JPS6336450B2 - - Google Patents

Description

[Detailed Description of the Invention] The present invention provides MTF (Modulation Transfer Function) of optical elements such as enlarger lenses and microlenses.
This invention relates to a signal processing method in an MTF measuring device that measures a transfer function.

Methods for measuring the MTF of a light transmission element such as a lens include the optical Fourier transform method that scans a grating chart, the electrical Fourier transform method that scans an edge or slit and electrically analyzes the electrical signal by Fourier analysis, and the interference method. Autocorrelation methods that utilize

However, this commercially available measuring instrument uses a grid chart,
Mechanical scanning of edges, slits, etc. is required, and a highly accurate scanning mechanism is required, whether it is rotational or translational. Therefore, this scanning mechanism tends to be expensive and relatively large, so the height of each image and the direction angle of the chart (so-called
Azimuth), etc., it is difficult to install the appropriate number of units. For this reason, it is necessary to set a set of scanning mechanisms, light sources, light receiving parts, etc. at the required positions and directional angles to perform measurements, so it is necessary to measure the optical axis and two off-axis points.
You have to change the settings at least five times just to measure MTF. Furthermore, since lenses generally have a manufacturing error called eccentricity, it is necessary to measure MTF in at least four diagonal directions for the same image height, so this alone requires 17 settings. Furthermore, if a measurement of an out-of-focus state called Defocus is added, the number of measurements required is equal to the number of times described above multiplied by the number of Defocuss. For these reasons, despite the importance of MTF measurement in determining lens performance, the spread of MTF at the inspection level has been slow.

Next, an example of a conventional MTF measuring device will be described. This device uses a copying lens, microlens,
It measures the MTF of optical elements such as lenses and fibers that are used with a finite object distance, such as enlarger lenses, and is shown in Figure 1. In the figure, reference numeral 1 denotes a light source section, and the main components are a light source and a slit, but a condenser lens, a diffuser plate, a wavelength selection filter, etc. can be attached as necessary. The slit of this light source section is configured so that it can be rotated to the desired direction and angle for measurement, and the entire light source section 1 can be slid on a sub-bench 2 and installed at a required object height.
Reference numeral 3 denotes a test lens holder, which consists of a device for attaching the test lens 4 and a device that can rotate this by any angle. Reference numeral 5 denotes a light receiving section, the main components of which are a scanning chart and a light receiver, but in addition to the chart's scanning mechanism and spatial frequency conversion mechanism, a relay lens, a finder, etc. are added as necessary. This light receiving section 5 can be rotated so that it can be installed at a direction angle corresponding to the slit direction of the light source section 1, and the entire light receiving section 5 slides on another sub-bench 6.
The image height can be set to correspond to the height of the installed object. The sub benches 2 and 6 and the lens holder 3 slide on the main bench 7 so that the necessary object distance and image distance can be set.

Although this type of MTF measuring device has excellent versatility, such as being able to measure the MTF under various setting conditions for a wide variety of lenses in small quantities, as well as measuring the MTF of optical elements other than lenses, it is not suitable for large quantities of lenses. Currently, it is almost useless as a so-called inspection machine that needs to efficiently determine pass/fail. In other words, in this device, the light source section 1 and the light receiving section 5 are relatively large, and the light receiving section 5 in particular requires a highly accurate mechanism and tends to be expensive, so only the number corresponding to the required object height (image height) is required. Difficult to install. Therefore, even if the lenses to be tested are of the same model, it is necessary to move the light source section 1 and the light receiving section 5 in accordance with the required object height each time the MTF measurement of an individual lens is performed. Furthermore, in areas other than the axis, radial direction (so-called radial) and tangential direction (so-called
The light source section 1 and the light receiving section 5 must be rotated and installed for each object height (image height) according to the directional angles in two directions (tangential).

Therefore, MTF, which uses solid-state scanning devices such as charge-coupled devices (CCDs) or photodiode arrays, is superior in operability and speed.
A measuring device is being considered. In this device, a chart image projected by the device under test is received by a solid-state scanning device, and the output signal is used to determine the reference maximum value A (corresponding to the transparent and opaque portions of the chart with a spatial frequency close to zero). Minimum value B, maximum value a corresponding to the transparent and opaque parts of the required spatial frequency of the chart
and the minimum value b, and from these, for example, (a-
b)/(A-B)×100 to calculate the value of the optical element to be tested.
Calculate MTF.

However, in this MTF measurement device, when the spatial frequency of the chart becomes high to a certain extent, even if the chart has a rectangular wave pattern of transparent and opaque parts, the time-series signal from the solid-state scanning element becomes sinusoidal, and its maximum value It becomes difficult to clearly distinguish between a and the minimum value b, making it difficult to calculate MTF. Furthermore, since solid-state scanning devices have a very large number of photo elements, local defects in the semiconductor may cause some photo elements to have abnormally higher sensitivity than other photo elements, or conversely to have too low sensitivity. There is. If an attempt is made to eliminate such abnormally sensitive photo elements, the yield of the product will deteriorate, so defects in a small number of photo elements are generally tolerated. Therefore, due to the influence of abnormal output due to abnormally sensitive photo elements,
MTF calculation becomes inaccurate.

Therefore, in the MTF measuring device described above, the output signal of the solid-state scanning element is greater than a specific threshold level Vth and includes the output signals of a plurality of continuous photo elements on the white side corresponding to the transparent part of the required spatial frequency of the chart. Let the maximum value in L ₃ be the true maximum value a, and the output signal of the solid-state scanning element is smaller than the specific threshold level Vth, and there are a plurality of consecutive black sides corresponding to the opaque part of the required spatial frequency of the chart. A method has been considered in which the MTF is calculated by setting the minimum value in the range _L4 that includes the photo element output signal as the minimum value b. According to this method, even if the spatial frequency of the chart becomes high, the maximum value a and the minimum value b can be distinguished. However, if the abnormal value a ₀ b ₀ due to the photo element output signal of abnormal sensitivity in the photo element output signal outputted in time series from the solid-state scanning element as shown in Fig. 9 is used instead of the true maximum value a and the true minimum value b. to
There is a high probability that it will be used for MTF calculation. This is because the abnormally sensitive photo element output signal _a0 , which has too high a sensitivity, becomes a maximum value wherever it is within the range _L3 , resulting in an incorrect MTF calculation.
b ₀ also causes a similar problem in the range L ₄ .

Furthermore, a method similar to the above method, in which a predetermined number of signals at both ends of each range L ₃ and L ₄ are discarded, and the average value of the remaining signals is taken and used as a and b has also been considered. However, with this method, each range
Since the signals at both ends of L ₃ and L ₄ are discarded, the probability of picking up abnormal signals due to abnormally sensitive photo elements decreases a little, but it is still quite high. Moreover, since the average value of the remaining signals is taken, another problem arises. In other words, even though the output signal of the solid-state scanning element is close to a sine wave, the remaining signals, which should have a magnitude relationship in principle, are averaged, so the true maximum and minimum values cannot be obtained, and the correct MTF cannot be obtained. Calculations will no longer be performed.

The present invention provides a signal processing method in an MTF measurement device that can perform correct MTF calculations while minimizing the probability of picking up abnormal signals, and can automatically adapt to any chart spatial frequency. The purpose is to

First, using the conventional optical Fourier transform method,
We will explain the MTF measurement principle and the MTF measurement principle using a solid-state scanning element. In the conventional optical Fourier transform method, a slit 9 is placed in front of a light source 8 as shown in FIG. The temporal brightness and darkness can be captured by the receiver 11 and displayed by the oscilloscope 12, or the slit 9 and the sine wave chart 10 can be exchanged and the sine wave chart 10 on the object side can be scanned as shown in FIG. 2b. Generates a temporal sine wave. Note that to obtain the MTF, an operation is performed in which the difference between the peak value and the trough value of the sine wave captured by the light receiver 11 is divided by the sum thereof.
On the other hand, as shown in Fig. 2c, the MTF measurement method using a solid-state scanning element places a sine wave chart 10 in front of the light source 8, similar to the method shown in Fig. 2b;
Unlike the above method, the sine wave chart 10 remains fixed, and the solid state scanning element 13 receives the projected image 10a of the sine wave chart 10 by the lens 4 to be inspected. Since the solid-state scanning element 13 has a self-scanning function, it converts a stationary spatial sine wave into a temporal sine wave, and the waveform is displayed on the oscilloscope 12 in the same manner as in the methods shown in Fig. 2 a and b. be done.

Figure 3 shows an example of an MTF measuring device using a solid-state scanning element, and is intended for inspection and adjustment of enlarger lenses and microlenses. In the figure,
Reference numeral 8 denotes a light source, and although a fluorescent lamp is shown here, it may also be a halogen lamp combined with a condensing illumination system. Reference numeral 14 denotes a diffuser plate, which is inserted for the purpose of eliminating unevenness in illuminance distribution over the entire screen and reducing the coherency of illumination light. 10 is a chart board with similar chart elements 10 ₁ at all positions corresponding to the required object height or image height as shown in FIG.
10 ₂ ... are arranged. These chart elements 10 ₁ , 10 ₂ . . . all show charts in the radial direction, but it goes without saying that they can also be arranged in the tangential direction on the same chart plate 10 at the same time, as shown in FIG. Also, each chart element 10 ₁ , 10 ₂ . . . may be a chart of a single frequency or a selected number of frequencies, but it must be a relatively wide transparent part for standardization. It has an opaque portion and is configured so that the MTF of the zero spatial frequency can be approximately normalized to 1 at each object height or image height. The lens 4 to be tested is attached to the lens holder 3, and is moved up and down by the focusing mechanism of the lens 4 itself or the focusing mechanism of the lens holder 3 to perform focusing. Reference numeral 15 denotes a plate on which the solid-state scanning elements 13 are installed on one plane, and in this example, nine solid-state scanning elements 1 are installed.
3 ₁ to 13 ₉ are chart elements 10 of the chart board 10
₁ , 10 ₂ . . . are installed corresponding to the array.

With such a device, all the information for each image height and direction angle can be obtained at once, so that electrical calculation processing makes it possible to determine the quality of the tested lens in a short time.

FIG. 6 shows the internal structure of the photoelectric conversion element CCD used as the solid-state scanning element 13, and FIG. 7 is its timing chart. Here, the photoelectric conversion section 16 consisting of 1728 photo elements has a photoelectric conversion ratio of 13 μm.
They are lined up at a pitch of m, photoelectrically convert the incident imaging light, and the photoelectrically converted signal is input to the analog shift registers 19, 20 through the transfer gates 17 and 18 by the transfer clock, and is output to the output gate by the shift clock φ ₁ and φ ₂ . 21 and is output in chronological order. In the apparatus shown in FIG. 3, light is emitted from a light source 8 and is diffused by a diffuser plate 14 to become uniform light on the chart plate 10.
to illuminate. Chart element 10 of chart board 10
₁ , 10 ₂ . . . is imaged as a black and white pattern on the CCDs _{13 1} to 13 ₉ via the test lens 4, and output signals are output from the CCDs _{13 1} to 13 ₉ in time series. At this time, from the output signal of the CCD 13, find the reference maximum value A, the reference minimum value B, the maximum value a and the minimum value b corresponding to the transparent and opaque parts of the required spatial frequency of the chart, and from these (a-b )/
The MTF of the lens 4 to be tested is calculated by the calculation (A-B)×100.

FIG. 8 shows the electrical part of this MTF measuring device.
CCD 13 ₁ to 13 ₉ are each chart element 10 ₁ , 10 ₂
The imaging light from... is photoelectrically converted and output in time series, but the analog multiplexer 22
CCDs 13 ₁ to 13 ₉ are selected one by one in sequence. The output signal of the selected CCD is converted into digital data by the A/D conversion circuit 23, and the above-mentioned A, B, a, and b are obtained by the arithmetic circuit 24 and then (a-b)/(A-B).
The MTF is calculated by the operation ×100. The control circuit 25 includes an analog multiplexer 22, A/
Controls the D conversion circuit 23, arithmetic circuit 24, and digital multiplexer 26, and displays the displays ₂₇₁ to _279.
It is selected in response to the selection of CCDs 13 ₁ to 13 ₉ and the calculation results of the calculation circuit 24 are displayed on the display. The measurement ends when all such operations are completed for the output signals of the CCDs 13 ₁ to 13 ₉ . Although not shown in FIG. 8, it is possible to determine whether the lens to be tested is acceptable by providing a determination circuit that determines acceptance or failure based on the above calculation results and set numerical values.

FIG. 10 shows an example of an output signal of the CCD 13 corresponding to a chart of a certain spatial frequency. As described above, when the spatial frequency becomes high to a certain extent, even if the original chart is a black and white rectangular wave pattern, the envelope of the output signal becomes sinusoidal. Therefore, it becomes difficult to clearly distinguish the signal for the white part of the chart from the signal for the black part. Therefore, the threshold level Vth is determined using the usual method, and a plurality of consecutive photo elements on the white side corresponding to the transparent part of the required spatial frequency of the chart are output based on the magnitude relationship with respect to the threshold level Vth of the output signal of the CCD 13. A white level region L ₃ containing signals and a black level region L ₄ containing a plurality of continuous photo element output signals on the black side corresponding to opaque parts of the required spatial frequency of the chart are divided. In the present invention, MTF
Regions L ₃ and L ₄ are used as output values a and b for calculation, respectively.
The feature is that the output signal of the photo element located in the center of the image is used. If there is an odd number of signals in the area, one signal is necessarily located in the center, but if there is an even number of signals in the area, one of the two signals closest to the center of the area is determined. Either the signal is unilaterally adopted according to prespecified conditions, or the following method is adopted. That is,
In region _L3 , the larger signal of the two signals closest to the center is used, and in region _L4 , conversely, the smaller signal is used. Alternatively, in order to avoid the trouble of determining the magnitude of a signal, the average value of two signals that are closest to the center of the area and which, in principle, do not have much difference in level may be used. In this way, the probability that the adopted signal is a signal from an abnormally sensitive photo element of the CCD is much smaller than in the conventional method. Also, as in the conventional method, when a and b are obtained by discarding both ends of the region and averaging the remaining three or more consecutive signals, a is smaller than the true maximum value and b is smaller than the true minimum value. will also get bigger
Although the calculated value of MTF will be smaller than the true MTF, in the present invention, there is almost no risk of this happening due to the above method.

In Figure 10, the threshold level
Region _L3 , where the signal is greater than _Vth , includes 10 photoelement output signals. this area
If the signals of _L3 are designated as _S1 ... _S10 from the right, there is no signal located exactly in the center of area _L3 , so we focus on signals _S5 and _S6 on both sides of the center. Since the envelope of the signal is sinusoidal, the envelope of the signal within region _L3 is close to a chevron-shaped symmetrical curve that reaches its maximum at the center position. Therefore, signals _S5 and _S9 are approximately at the same level. Therefore, let S ₅ = a ₁ and S ₆ = a ₂ , and use the following methods to determine the maximum value a used for MTF calculation.
Determine either of these using a circuit.

The signal that appears first in time series is used.

a=a ₁ Use a signal that appears later in time series.

a=a ₂ Compare a ₁ and a ₂ and use the larger one.

a= _max (a ₁ , a ₂ ) Since a ₁ and a ₂ are approximately at the same level, the average of both is taken.

a = 1/2 (a ₁ + a ₂ ) In Figure 10, the shock hold level
Since nine signals are included in the region _L4 where the signal is smaller than _Vth , only one signal located in the center is determined and this is designated as b. Since each photo element (photoelectric conversion unit) of the CCD 13 has a slight difference in sensitivity, there is no guarantee that this b is the minimum value in the region _L4 , but it has the highest probability of being the minimum value, and if Even if it is not a local minimum value, it can be guaranteed that it is very close to the local minimum value. If the number of signals included in region _L4 is an even number, b may be determined using a method similar to the method for determining a described above. However, instead of the above, compare 3a b ₁ and b ₂ and use the smaller one.

Let b=m _io (b ₁ , b ₂ ).

Next, embodiments of the present invention will be described along with specific circuit configuration examples. Although this example is an example in which all signals are processed in analog form, the output signal of the solid-state scanning element may be amplified, A/D converted, and then digitally processed.

In this embodiment, a white level area L ₃ and a black level area L ₄ are measured in an MTF measuring device using the CCD 13.
MTF calculation is performed by sampling each center signal (L ₃ /2, L ₄ /2 in time) as a and b, but using the single spatial frequency chart at that time, the first set of white The centers of the level area and the black level area are detected and stored as counts, and the centers of the second and subsequent sets of white level areas and black level areas are sampled.

FIG. 11 shows the circuit configuration of this embodiment, and the 12th
The figure shows the timing chart. CCD13
The output signal is amplified by the amplifier 28, and the threshold hold level V _th
are binarized into white level area _L3 and black level area
L ₄ are distinguished. Flip flop 30, 31
A circuit 33 consisting of a gate 32 and a gate 32 selects the first set of white level regions from the output signal of the comparator 29.
L ₃ and black level region L ₄ are detected and their output signals are sent to gates 34 and 35. The gate 34 receives the output signal of the comparator 29 and becomes valid in the first white level region _L3 corresponding to the first transparent part of the transparent part and opaque part of the chart's required spatial frequency, and the CCD clock is set to 1/2. 1/ with circuit 36
The frequency is divided into 2, passes through a gate 34, is counted by a binary counter 37, and is stored. At this time, the binary counter 37 counts at 1/2 of the advance amount of the CCD clock, so that the center of the white level area _L3 is detected and stored.
Further, the output signal of the comparator 29 is inverted by the inverter 38 and inputted to the gate 35, and the first black level region _L4 corresponding to the first opaque part in the transparent part and opaque part of the required spatial frequency of the chart is input.
becomes valid. Therefore, the CCD clock is divided in half by the 1/2 frequency divider 36, passes through the gate 35, is counted by the binary counter 39, and is stored, whereby the center of the black level area _L4 is detected and stored. Ru. Next, from the circuit 33 to the gate 4
0 and 41, and the set circuit 42 detects the point at which the output signal of the comparator 29 switches. And the down counter 43 is set circuit 4
The stored data of the binary counter 37 is set by the output of 2, and is counted down by the CCD clock. The gate 40 receives the outputs of the down counter 43 and the comparator 29, and when the contents of the down counter 43 reach 0, outputs an output signal, opens the analog switch 44, and outputs the signal from the amplifier 28. The output signal from this analog switch 44 is the central signal in the second white level area _L3 corresponding to the second transparent part of the transparent part and opaque part of the chart's required spatial frequency, and is the central signal of the chart's required spatial frequency. The third and subsequent white level areas corresponding to the third and subsequent transparent areas in the transparent and opaque areas are output in the same manner. Further, the down counter 45 has the stored data of the binary counter 39 set by the output of the set circuit 42, and counts down by the CCD clock. The gate 41 receives the output of the down counter 45 and the output of the comparator 29 via the inverter 38, and outputs an output signal when the content of the down counter 45 reaches 0 at the center of the black level area _L4 . The analog switch 46 is opened to output the signal from the amplifier 28. This analog switch 4
The signal from No. 6 becomes the center signal in the second and subsequent black level regions. MTF calculation is performed by a circuit (not shown) using signals from analog switches 44 and 46. binary counter 37,3
9. Flip-flops 30 and 31 are reference black level regions having only the minimum value B corresponding to the opaque part of the chart in FIG. 12 with a spatial frequency close to zero.
Since it is reset at _L2 , even if multiple CCDs are used, each area _L3 and _L4 is reset for each CCD.
The median value of can be sampled.

Incidentally, in the above embodiment, the output of the CCD is processed as an analog signal, but it is also easy to digitize it with an A/D converter and process it.

As described above, the present invention includes a projection chart, an illumination means for illuminating the chart, and a large number of photo elements for photoelectrically converting the image of the chart projected by the test element. In an MTF measurement device that is equipped with a solid-state scanning element that outputs an output signal in time series, and calculates the MTF of the device under test by finding the minimum and maximum values of the output signal of the solid-state scanning element, the output of the solid-state scanning element is The signal is compared with a predetermined level and divided into a white level area containing the output signals of a plurality of continuous photo elements on the white side and a black level area containing the output signals of a plurality of continuous photo elements on the black side. counting the number of each photo element output signal in the white level area and the black level area, and detecting the position of each center photo element output signal or the photo element output signal closest to each center in the white level area and the black level area; The photo element output signals at the center of the second and subsequent white level areas and black level areas or the photo element output signals closest to each center are detected as the true maximum and minimum values, and these white and black areas are detected. Since the MTF is calculated by
Not only can you perform MTF calculations, but also
It can automatically adapt to any spatial frequency.

[Brief explanation of the drawing]

FIG. 1 is a perspective view showing an example of a conventional MTF measuring device, FIG. 2 is a diagram for explaining the principle of MTF measurement, and FIG. 3 is a perspective view showing an example of an MTF measuring device using a solid-state scanning element. Figures 4 and 5 are plan views showing examples of chart boards, Figure 6 is a block diagram showing the internal structure of the CCD, Figure 7 is a timing chart showing the operation of the CCD, and Figure 8 is the diagram shown in Figure 3. Block diagram showing the electrical circuit used in the device, No. 9
10 is a waveform diagram showing the CCD output signal of the device shown in FIG. 3. FIG. 11 is a block diagram showing an example of the circuit according to the present invention.
FIG. 12 is a timing chart showing the operation of the same implementation circuit. 29... Comparator, 33... First set detection circuit, 34, 35, 40, 41... Gate, 3
6...1/2 frequency divider circuit, 37, 39... Binary counter, 38... Inverter, 42... Set circuit, 43, 45... Down counter, 4
4,46...Analog switch.

Claims (1)

[Claims] 1 A projection chart, an illumination means for illuminating the chart, and a large number of photo elements for photoelectrically converting the image of the chart projected by the test element, and outputting the output signals of these photo elements in time series. In an MTF measurement device that is equipped with a solid-state scanning element and calculates the MTF of the device under test by determining the maximum and minimum values of the output signal of the solid-state scanning element, the output signal of the solid-state scanning element is compared with a predetermined level. The first white level area and black level area are divided into a white level area containing the output signals of multiple continuous photo elements on the white side and a black level area containing the output signals of multiple continuous photo elements on the black side. The number of output signals of each photo element is counted in the white level area and the black level area, and the position of the photo element output signal at the center or the photo element output signal closest to each center is detected. The feature is that the photo element output signal at each center in the white level area and the black level area or the photo element output signal closest to each center is detected as local maximum and local minimum values, and these local maximum and local minimum values are used for MTF calculation. Signal processing method for MTF measurement equipment.