GB2076532A - Correcting Non-linearities in Optical Scanners - Google Patents

Abstract

Correction of recurring non- linearities in the scan time of an optical image scanner 11 includes a calibration stage in which an accurately located grating 14 having alternating optical properties at known intervals is scanned and the times at which the changes are detected are compared with the stored times representing the times at which detection should occur for the grating to give a time error for each of the times measured. The errors are stored in a store 27 at locations defined by the measured times. In scanning an object each measured scan time at which a change in optical properties is detected addresses the store and retrieves the error for that time, which is added to the measured scan time to give a linearity corrected value. The times are represented by clock pulse counts and the manipulation and storage is digital. <IMAGE>

Description

SPECIFICATION Correction for Recurring Non-linearities in Optical Image Scanners This invention relates to correction for recurring positional non-linearities in optical image scanners.

One example of an optical image scanner is the cathode ray tube flying spot scanner in which a spot of light is generated on the screen of a c.r.t.

and is swept repetitively across the screen in a line scan. Light from the spot is detected after engagement with an object and the detected level is indicative of some optical property of the object, such as its reflection, diffusion, transmission or absorption property. More particularly any change in detected level is indicative of a change in the optical property. The distance of the spot from the beginning of the scan is a function of the deflection voltage applied to the c.r.t. electron beam and this is varied as a predetermined function of time to give a linear scan of fixed duration, that is, the spot position is a function of the time elapsed since the beginning of the scan, hereinafter referred to as the scan time.

By utilising knowledge of the spot position throughout the scan the image scanner is then suited for image coding, that is, making quantitative measurements of the optical properties by sampling the detector signal upon the occurence of it crossing one or more predetermined threshold values in relation to the scan time of the occurrence. These samples may be recorded or stored and reproduced or further processed as a representation of the behavour of the image, produced by the spot, in the scan.

However in this and in other types of image scanners, such as television camera tubes or rotating mirror or prism scanners, non-linearities can occur in the relationship between the spot position (or its equivalent) and the scan time which recur from scan to scan but which are a function of the individual scanner. Consequently the scan time at which samples are recorded as happening can be in error and possibly by an amount which renders a scanner unsuitable for accurate quantitivative measurement.

Furthermore as such non-linearities are peculiar to individual pieces of apparatus a general compensation cannot readily be applied to the scanning drive particularly in the case of rotating mechanical scanners where inertia prevents significant charges of scanning speed within a scan period.

While it is impracticable to attempt to prevent the occurrence of non-linearities it is an object of the invention to provide a method of, and apparatus for, correcting for recurring nonlinearities.

According to one aspect of the present invention a method of correcting for recurring non-linearities in the scan time of optical image scanners comprises using the scanner to scan a calibration object having accurately positioned optical discontinuities, comparing the actual scan times at which said discontinuities are detected with their theoretical detection times to derive a plurality of scanning time error signals by which the scanner departs from said theoretical detection times and storing said time error signals, then in each scan of an object under observation detecting each crossing of predetermined threshold levels by the image and the scan time of each detection, connecting each scan time in accordance with a stored time error signal relating to that scan time and producing information representing each detection of a threshold level with the linearity corrected scan time of the detection.

Preferably the scan times are measured as the number of pulses, derived from a clock, counted since the start of the scan, and the corresponding time error signals represented as numbers of pulses recorded digitally as error words in a digital store.

According to another aspect of the present invention apparatus for correcting for recurring non-linearities in the scan time of optical image scanners comprises a photodetector operable in each scan of an object to produce a detector signal having a parameter related to light received from an object being scanned, thresholding means operable to provide a sample signal indicative of a detector signal crossing a predetermined threshold magnitude from one range of values to another, timing means responsive to the commencement of each scan to produce a scan time signal representative of the scan time continuously during the scan, gating means responsive to each sample signal to cause a scan time signal to be read from the timing means, calibration means including a calibration object arranged to be located for optical scanning by the scanner such that changes in optical properties of the calibration object occur at positions for which corresponding theoretical scan times are stored, subtraction means operable in a scan of the calibration object to subtract scan time signals, as gated by the thresholding means, from stored theoretical scan times to produce error signals indicative of the error by which each measured scan time departs from its theoretical value and an error signal store operable to store said error signals at locations specified by said measured scan time signals, and scan time signal correction means responsive to each sample signal produced in scanning an object under inspection to apply the associated scan time signal both to the error store to retrieve the error signal corresponding to the scan time and to addition means in which the error signal is added algebraically to produce a linearity-corrected value of the scan time signal.

The detector signal parameter upon which threshold crossing is detected may be the signal magnitude.

The timing means may comprise an oscillator operable to produce a train of clock pulses and counter responsive to the commencement of each scan to count clock pulses during the scan and produce a scan count signal representative of the scan time signal. The scan count signal may be produced as a binary word and the theoretical scan times may be stored as numbers of clock pulses in binary form in a digital store.

Preferably the calibration object is a ruled grating which is divided into alternate regions having optical properties to cause a detector signal to be either above or below a calibration threshold level as applied to the thresholding means.

An embodiment of the invention will now be described by way of example with the reference to the accompanying drawings, in which: Figure 1 (a) shows a graph of detector magnitude against time for a single scan of an image scanner and its relationship with three threshold levels of detector signal magnitude, Figure 1 (b) shows a graph of the detectgr signal sampled at said threshold levels and plotted against scan time as represented by clock pulses, Figure 2(a) is a sectional elevation through a calibration grating used in a calibration step, Figure 2(b) is a-graph showing the relationship between the detector signal produced by part of a scan of the calibration grating and the scan time as represented by a number of clock pulses, Figure 3 is a graph showing the variation of scan time error E as a function of scan time or spot displacement.

Figure 4 is a schematic representation of an image scanner and the non-linearity correction circuitry of the present invention as used in a calibration mode, and Figure 5 is a schematic diagram, similar to Figure 4 but showing the correction circuitry as used in a correction mode.

Referring to Figure 1 the graph waveform shows a detector signal as produced by a photodetector during one scan of an object by a light spot generated by a flying spot scanner, the magnitude of the signal throughout the scan being a function of the light transmitted by an object. Such scanners and their detector signals are well known. The detector signal is shown plotted against scan time for a single scan. Also shown are three predetermined voltage levels V to V, defining four ranges of detector signal magnitude;-- 0 to V1, V1 to V2, V2 to V3 and greater than V3. The times at which the detector signal crosses the threshold levels and enters a different range of values, t1, t2, etc. are also shown.

It will be appreciated that the detector signal can be digitised by sampling the signal at crossings of the threshold and recording the times tt, t2 etc. of such crossings in terms of a number n1, n2, etc, of clock pulses, requiring a relatively small number of samples to be recorded for further processing or future use. From a small number of samples recorded a simplified form of the graph may be realised as shown in Figure 1 (b) and illustrating the effects of digitisation. The coarseness of the digitisation steps along the vertical axis can be varied by varying the values of the threshold voltages and the numbers of thresholds employed.

The present invention involves the use of a calibration step employing a calibration grafting, such as shown in sectional elevation in Figure 2(a), which has regular and well defined alterna transparent and opaque regions. The detector signal from a scan of the calibration grating is shown in Figure 2(b) plotted against number of scan pulses, the peaks due to transmission clearly being of greater magnitude than a calibration threshold voltage Vc. With the calibration grating positioned accurately with respect to the scanner then each or the signal discontinuities should occur after a known constant number of clock pulses and any differenences between the count at the theoretical detection points and at the actual detection point represent errors due to scan non-linearity.

Figure 3 shows in graphical form a nonlinearity scan time error E as a function of scan position (or scan time) for a typical scanner and the error may be considered in terms of a number of clock pulses nE, by which the scan time of a discontinuing in error and the position of the error np.

Referring to Figure 4, an image scanner 10 including non-linearity correction is shown schematically in a calibration mode of operation.

A cathode ray tube 11 is caused to generate a scanning beam of light 12 by a sweep signal generator 1 3. A calibration grating 14 is supported on a transparent table 1 5 beneath which is a light integrating enclosure 1 6 for directing light transmitted by the grating onto a photodetector 1 7. A signal level threshold detector 18 has one input 19 connected to the photodetector to receive the detector signal and another input 20 connected to a calibration threshold voltage Vc by way of a potentiometer 21. A clock pulse generator 22 is connected to supply clock pulses to a counter 23. The counter 23 is arranged to count in each scan the number of pulses generated from the beginning of the scan and provide continuously in parallel binary form the accumulating total to gating means 24.

The gating means 24 is arranged to be triggered by the threshold means 1 8 each time the detector signal crosses the calibration threshold value and pass the current scan count. The gating means 24 is connected to supply the scan count both to subtraction means 25 and to an address input 26 of an error store 27. A pre-programmed store 28 contains words representing theoretical scan counts for the known and accurately positioned discontinuities in the calibration grating and these are fed in paraliel binary form to the subtraction means 25 in synchronism with the scan for comparison with the actual scan counts produced.The difference for each count, representing the count error nE is applied to an input 29 of the error store 27 and stored at a location defined by the scan count addressing the address input 26.

To recapitulate on the calibration operation, in the calibration mode the calibration grating 14 is located on the table 23 accurately such that the discontinuities occur at locations for which the theoretical stored scan times are appropriate.

The grating is scanned by the scanner and during the scan each detection of a discontinuity in the grating pattern causes a scan count to be produced from gate 24 and applied to subtraction means 25 with the theoretical scan count. The difference between them, that is, the scan time error, is stored in the error store 27 at a location unique to the measured scan count addressing the input 26. At the end of the scan the error store will be filled with scan error counts nE for all scan position count np.

Referring now to Figure 5 which shows the scanner and processing circuit in a correction mode of operation when scanning an object 14' placed on the table 1 5. The clock pulses generator, counter and gating means 22, 23 and 24 are still employed but now there are three threshold detectors 30, 31 and 32 each connected to receive signals from the photodetector 17 and compare them respectively with the threshold voltages Va, V2 and V3 referred in connection with Figure 1(a). It will be understood that one of the threshold detectors may be used in the calibration mode as the threshold detector 14 with a suitable threshold voltage.

Each of the threshold detectors 30-32 has an output terminal 33-35 respectively connected to a coding arrangement 36 which produces a signal indicative of which thresholds are crossed, that is, which range of values the detector signals is in, for further processing or storage.

The signal passed by the gating means 24 is applied to the address terminal 26 of the error store 27 and also to one input 37 of an addition circuit 38. A second input 39 of the addition circuit 38 is connected to an output terminal 40 of the error store. The addition circuit 38 has an output terminal 41.

Thus in a measuring operation scan, the counter 23 counts clock pulses from the beginning of the scan and the count at any time, the scan count, represents the scan time. The threshold detectors 30-32 receive the photodetector signal and each time the signal crosses a threshold from one range of values to another the coding arrangement 36 produces a signal in binary form indicative of the range of values, and the threshold detection operates gating means 24 to give the numbered scan count for that threshold crossing. The scan count is applied to the first input 37 of the addition circuit 38 and also to the address input of error store 27 from which the count error signal corresponding to that scan count is retrieved and applied to the second input 39 of the addition circuit.The addition circuit produces a corrected linearity scan count which is coupled with the coded signal from coding arrangement 36 to identify the sample point of the detector signal.

The sampled detector signals from the coding arrangement and the scan time signals both being in digital form are suitable for storage, further processing or reproduction in graphical form. As with other forms of digitisation of continuous events by sampling the accuracy of reproduction depends on the sampling interval, in this instance in the number and separation of threshold levels, and on the number of scan times for which correction is available. This latter feature depends on there being a sufficient number of error words generated at sufficiently small time intervals (a function of grating resolution and clock frequency) and on storage locations for each word generated.

The error word store must contain words of sufficient length to cater for the worst case error expected in the apparatus and a sufficient number of words to cater for the worst case rate of change of error amplitude.

The actual values chosen for the various parameter depend on the type of scanner, the optical properties being detected and on the variation of the optical properties.

In the above example the detector signal parameter chosen for comparison with threshold values is the signal magnitude. It will be appreciated that the detector signal may be produced having some other parameter, such as frequency, related to the amount or intensity of light and the threshold values will be changed accordingly into a suitable form, such as frequencies.

Similarly the light is described by way of example as being transmitted by the calibration grating and the object under inspection and all light collected. It will be understood by those experienced in the field of optical inspection scanners that light (whether within or outside of the visible spectrum) emanating from the grating and object by any optical property may be detected totally or in part by location of the photodetection means and light collection means in a suitable position.

Claims (9)

Claims

1. A method of correcting for recurring non linearities in the scan time of optical image scanners comprising using the scanner to scan a calibration object having accurately positioned optical discontinuities, comparing the actual scan times at which said discontinuities are detected with their theoretical detection times to derive a plurality of scanning time error signals by which the scanner departs from said theorectical detection times and storing said time error signals, then in each scan of an object under observation detecting each crossing of predetermined threshold levels by the image and the scan time of each detection, correcting each scan time in accordance with a stored time error signal relating to that scan time and producing information representing each detection of a threshold level with the linearity corrected scan time of the detection.

2. A method as claimed in claim 1 in which the scan times are measured as pulse counts of pulses derived from a clock and counted from the start of the scan and the corresponding time error signals are represented as numbers of clock pulses recorded digitally as error words in a digital store.

3. A method of correcting for recurring non linearities in the scan time of optical image scanners substantially as herein described with reference to the accompanying drawings.

4. Apparatus for correcting for recurring non linearities in the scan time of optical image scanners comprising a photodetector operable in each scan of an object to produce a detector signal having a parameter related to light received from an object being scanned, thresholding means operable to provide a sample signal indicative of a detector signal crossing a predetermined threshold magnitude from one range of values to another, timing means responsive to the commencement of each scan to produce a scan time signal representative of the scan time continuously during the scan, gating means responsive to each sample signal to cause a scan time signal to be read from the timing means, calibration means including a calibration object arranged to be located for optical scanning by the scanner such that changes in optical properties of the calibration object occur at positions for which corresponding theoretical scan times are stored, subtraction means operable in a scan of the calibration object to subtract scan time signals, as gated by the thresholding means, from stored theoretical scan times to produce error signals indicative of the error by which each measured scan time departs from its theoretical value and an error signal store operable to store said error signals at locations specified by said measured scan time signals, and scan time signal correction means responsive to each sample signal produced in scanning an object under inspection to apply the associated scan time signal both to the error store to retries the error signal corresponding to the scan time and to addition means in which the error signal is added algebraically to produce a linearitycorrected value of the scan time signal.

5. Apparatus as claimed in claim 4 in which the detector signal parameter upon which the threshold crossing is detected is the signal magnitude.

6. Apparatus as claimed in claim 4 or claim 5 in which the timing means comprises an oscillator operable to produce a train of clock pulses and a counter responsive to the commencement of each scan to count clock pulses during the scan and produce a scan count signal representative of the scan time signal.

7. Apparatus as claimed in claim 6 in which the scan count signal is produced as a binary word and the theoretical scan times are stored as numbers of clock pulses in binary form in a digital store.

8. Apparatus as claimed in any one of the preceding claims in which the calibration object is a ruled grating which is divided into alternate regions having optical properties to cause a detector signal to be either above or below a calibration threshold level as applied to the thresholding means.

9. Apparatus for correcting for recurring nonlinearities in the scan times of optical image scanners substantially as herein described with reference to, and as shown by, Figures 4 and 5 of the accompanying drawings.