JPS6314887B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a method for correcting sensitivity between photoelements in a solid-state sensor in which a plurality of photoelements such as photodiodes or phototransistors are arranged in parallel.

In optical measurement devices such as MTF (modulation transfer function) inspection machines that inspect lens aberrations, etc., solid-state sensors such as CCD (charge coupled device) sensors are used to measure light intensity distribution. A CCD sensor, for example, consists of 2048 (2048-bit) photoelements arranged in parallel with a center-to-center distance of 13 μm, and transmits the output corresponding to the exposure amount of each photoelement to a signal processing device to calculate the exposure distribution, etc. It can be used for measuring. Therefore, each photoelement of a CCD sensor has variations in its sensitivity due to variations in manufacturing parameters during the manufacturing process and properties of the material itself such as wafer defects, which causes errors in light intensity measurement of the CCD sensor. . However, conventionally,
Although it was possible to correct the sensitivity between each CCD sensor, the non-uniformity of the optical response of each photoelement ( so-called
PRNU) and the non-uniformity of the dark current of each photoelement (so-called DSNU), it was not possible to correct the sensitivity between each photoelement.

The present invention has been made in view of the above points, and in addition to correcting the sensitivity between each solid-state sensor, it is also possible to correct variations in sensitivity between each photoelement in each solid-state sensor, and enables high-precision optical measurement. The purpose of this invention is to provide a method for correcting the sensitivity of a solid-state sensor that can be used to implement a device. That is, the feature of the method of the present invention is that, in the sensitivity correction method for a solid-state sensor having a plurality of photoelements, the sensitivity characteristics with respect to the exposure amount of a reference photoelement selected as a sensitivity standard and other photoelements are determined in advance. Another feature of the present invention is that, when measuring the amount of light of the object to be measured, the output of each photoelement other than the reference photoelement is corrected based on the difference in the sensitivity characteristics with respect to the reference photoelement.

Solid state sensors such as CCD sensors, for example, 2048
photoelements are arranged side by side with a center-to-center distance of 13μm, and the sensitivity characteristics of each photoelement are
As shown in the figure, where the horizontal axis represents the exposure amount and the vertical axis represents the photoelement signal output (volts), there is linearity with respect to the exposure amount when used within the guaranteed operating frequency range. are doing. Therefore, the output v of each photoelement is expressed by a linear function of the exposure amount E, but as mentioned above, there are differences in the photoresponse and dark current of each photoelement, and the constant of the linear function is different for each photoelement. It depends.
This can be expressed mathematically as shown in equation (1) below.

V _i =A _i E+B _i (1) where A _i and B _i are constants for the i-th photoelement. Note that dark current is a signal component that is detected during non-exposure and is generated by heat in the CCD sensor. In Figure 1, the value of the vertical axis intercept, and in equation (1), B _i represents this dark current. ing. The photoresponsiveness is expressed by the slope of the straight line in FIG. 1 and A _i in equation (1). And for each photoelement A _i
Since the and B _i values are different, even when each photoelement is irradiated with light at the same amount of light, there is a difference in output, and a difference in sensitivity exists.

In the present invention, the sensitivity characteristics of each photoelement are determined in advance as A _i and B _i values, and when measuring the amount of light of the measurement target, the output of each photoelement is
The correction is made to match the sensitivity characteristics of a reference photoelement j appropriately selected as a sensitivity reference. That is, in the method of the present invention, in the sensitivity correction method for a solid-state sensor having N photoelements, the sensitivity characteristics of each photoelement are expressed by the following formula: V _i =A _i E+Bi (1) Measurement of light intensity of the measurement target Prior to this, each photoelement is exposed twice with different uniform exposure amounts, and the constants A _i and B _i of the formula are determined from each output,
The present invention is characterized in that each output v _i of each photoelement during light intensity measurement is corrected according to the following formula Vi ₌ (A _j /A _i ) v _i +B _j - (A _j /A _i )B _i It is something.

However, v _i : Output of photoelement E : Exposure amount A _i , B _i : Constant for i-th photoelement (i=1, 2,...N) V _i : Output of photoelement i after correction A _j , B _j : Constant for the reference photoelement j selected as a sensitivity reference In this case, one of the different uniform exposure amounts is zero light amount, and therefore the output of each photoelement at that time is the dark current. It is preferable that the This is because the B _i can be found in direct correspondence to the output of the photoelement.
Note that when the sensitivity characteristic of each photoelement is not expressed by the above equation (1) and another function is used, the above-mentioned exposure with different uniform exposure amounts is not necessarily limited to two times. The number of uniform exposures may be appropriately set depending on the number of constants used in the sensitivity characteristic equation to be set.

The present invention will be explained in detail below. Assume that the number of photoelements in the CCD sensor is N (N bits), and each photoelement is uniformly exposed with an exposure amount E. Each photoelement i(i
= 1, 2, ..., N; i≠j) output v _i and the output v _j of the reference photoelement j selected as the sensitivity standard
is expressed by the following formulas (2) and (3).

v _i =A _i E+B _i ...(2) v _j =A _j E+B _j ...(3) Since the output of other photoelement i is adapted to the sensitivity characteristic of photoelement j as a reference, the exposure In quantity E, the output v _i of photoelement i should be modified to the output v _j of reference photoelement j. This correction formula is expressed by the following formula (4).

V _i = (A _j /A _i )v _i +B _j - (A _j /A _i )B _i ...(4) That is, the actual output v _i (=A _i E+
By substituting B _i ) into equation (4), V _i = v _j (=A _j E+B _j )
Then, the corrected output V _i of photoelement i is v _j
matches. Therefore, when measuring the light intensity of the measurement target using a CCD sensor, each photoelement i
If the output v _i of (i=1, 2,...N; i≠j) is corrected according to equation (4), the exposure amount can be converted to voltage for all photoelements using the sensitivity characteristics of photoelement j. This means that it has been converted, and variations in sensitivity can be corrected.

Therefore, in order to perform the correction calculation based on equation (4), it is necessary to obtain the constants A _i and B _i (including A _j and B _j ) in advance, and to obtain the sensitivity characteristics of each photoelement. Therefore, prior to measuring the light intensity of the measurement target, each photoelement of the CCD sensor is exposed twice with different uniform exposure amounts, and A _i and B _i are calculated from each output of each photoelement at that time. In this case, one of the two uniform exposures is made with zero exposure, that is, the output of each photoelement is detected in a dark state during non-exposure,
It is better to find the dark current component B _i directly from this output. Then, by eliminating E from equations (2) and (3), we get the following (5)
The formula is obtained, A _j /A _i = (v _j - B _j ) / (v _i - B _i ) ...(5) For each photoelement in another uniform exposure (exposure amount is not zero) A _j /A _i is determined from the outputs v _i and v _j according to equation (5).

Next, a sensitivity correction device capable of performing the sensitivity correction as described above will be described. FIG. 2 is a block diagram showing one embodiment of the sensitivity correction method of the present invention incorporated into the Fourier calculation circuit of the MTF tester.
In addition, during Fourier calculation, among the output signals of each photoelement, components that change due to exposure,
That is, it is necessary to use only A _i E and exclude the dark current component B _i which is the DC component of the signal, so in the following description, a device for correcting v _i −B _i will be described. Therefore, the correction equation of equation (4) is transformed as shown in equation (6) below.

V _i −B _j = (A _j /A _i ) (v _i −B _i ) ...(6) In other words, A _j /A _i calculated in advance using equation (5) and the dark current component
From B _i , correction calculation of v _i can be performed according to equation (6).

In the example shown in Figure 2, a 256-bit CCD
Six sensors 100 are arranged in parallel, and each CCD sensor 1
00 is connected in parallel to multiplexer 101,
Each photoelement i (i=
The output of (1 to 256) x 6) is the sample and
The data passes through a hold circuit 102 and an analog/digital converter (hereinafter abbreviated as ADC) 103.
It is input to RAM 104 and stored. The output of data RAM 104 is adder/subtractor 1
10 and multiplier 111;
0 and the output of multiplier 111 are input to input port 112
The data is input to the arithmetic unit 113 via. The output terminal of the arithmetic unit 113 is connected to the data RAM 104, the constant RAM 106, and the constant through the output port 114.
Connected to RAM108, constant RAM106,1
The output of 08 is input to an adder/subtracter 110 and a multiplier 111. Constant RAM106
is a RAM for storing correction constants A _j /A _i and B _i , and a constant RAM 108 stores sin and cos tables used when performing Fourier calculations on the output of the photoelement after correction. There is. 105, 107 and 109 are data respectively
RAM104, constant RAM106 and constant RAM1
08 address counter. Further, as shown in the figure, gates G1 to G14 are installed between the circuits, and each gate G1 to G14 is normally closed and is controlled to open or close by a controller (not shown). Note that the signal input to the adder/subtractor 110 via the gate G9 is a zero signal.

Next, the operation of this sensitivity correction device will be explained.

() When calculating each correction constant prior to MTF measurement.

(1) Each photoelement of the six CCD sensors 100 is uniformly exposed with the same amount of light, and gate G1
and open G3 to output the output of each photoelement i.
v _i (i=(1~256)×6) as data RAM10
Store in 4.

(2) Next, close gates G1 and GG3, open gates G9 and G12, and read data RAM1.
04 is input to the arithmetic unit 113 via the input port 112.

(3) Then close gates G9 and G12,
The CCD sensor 100 is set to zero exposure, that is, in a dark state, gates G1 and G3 are opened, and the output B _i of each photoelement i (i = (1 to 256)
×6) is input to the data RAM 104, and the stored contents of the data RAM 104 are rewritten with B _i .

(4) Next, close gates G1 and G3, open gates G9 and G12, and read data RAM 10.
4 is input to the arithmetic unit 113 via the input port 112.

(5) The arithmetic unit 113 receives the photoelement i (i=(1~256)×6) input as described above.
Based on the appropriately selected reference photoelement j from v _i and B _i for
Calculate A _j /A _i and obtain the obtained A _j /A _i and B _i
For example, store it on a floppy disk. Gates G9 and G12 are arithmetic unit 113
is closed after reading the data.

() When correcting the output of each photoelement in MTF measurement (1) The calculation device 113 reads the A _j /A _i and B _i values stored in the floppy disk, and opens gates G14 and G4. These values are stored in constant RAM 106. In addition,
As is clear from the above, constant RAM106
has 256×6×2 addresses since each of the six CCD sensors 100 has 256 bits. Then gates G14 and G4
Close.

(2) Gates G1 and G3 are opened, and the output v _i of each photoelement of the CCD sensor 100 receiving exposure from the measurement target is stored in the data RAM 10.
4 and store it. Then, gates G1 and G3 are closed.

(3) Open gates G10, G2 and G3 and output v _i stored in data RAM 104;
Constant B _i stored in constant RAM 106
are passed through the adder/subtractor 110 to calculate v _i −B _i , and the result of the calculation is stored in the data RAM 104.
, and rewrite the stored contents of the data RAM 104 as v _i -B _i . Then, gates G10, G2, and G3 are closed.

(4) Open gates G7, G6 and G3 and read the data.
The v _i -B _i value stored in the RAM 104,
Constants stored in constant RAM 106
A _j /A _i is passed through the multiplier 111 to perform the calculation of equation (6), and the calculation result is stored in the data RAM.
104 to rewrite the stored contents of the data RAM 104 as (A _j /A _i )(v _i −B _i ) [see equation (6)]. And gate G7,G
Close 6 and G3.

Therefore, the content stored in the data RAM 104 at this point is (A _j /A _i )(v _i −B _i ) for each photoelement, and the actual measured value v _i of the output is calculated using equation (6). The result of the correction calculation according to is stored in the data RAM 104. Therefore, (A _j /A _i )(v _i −B _i ) stored in the data RAM 104 and sin and
cos table and gates G8 and G13
By opening the multiplier 111, the data is multiplied by the multiplier 111, and then input to the arithmetic unit 113, and the MTF test program built in the arithmetic unit 113 performs necessary operations such as Fourier operation to obtain the MTF test result.

The opening and closing operations of each of the gates G1 to G14 as described above can be performed by a controller using a microcomputer or the like. One example is shown in FIGS. 3 to 6. Figure 3 shows the controller 12
A block diagram illustrating data input/output control using 0, and FIGS. 4 to 6 are control timing diagrams. For the sake of simplicity, CCD sensor 1
For example, 00 is illustrated for the case where one 256-bit bit is used. Further, in FIG. 3, the same parts as in FIG. 2 are given the same reference numerals.
In FIG. 3, 121 is an output circuit for a unit signal "1";
122 is a word counter. Further, 1 is a data bus, and 2 to 19 are controllers 120.
bus for transmitting control signals from to each circuit, 20 to 2
A bus 4 transmits various commands to the controller 120.

Then, when a command to read the output of the CCD sensor 100 into the data RAM 104 is input to the controller 120 from the bus 20, the output v _i (zero) of the 256 photo elements of the CCD sensor 100 at the timing shown in FIG. (at the time of light intensity exposure and at the time of exposure with uniform light intensity) are stored in the data RAM 104. Next, the storage contents of the data RAM 104 are transferred to the arithmetic unit 1.
13 and calculate A _j /A _i , then A _j /
Store A _i and B _i in constant RAM 106.

For MTF measurement, the CCD sensor 1 to which the measurement target was exposed was similarly placed at the timing shown in Figure 4.
The output v _i of each photoelement of 00 is stored in the data RAM
104, and at the timing shown in FIG.
From this v _i and B _i of the constant RAM 106, v _i −B _i is calculated, and the calculation result is stored in the data RAM 104. Then, at the timing shown in FIG. 6, v _i −B _i stored in the data RAM 104 and the constant
From A _j /A _i in RAM 106, the two are multiplied and the multiplication result is stored in data RAM 104. As mentioned above, when measuring MTF, the actual measured output v _i
is subjected to correction calculation according to equation (6), and the output after correction is
V _i -B _j is determined as the storage content of the data RAM 104.

The output V _i of each photoelement i (i≠j), which has been corrected as described above, matches the sensitivity characteristics of the reference photoelement j. If the exposure amount is uniform, the output signals of all photoelements (including photoelements of separate solid-state sensors) will have the same output value, but due to variations in photoresponsivity and dark current in each photoelement, The variations in sensitivity between the two are corrected. A solid-state sensor to which the sensitivity correction of the present invention is applied makes it possible to perform highly accurate optical measurements.

Note that the present invention should not be limited to the specific embodiments described above, and the present invention is applicable not only to CCD sensors but also to solid-state sensors having a plurality of photoelements. Of course, various modifications are possible within the technical scope.

[Brief explanation of the drawing]

FIG. 1 is a graph showing the sensitivity characteristics of the photoelement, FIG. 2 is a block diagram showing one embodiment of the device of the present invention, FIG. 3 is a block diagram explaining the data transfer function of the controller, and FIGS. FIG. 6 is a data control timing diagram of the controller. (Explanation of symbols) 1...Data bus, 100...
CCD sensor, 103...analog/digital converter, 104...data RAM, 106...constant RAM,
110... Adder/subtractor, 111... Multiplier, 113... Arithmetic device, 120... Controller.

Claims (1)

[Claims] In a sensitivity correction method for a solid-state sensor having 1N photoelements, the sensitivity characteristics of each photoelement are expressed by the following formula, v _i = A _i E + B _i Prior to measuring the light amount of the measurement target, the above-mentioned Each photoelement is exposed twice with different uniform exposure amounts, and the constants A _i and B _i of the above formula are determined from each output.
The present invention is characterized in that each output v _i of each photoelement during light intensity measurement is corrected according to the following formula Vi ₌ (A _j /A _i ) v _i +B _j - (A _j /A _i )B _i Sensitivity correction method for solid-state sensors. However, v _i : Output of photoelement E : Exposure amount A _i , B _i : Constant for i-th photoelement (i=1, 2,..., N) Vi _: Output A of photoelement i after correction _i , B _i : constants for the reference photoelement j selected as the sensitivity reference. 2. The solid-state sensor sensitivity correction method according to claim 1, wherein one of the different uniform exposure amounts is zero light amount.