This invention relates to the measurement of the position of an object boundary enabling, for example, the exact measurement of the size of the object.

The measurement of position of an edge of optically formed image is conventionally carried out with the aid of an optical system equipped with a co-ordinate measuring stage and cross-wire or some form of reference marker which is aligned to coincide with the edge under investigation and the position of the edge can thus be recorded.

This method has a number of disadvantages. It is slow, dependent on subjective judgment and fatigue of operator and therefore may produce inconsistent results.

With the growing requirement for automated non- contact measuring systems there is need for rapid, precise and consistent measurement of edge position on optical images under dynamic conditions. Such measurements find use, for instance, in high precision metrology applications, where it is necessary to obtain dimensional information on engineering components.

It is an object of the present invention to provide an improved method and apparatus for measuring the position of an object boundary.

According to a first aspect of the present invention there is provided a method of measuring the position of an object boundary in which the object boundary is moved relative to a detector and an indication of the position of the object boundary relative to a datum point fixed in relation to the detector is produced continuously for recording in a recording means, and the value of the indication is recorded by the recording means on detection by the detector of the object boundary, the detection signal and the indication undergoing equal delays in their transmission to the recording means.

SUBSTITUTE SHEET

According to a second aspect of the present invention there is provided apparatus for measuring the position of the boundary of an object including a support for the object, optical means for forming an image of at least 'part of the object boundary on a screen which -may be virtual, a first optical fibre of relatively small diameter coupling first area of the screen to a first light intensity responsive detector, a second optical fibre of relatively large diameter coupling from a second area of the screen substantially concentric with the first area to a second light intensity responsive detector, the support and the screen being relatively movable, comparator means for comparing the signals from the first and second detectors and producing an output signal when the two detector signals have a predetermined relationship, transducer means coupled to the support producing an electrical signal representing the position of the support relative to the screen, and recording means connected to the transducer means for recording the position of the support at the instant the comparator means produces the output signal.

According to a third aspect of the present invention there is provided a method of measuring the position of an edge comprising the steps of: placing the edge on a movable support, illuminating the edge by means of a light source, forming an image of the illuminated edge on a screen, conveying the light from a particular fixed area on the screen to a light intensity responsive detector, deriving a first electrical signal representing the light intensity at the area from the detector, deriving a second electrical signal in response to light from the light source including light other than from the area, passing the first and second electrical signals through first and second channels respectively to a comparator, the comparator producing an output signal when the first and second electrical signals have a predetermined relationship, moving the movable support so that the image of the edge on the screen moves across the particular area, producing a third electrical signal representing the instantaneous position of the support as it moves, conveying the third electrical signal along a third

channel to a recording means, and recording a representation of the position of the support in the recording means at the instant the output signal from the comparator is applied to the recording means, the time delay for the transmission of an electrical signal representing the edge from the detector to the recording means being substantially equal to the time delay for the transmission of the third electrical signal from the support to the recording means.

According to a fourth aspect of the present invention there is provided apparatus for measuring the position of an edge including a movable support on which the edge is mounted, a light source for illuminating the edge, means for forming an image of the edge on a screen, optical fibre means facing a particular fixed area of the screen for coupling light it receives from that area to.a light intensity responsive detector for producing a first electrical signal representing the light received, a second means responsive to .light from the source including light not from the particular area for producing a second electrical signal, a comparator for producing an output signal when the two signals applied to it have a predetermined relationship, first and second channels respectively for transmitting the first and second electrical signals to the comparator, transducer means coupled to the support for producing a third electrical signal representing the position of the support, recording means connected to receive the third electrical signal via a third channel and responsive to the output signal from the comparator to record a representation of the position of the support at that instant, the time delays imposed on the representation of the edge conveyed by the first electrical signal from the detector through the comparator to the recording means and on the representation of the position of the support conveyed by the third electrical signal from the transducing means to the recording means being arranged to be equal.

According to a fifth aspect of the present invention there is provided apparatus for measuring the position of an edge including a movable support on which the edge is mounted, a light source for illuminating the edge, means for

forming an image of the edge on a screen, light splitting means for directing light from a first relatively small circular area of the screen to a first light responsive detector and from a relatively large circular area of the screen substantially concentric with the small circular area to a second light responsive detector, the first and second detectors being for producing first and second electrical signals respectively representing the light received, a comparator for producing an output signal when the two signals applied to it have a predetermined relationship, first and second channels respectively for transmitting the first and second electrical signals to the comparator, transducer means coupled to the support for producing a third electrical signal representing the position of the support, recording means connected to receive the third electrical signal via a third channel and responsive to the output signal from the comparator to record a representation of the position of the support at that instant, the time delays imposed on the representation of the edge conveyed by the first electrical signal from the detector through the comparator to the recording means and on the representation of the position of the support conveyed by the third electrical signal from the transducing means to the recording means being arranged to be equal.

Preferably the object is mounted on a movable support so that the area of interest always lies close the optical axis of the lens used to form the image and the photodetectors with their associated optical fibres can be rigidly mounted. Although reference is made to a screen on which an image of the object boundary is focussed, it is not necessary for the screen to be a physical entity, it may merely be the focal plane of the lens forming the image.

The invention can measure the position of the edge of an optical image precisely and rapidly, avoiding the need for manual alignment of a workpiece with a cross-wire. The measurement can be done dynamically without stopping the edge under the detector.

The light used may be visible, ultra-violet or infra-red, the detectors and optical components being chosen suitably. The use of visible light would enable an operator to observe the operation.

The transducers providing electrical signals repre¬ senting the position of the movable support on which the object is carried may each consist of a diffraction grating scale, of say 10 micron pitch, with quadrature cursor gratings producing sine and cosine output signals. The complete cycles of the output signals may be counted digitally and one or more additional digits calculated from the sine and cosine signal values. The recording means may be digital registers, one for each axis of movement of the support, for recording the complete cycles, with analogue to digital converters for the sine and cosine values staticised at the sampling instant from which the interpolated digits are calculated and inserted at the ends of the registers.

The operation of the apparatus may be controlled by a microprocessor or small computer programmed to move the object on its support so that the positions of parts of the boundary of the object can be measured and the size of the object calculated from the measurements. For example, if the object is circular the positions of three or more points on its periphery would enable its diameter to be calculated, using least square fitting if more than three points are used.

The use of a comparator or differential amplifier to detect the transition from light to dark or vice versa at the edge means that the edge detection is substantially without hysteresis, so that the edge position can be measured accurately in either direction.

In a typical application the object under investi¬ gation is placed on a co-ordinate stage at the focal plane of an optical system and the stage is moved, e.g. under manual or computer control, so that the edge of the image crosses the sensor aperture, the area sensed by the light detector.

At the instant the edge passes the sensor aperture a signal is sent from the detector to the co-ordinate position register unit and the position of the stage is recorded at that instant.

The recorded information together with the polarity of transition can be transmitted to a computer for further processing and data logging.

In practical application it is most desirable to b ^' e able to read the edge position "on-the-fly" at different traversing speeds.

In order to ensure that the co-ordinate information is registered correctly the signal propagation delay through the edge sensor circuitry must be equal to that through the co-ordinate circuitry.

In other words it is necessary to ensure that the trigger pulse arrives at the co-ordinate value register without delay or if this is impossible, as it is in this case mainly due to exceedingly low signal level, it is sufficient to ensure that the propagation of the positional information arriving at the co-ordinate value register is delayed by the same amount so that both signals are coincident at the co-ordinate value register.

In principle it should be sufficient to increase the bandwidth of the trigger circuitry until the delay becomes insignificant. In practice, however, the bandwidth cannot be made arbitrarily wide. The amount of light energy available in these applications is usually very small and the generated signal requires substantial amplification. Even the most advanced analogue amplifiers are too 'noisy' for this appli¬ cation. Excessive noise produces scatter and essentially reduces the accuracy of measurement. To achieve the required accuracy it is necessary and sufficient to reduce the scatter to a level below the limit of resolution of the measuring system.

In these circumstances, bandwidth reduction is the only way to control the noise. This, however, has the undesirable effect of introducing a group delay which makes the edge position measurements speed dependent. This speed dependence makes accurate measurement impossible except in

applications where the speed can be controlled during the transition over an edge. Even in motor driven applications it is almost impossible to control the speed at all times.

For instance, when the edge is encountered at an oblique angle, velocity components in any two orthogonal directions will depend on the angle. Therefore in order to control the velocity of the transition it is necessary to know, in advance, not only the co-ordinate position at which the edge will be encountered but also its orientation.

In most practical situations this is not possible since such information is not known in advance.

In examples of the present invention this problem is overcome by providing matched filters in the signal paths so that the group delays are equal.

The examples of the invention are designed to reduce noise to a minimum, since by keeping noise down it is possible to increase bandwidth to a level which allows the edge to have speeds of several metres per second at the image plane, without adversely affecting accuracy.

In order that the invention may be fully understood and readily carried into effect examples of it will now be described with reference to the accompanying drawings, of which:-

FIGURE 1 is a diagram of the physical layout of the non-electrical part of an example of the invention;

FIGURE 2 is a block circuit diagram of the electrical part of an example of the invention;

FIGURE 3 shows one alternative optical configuration;

FIGURE 4 shows a second alternative optical configuration;

FIGURE 5 is a block circuit diagram of the electrical part of another example of the invention;

FIGURE 6 is a flow diagram of an error detecting routine usable with the circuit of Figure 5; and

FIGURES 7(a), 7(b), 7(c) and 7(d) are diagrams explaining the basis of the routine shown in Figure 6.

A possible physical layout of the apparatus is shown in Figure 1. An object 106 to be measured is carried on a movable support 107, such as a microscope stage and illumi¬ nated by light from a lamp 102 directed by a collimating lens 103. The brightness of the light from the lamp 102 is monitored through a light guide 101, which may be an optical fibre. The brightness of the lamp 102 is controlled by a potentiometer 5.

The position of the support 107 is measured by X and Y transducers 104 and 105 producing its displacement in two orthogonal directions. Each of the transducers may comprise a scale diffraction grating with two cursor gratings disposed in quadrature so that light transmission through the scale grating and then through the two cursor gratings separately follows sine and cosine curve respectively. The pitch of the gratings may be, for example, 10 microns.

An image 111 of the object 106 is focussed by optical system 108 on a screen 109. A light guide 112, which may be an optical fibre of small diameters, e.g. 0.04 mm, has its end 110 over a fixed area of the screen 109, and forms the aperture for sensing the edge of the image 111 which represents the boundary of the object 106 which is of interest. The screen 109 need not be a physical entity but may merely be the focal plane of the optical system 108.

The circuit diagram shown in Figure 2 will now be described, with occasional references to items shown in Figure 1. An edge moving across the edge sensor aperture 110 positioned in the image plane 109 of an optical system 108 produces a change in the light intensity which is transmitted along the light guide 112, 201 to a photosensitive detector 202. The detector generates an electric current proportional to light intensity falling upon it. The current is fed into a low noise amplifier 203 which converts it into a voltage level with sufficient amplification to operate a voltage comparator 206. The amplification is made large enough to exceed the input sensitivity of the comparator. Attenuator 204 is used to adjust the signal gain to an appropriate level to allow for changes in light intensity when different objective lenses are used as the optical system 108. A low

pass filter 205 limits the total electronic noise associated with the amplification path 201, 202, 203, 204, 205 including the comparator 206 below the desired limit of resolution of the measuring system.

A similar channel is provided in a reference path formed by optical fibre 211, detector 212, amplifier 213, attenuator 214 and low pass filter 215. This channel monitors changes in the light intensity of the light source 102. Any variation in the light intensity, caused by power variation, ageing or thermal effect, is automatically compensated, ensuring that such changes in the intensity have no influence on measuring accuracy. The bandwidth of the low pass filter 215 in the monitoring channel is adjusted to reduce the electronic noise generated in the signal path 211, 212, 213, 214, 215 to an acceptable level, i.e. below the desired limit of resolution of the measuring system. At the same time the bandwidth is made wide enough to allow slow variations in the light intensity, due to for example, source ageing, power supply instability, alternating current power supply ripple, to be passed to the comparator 206 without attenuation.

Both reference and signal currents emerging from their respective low pass filters 205, 215 are also routed through high impedance buffers 207 and 216 to a difference amplifier 209. A potentiometer 208 adjusts the proportion of the output of the buffer 207 which is applied to the amplifier 209 to compensate differences in gain between the signal and reference paths. In normal operation when the optical system has been configured to give a desired image magnification the attenuator 5 is adjusted until the brightness of the lamp is such that a null meter 210 shows no deflection. Any deviation from zero reading on the nullmeter during normal operation indicates incorrect setting. The potentiometer 208 may be used to adjust the relative edge trigger level in relation to reference voltage. This feature enables adjustment of the edge as represented by an electronic voltage level to coincide with the real edge lying exactly across the middle of the sensing aperture 110 in the image plane 109.

The output from the comparator 206 is used to latch the co-ordinate value information from the transducers 104, 105 in the co-ordinate register 217. Each co-ordinate value signal is conveyed to its register via a path including an amplifier 222 and a low pass filter 224 matched to the filter 205. There may be more than one position co-ordinate register depending on the number of measurement axes installed on the co-ordinate stage 107. The output signal from the comparator 206 may also be used to set a state register 218 which can be used in turn to inform the outside environment such as computer or data logger that a co¬ ordinate information corresponding to an edge transition is ready and available in the co-ordinate register and the sense of the transition (light to dark or dark to light) . After the co-ordinate information has been read an acknowledge signal 220 can be sent to clear the service request latch and make it ready for next edge transition.

Alternative ways of providing the reference signal are shown in Figures 3, 4 and 5.

A split beam sensor is shown in Figure 3. The arrangement has the distinct advantage over a concentric fibre aperture in that the symmetry of apertures can readily be obtained since only single fibres are involved.

Light from an objective lens travelling along the optical axis 301 is partially reflected and deviated by the first beam splitter 308 towards the image plane 304 in which the small fibre is positioned. The light captured by this fibre is guided to photodetector 501 in Figure 5.

The remainder of the light travelling along the optical axis is again partially deflected by the second beam splitter 307 and directed towards the second thicker fibre 303 positioned in the image plane 305. A certain degree of blurring may be advantageous. This light is guided by the fibre to the photocell 507 in Figure 5. The remaining light emerging from the second beam splitter 307 may be passed on to other uses such as TV camera an eye piece for direct viewing in the image plane 306.

The. fractional reflection of each beam splitter can be chosen to suit the requirement of the optical system in

use .

A concentric fibre configuration is shown in Figure 4. It consists of two light guides 403 and 404 the inner and the outer fibres respectively. The inner fibre can be a single strand of plastic type. The outer is normally constructed from a bundle of fine glass fibres and is bonded together with a suitable adhesive. The face 405 of the fibres is polished. The polished fibre face is normally placed in the image plane 401.

The arrangements of Figures 3 and 4 could be used with the circuit shown in Figure 5. This is essentially a differential arrangement which has the following advantages:

1. It can operate with transmitted light or reflected light from the object.

2. The accuracy is not impaired by ambient light changes.

3. It can operate with low edge contrast.

The smaller central fibre 302, 403, 501 guides the light to photodetector 502. The current produced by the detector 502 is amplified by a wide band amplifier 503 and the signal is fed into a variable gain amplifier 504. There is a similar signal path for the larger reference fibre 506 with a detector 507, an amplifier 508 and a variable gain amplifier 509 The variable gain amplifiers 504 and 509 are used to adjust the gains of the two channels so that both inputs to a difference amplifier 511 used as a comparator are the same for a given light level. The inputs to the amplifier 511 are through low pass filters 505 and 510 designed to limit the bandwidth of the two channels in order to reduce the electronic noise to a level below the desired limit of resolution of the system.

A variable gain amplifier 512 is used to adjust the difference signal from the amplifier 511 to a suitable level for comparators 513, 514 and 515, which compare the difference signal with reference levels A, C and B respectively. The difference signal level is monitored on meter 524 to which it is applied via a peak detector and hold amplifier 523.

The difference signal, shown in Figure 7(d) requires three levels of discrimination levels A and B to detect the positive and negative-going peaks and level C to detect the zero crossing. This is discrimination is achieved by means of the three voltage comparators 513, 514 and 515 and their corresponding reference levels 533, 534 and 535. The output from the comparators are routed to three state registers 516, 517 and 518 which interface to an external processing device, such as a computer, for validation of the difference signal, for example by the routine shown in the flow chart of Figure 6.

The output from "C" state register 517 also is used as the trigger signal for a co-ordinate value register 522 or plurality of registers. This ensures that the co-ordinate value register will only be triggered after .he difference signal sequence has been validated by the flow chart of Figure 6. Note that the validation process used in the flow chart produces a "valid" indication immediately on receipt of the output from the "C" comparator 514.

The positional information from one of the stage transducers 104, 105 enters at 519 and passed through an amplifier 520 and a low pass filter 521 to the co-ordinate value register 522. The information from the other transducer is handled in the same way. When the values have been recorded in the registers they are transferred to the computer or data logger when all those state registers 516, 517 and 518 have been set.

The three low pass filters 505, 510 and 521 are matched so that the total group delay along each of the three channels, the signal channel, the reference channel and the positional information channel, is the same.

The flow chart shown in Figure 6 looks for C following A without B or C following B without A and when C occurs the co-ordinate values are stored in the register (s). If the signals appear in any order other than ACB or BCA an error is indicated.

In Figure 7(a) a light to dark transition is shown, which is aligned with the centre of concentric signal and reference apertures 707 and 706. The detector outputs

associated with the two apertures as the transition moves over them are shown superimposed in Figure 7(c) and the resulting difference signal is shown in Figure 7(d).

Instead of programmable computer, the flow chart shown in Figure 6 could be performed by a hard-wired special purpose logical unit. The terms computer and computing means used herein are intended to include such a unit.