments of Figs. 1-4.

Pre-detector means 44 may, depending on d e application, subject the light beam 12 to optical conditioning prior to detection. Typically die pre-detector means 44 may include interference filters, screens, integrating spheres, etc.

The detector 46 proper may be a photodiode or a photomultiplier tube. Other devices which can be used include charge-coupled devices, light-dependant resistors, etc. The particular detector used would depend on die required performance and the particular embodiment under consideration. A detector bias and control circuit 48 provides power and bias requirements for die detector.

Finally, detector signal conditioning and amplification circuitry 50 may provide various functions such as demodulation, dark current offset, current-to- voltage conversion, amplification etc., according to die application. The output of die circuit 50 is the intensity signal which is fed back to die controller 20 and represents the instantaneous intensity of the light spot 13 as measured at die detector 46.

Figure 7 is a block diagram of a controller 20 for use in the embodiments of Figs. 1-4.

An efficient type of controller 20 for use in die present invention is a known type of PID loop controller. At its input die controller has a comparator 52 which receives the intensity signal 16 as one input and d e reference value 54, for example (Al + A0)/2, as the second input. From these inputs it derives an error signal 56 which is a measure of the difference between the two input signals 16 and 54. The tiiree controller components 58, 60 and 62 act upon die error signal 56 to produce die spot control signal 18 for the signal generator 38 of the deflector 36 (Fig. 5).

The proportional component 58 provides correction of the beam position based on die magnitude of me error, the integral component 60 provides correction of the beam position based on die error value averaged over time, and die differential component 62 determines d e rate of approach to die reference value. The action of tiiese components 58, 60 and 62 are combined at 64 to provide die spot control signal 18 which, as has been described, tends to reduce d e error signal 56 by positioning the light beam 12 (or spot 13 in the case of die embodiment of Fig. 3) on die edge 24 of the object 22 as seen by the detector 14. As indicated in Fig. 7, the gain and time constant values can be set and adjusted to vary the response of the PID loop according to conditions.

While the PID controller is a very suitable controller for use in the present invention, otiier kinds of controller could be used. One possibility is to use a controller that simply causes the beam 12, and hence d e spot 13, to move in one direction when die error signal 56 is positive, and in the otiier direction when die error signal is negative. Provided the response of die feedback loop were sufficiently fast, this would cause d e beam to "hunt" at the edge of die object, moving back and forth across die edge widiout actually settling down to a fixed position. This would be acceptable in some applications provided die amplitude of the hunting were not too large. Another possibility is to use information about die position of die spot relative to the edge, if known, in addition to d e intensity error signal in the controller.

As stated, die above embodiments have assumed that the spot 13 is deflected in only a single (Y) direction and, accordingly, die apparatus is only capable of determining a single point on the edge of die object 22. As referred to above, a typical application for such one-dimensional edge detection is fluid boundary tracking.

However, where it is desired to determine the two-dimensional profile of an edge, it is necessary that the beam 12, and hence die spot 13, be deflectable in two directions at an angle to one anodier, preferably orthogonal X-Y directions. In tiiis case the beam deflector 36 (Fig. 5) will have two deflectors for independentiy deflecting die beam in die X and Y directions respectively, for example single-axis galvoscanners in the case of a laser beam. Figure 8 shows a controller 20 suitable for controlling such an arrangement.

The controller of Fig. 8 includes a PID controller as already described in relation to Fig. 7 to provide a spot control signal 18 controlling deflection of die beam 12, and hence die spot 13, in the Y direction. Simultaneously the beam 12 is deflected in die X direction under die control of a second spot control signal 18' derived from a timebase 66. In tiiis case both signals 18 and 18' are input to die signal generator 38 (Fig. 5) which generates corresponding beam deflection signals to the X and Y deflectors of die beam deflector 36. Thus in die Y direction die beam is deflected as described above under d e control of the PID controller, while in die X direction die beam is deflected under die control of the timebase 66.

This permits the beam 12 to track along die edge of a stationary object. However, if die scanning frequency in me X-direction is too high the beam will not be able to follow steep variations in die edge, and therefore the frequency of the timebase should be adjusted in relation to die response time of the PID feedback loop to allow the beam to follow the edge closely. Alternatively, this problem could be overcome by freezing the X direction spot control signal 18' when the magnitude of die error signal 56 exceeds some preset threshold, and only resuming beam movement in the X direction when die magnitude of die error signal falls below the threshold.

In this case the position detector and output generator 42 will output two position signals 26, corresponding to die instantaneous X and Y coordinates of the edge 24 of die object 22, these being derived from the instantaneous position of the two X and Y deflectors in die beam deflector 36. Alternatively, these two position signals could be derived directly from die spot control signals 18 and 18' .

As an alternative, Fig. 9, the controller 20 could include a PID controller (Fig. 7) providing a radial spot control signal 68 for radial deflection of die beam 12 relative to an origin, and an angular timebase 70 providing an angular spot control signal 72 defining die angular position of die spot relative to an axis extending from the origin. These radial and angular spot control signals 68 and 72 are tiien converted to X and Y direction spot control signals 18' and 18 by a polar-cartesian coordinate converter 74 for control of the X and Y direction deflectors in the beam deflector 36 (Fig. 5).

This will provide a radar-like sweep of die beam 12, and hence die spot 13, around die origin, the radial position of die beam at any instant being determined by die PID controller so that the beam follows die edge of die object. Again, the frequency of the timebase should be appropriately adjusted in relation to the response time of the PID feedback loop to allow die beam to follow the edge closely, or the angular spot control signal 72 can be frozen when the magnitude of die error signal 56 exceeds a preset threshold, as described above.

This embodiment is useful where one wishes to follow the edge on all sides of an object, and can be used to track botii outside and inside edges of a hollow object, such as a machine nut, Fig. 13, depending upon whedier the beam is deflected radially outward or inward when die controller 20 detects tiiat the spot is obscured from die detector 14 by die object. In both cases the origin is located at or near the centre of the hole in die nut as seen by the detector 14.

However, die edge of some shapes, such as d e "L" shape shown in Fig. 10, cannot be followed on all sides of die object using a single origin. For example, using die point 01 as an origin and assuming a clockwise angular scan is used starting at the lower left corner of the object, the edges S6, SI and S2 are followed in their entirety, but the edges S3 and S4 are missed and die beam jumps from die bottom of edge S2 directly to d e point on edge S5 radially outward relative to the origin so that only the portion D of the edge S5 is followed. However, die missing edges of die object can be tracked by choosing a second origin, for example 02, and performing a second scan. By correlating the two sets of results die outline of the entire object can be reconstructed.

Figure 11 is a block diagram of circuitry which uses the position signal(s) derived from the position detector and output generator 42 (Fig. 5). It will be recalled that either one or two such signals are generated, depending upon whedier the spot 13 is deflected in only the Y direction (Fig. 7) or in botii X and Y directions (Figs. 8 and 9).

The position signals can be used to drive a visual display device 76, for example a CRT, to provide an image for visual inspection. Examples of images which can be produced in this way are shown in Figures 12 and 13.

The images in Fig. 12 were produced using a controller as described witii reference to Fig. 8. The left hand side of Fig. 12 shows die image of a comb with one missing tootii in die middle and several at the right hand end, and me right hand side of Fig. 12 shows die profile of a "Yale" key. The image in Fig. 13 was produced using a controller as described with reference to Fig. 9, and shows botii inside and outside edges of a machine nut. Alternatively or in addition die signals can be used witii analog circuitry 78 for on-line control or inspection.

The position signals, if analog, can also be digitised in an analog to digital converter 80 to provide die edges of die object under examination as a series of coordinate values. These coordinates can then be analysed 82 or used in machine vision applications such as process control 84, parameter generation, product inspection, etc.

The size of the spot 13 should preferably be small in relation to the size of the object under examination and of die same order of magnitude as variations in the edge which it is desired to distinguish.

A further embodiment will now be described witii reference to Figure 14. In this embodiment an oscilloscope 86 is used to produce a spot 13 on its screen 88 which is viewed by die detector 14. The timebase of the oscilloscope is switched out, and die vertical (Y) deflection of die spot 13 is controlled by d e spot control signal 18 from the controller 20 which, in this case, is constructed as described in relation to Fig. 7. It will be seen that tiiis embodiment is essentially the same as that of Fig. 3, except that the oscilloscope screen 88 replaces die screen 28 and die deflecting circuits of the oscilloscope itself are used as die beam deflector. Thus, as die beam is deflected across die screen 88 internally of the oscilloscope tiiere is no substantial change in the size or intensity of die spot 13 on die screen itself. However, the proportion of the spot area, and correspondingly its intensity, as seen on the screen 88 by the detector 14 will change at that point within die range of deflection of die beam 12 where the line of sight 29 between the spot 13 and die detector becomes obscured by die edge of die object 22. Therefore, as the area of die spot 13, as seen by the detector 14, varies from its full area to zero the amplitude A of die intensity signal 16 will vary accordingly. This signal 16 is fed back to the controller for Y direction control of the oscilloscope beam in the manner previously described. This arrangement has been used to examine integrated circuit (IC) legs at a rate of one million legs per minute. The ICs were placed around die circumference of a large wheel or carousel 90 which was then rotated in front of the oscilloscope screen 88 at high speed. Thus, while the controller 20 determined die position of die spot 13 in die Y (vertical) direction, die rotation of the carousel 90 effectively provided an X direction scan of the spot 13 relative to the integrated circuits. The Y direction position signal 26 derived from the oscilloscope deflector circuits, or from the spot control signal 18, together with an X direction signal derived from the rotation of the carousel, provided a two dimensional profile of the IC legs around die circumference of the carousel.

This embodiment used a standard side-window photomultiplier tube - the EMI 978 IB - as the detector housed within a limiting aperture, and d e oscilloscope used was a standard Hitachi V212 20MHz model.

Of course, it will be understood that for a stationary object the X deflector of the oscilloscope could be operated under die control of the horizontal timebase of the oscilloscope, thereby giving two dimensional spot movement and allowing the spot 13 to track die edge of die object generally in the manner described above for the laser embodiments. Also, the radial/angular scanning technique described above can similarly be used witii an oscilloscope or other CRT based device.

Otiier embodiments of die invention are possible in which a spot of hght on a screen is generated by otiier tiian a light beam. For example, a spot of light can be generated on die screen of a display panel consisting of a high density matrix of LEDs, and die spot can be deflected by using the drive circuitry of the display panel. Such an arrangement can be substituted for the oscilloscope in Fig. 14.

As mentioned above, die use of filters and/or modulation can be used to mitigate the effect of ambient light. When using a laser as the light source it is possible to use a laser witii a narrow bandwidtii; e.g. a laser diode at 636nm. An interference filter with a central passband at die same wavelengtii can then be placed in front of the detector which then effectively only sees the light from the laser, and otiier ambient light is filtered out.

Alternatively, the laser output can be modulated. Laser diodes can be modulated at frequencies in excess of 20MHz. Signals of this frequency incident on the detector are unlikely to originate from light sources other than that used to produce die spot and such frequencies are also much higher than the bandwidth required for imaging. The signal to the detector then contains both normal incident light (ambient light) and d e laser light. Using AC demodulation die modulated laser signal can be extracted from the detector signal. Modulation and filter techniques can also be used with CRT-based devices such as die oscilloscope described above.

Another application of the use of filters and/or beam modulation is where the object is examined simultaneously with two beams of light, for example to provide stereo views of an object to obtain spatial/range information. An embodiment is shown in Figure 15.

In Fig. 15 there are two systems 10', 14', 20' and 10", 14", 20" as described witii reference to Fig. 2, but sharing a common translucent ground glass screen 28. The light sources 10', 10" are offset from one another so that when the beams are grazing the same point on the edge 24 of die object 22 the light spots 13', 13" are separated on die screen 28. Then the distance of die object 22 from the screen 28 can be readily determined by a simple geometrical calculation from the vertical (as seen in Fig. 15) separation of light sources 10, 10", the horizontal distance of the light sources from the screen and die vertical separation of the two spots 13' , 13" on the screen.

In this case it is important that each detector 14', 14" only responds to light from its own source 10', 10". This can be achieved by using a different coloured light for each beam 12', 12" and providing appropriate filters in front of each detector. Alternatively the light beams can be differently modulated, each detector responding to light only of the correct modulation for that detector. However, it is not necessary to use two systems to provide range information. In Figure 16 a single system is used and first a measurement of the position of die spot 13' is taken along the main beam path 12' . Next a mirror 92 is switched into the path of the beam 12' which deflects die beam onto a second mirror 94 and tiience along a new path 12" which approaches the object at an angle to the first path 12' . The position of the spot 13" resulting from the new light path is now measured and, as before, range information can be readily calculated. Clearly die ranging principles described in Figs. 15 and 16 can be applied to other of the embodiments.

All the above embodiments have assumed that it is the position of the edge of die object which is of interest. However, the principles described above can be used for measuring the height of a surface, where the surface under examination is not a surface of die object, and die object is used simply as a component of the system. Figure 17 shows a basic embodiment.

In this case die beam 12 is reflected off a surface 96 whose height is to be determined. A knife edge 98, corresponding in the previous embodiments to the object 22, is located at a fixed position relative to the light source 10 and translucent ground glass screen 28. Thus die angle of deflection of die light beam 12 will be adjusted by die feedback control mechanism previously described so tiiat it just grazes the top of die knife edge 98 and forms a spot 13 on the screen 28.

If the height of the surface is now changed, say by moving the surface upward to d e position 96' indicated in broken lines, die beam 12 reflected off d e surface 96 will travel along a new patii 12', again just grazing d e top of d e knife edge 98 and forming a spot 13' at a new position on the screen 28. The coordinates of the two spots 13, 13' are derived in the manner described previously, and die distance between them is directly related to the difference in height, delta h, between the two positions of the surface. If die surface 96 is not reflective and die image of the spot is formed on die surface 96 itself, then the screen 28 is omitted and d e detector 14 measures die intensity of the spot as seen on the surface 96. This technique may be used for surface profiling. Although the preceding embodiments have described die detection of a physical edge of an opaque object, the technique is not limited to such cases. The object may be any solid or fluid entity and may be opaque, translucent, transparent or reflective, or more than one of these in different regions. Provided the object presents a line demarcating regions of different transmissivity or reflectivity which is able to modify die amplitude of die intensity signal 16 as die beam or spot crosses that line, it is possible to detect tiiat line according to die above principles by setting the controller reference value to a value between Al and AO and preferably (Al + A0)/2. In this case, Al is the amplitude of die intensity signal 16 on one side of die demarcation line and AO die intensity on the otiier side of the demarcation line.

The relative locations of the various components in the embodiments described are not restricted to die arrangements shown. By using beam splitters and/or mirrors the components may be placed anywhere convenient. For example, using beam splitters, in the embodiment of Fig. 2 die position-controlled light source 10 and die detector 14 can be placed in the same housing, so that die detector 14 looks at the spot 13 on die screen 28 along d e same pati as the outgoing light beam 12.

The present invention has many useful applications. The radial scan technique (Fig. 9) can be used for grading or sizing or checking the shape of many mass-produced products. By comparing a product witii a copy stored in memory die technique could be used to detect flawed products in a production line. The linear scan technique (Figs. 7 and 8) could be used to detect missing teed on saw blades, missing pins on electrical connectors and broken or fused teetii on a comb, and indeed all departures from a standard profile.

Claims

1. A method of determining me location of at least one point on an object boundary, d e method comprising d e following steps not necessarily in the order stated:

(a) generating a spot of light deflectable along at least one direction,

(b) providing a control means for generating a first signal ("spot control signal") which controls the deflection of die spot along die one direction,

(c) providing a light detector to generate a second signal ("intensity signal") related to die intensity of the spot as measured at die detector,

(d) placing the object such that the boundary is located at a position where it is able to modify the intensity of the spot, as measured at die detector, at a point within the range of deflection of die spot, and

(e) feeding die intensity signal back to the control means, the control means being responsive to the intensity signal to adjust die spot control signal in an attempt to deflect die spot to a position wherein the intensity signal equals, or falls within a range including, a value corresponding to the modification of die spot intensity by die said boundary.

2. A metiiod as claimed in claim 1 , wherein the value of the spot control signal determines a particular position of the spot along die one direction.

3. A metiiod as claimed in claim 1 or 2, wherein the spot is also deflectable along a further direction under die control of a further spot control signal derived from a timebase.

4. A method as claimed in claim 3, wherein the further direction is normal to the one direction.

5. A method as claimed in claim 3, wherein the one direction is a radial direction relative to an origin and die spot is also deflectable angularly about the origin under the control of a further spot control signal derived from a timebase.

6. A mediod as claimed in any preceding claim, wherein the spot of light is formed by a light beam which falls directly onto die detector for detection of die intensity of the spot and die object is located along die path of the light beam to the light detector.

7. A mediod as claimed in any one of claims 1 to 5, wherein the spot is formed by a light beam which falls on a screen, wherein the detector detects d e intensity of the spot on the screen, and wherein the object is located between the screen and the detector.

8. A mediod as claimed in any one of claims 1 to 5, wherein the spot is formed by a light beam which falls on a screen, wherein the detector detects the intensity of the spot on the screen, and wherein the object is located along the path of the light beam to die screen.

9. An apparatus for performing the mediod claimed in claim 1, including means for generating a spot of light deflectable along at least one direction, a control means for generating a first signal ("spot control signal") which controls the deflection of die spot along die one direction, a light detector for generating a second signal ("intensity signal") related to die intensity of the spot as measured at die detector, and means for feeding die intensity signal back to the control means, the control means being responsive to the intensity signal to adjust the spot control signal in an attempt to deflect die spot to a position wherein die intensity signal equals, or falls within a range including, a settable reference value.