JP7046994B2 - Image reader and image reading method - Google Patents

Description

The present invention relates to an image reading device and an image reading method for reading image data from an image sensor.

The output value (hereinafter referred to as pixel output value) read from each pixel of the image sensor varies from pixel to pixel due to the difference in wiring length structurally caused by the pixel arrangement. Conventionally, the pixel output value of the image sensor has been corrected for the variation of the pixel output value of the image sensor for each pixel by using a moving average or a filter according to the structure such as the pixel arrangement of the image sensor. For example, see Patent Document 1).

In Patent Document 1, the pixel output variation component of the line sensor and the light distribution unevenness component of the illumination optical system are separated, and the shading correction coefficient is obtained from the pixel output variation component of the line sensor, and the light distribution unevenness component of the illumination optical system is obtained. It is disclosed that the optical system light distribution correction coefficient is calculated from the above and the reading data of the line sensor is corrected.

Japanese Unexamined Patent Publication No. 2000-224419

However, the variation (error) in the pixel output value of the image sensor for each pixel varies depending on the amount of input light to each pixel and the charge accumulation time in each pixel. The principle of its generation is as follows.

As shown in FIG. 1, in general, the image sensor 10 is stored from a pixel 12 composed of a photodiode 13 that converts light into electric charges, a charge amplifier 14 that stores the converted electric charge, and a plurality of pixels 12. The shift register 15 for reading the electric charge and the shift register reading circuit 16 are provided. The image sensor 10 of FIG. 1 includes 256 pixels 12 ₁ , 12 ₂ , ..., 12 ₂₅₆ . Generally, the wiring 17 between the photodiode 13 and the charge amplifier 14 differs for each pixel depending on the charge amplifier arrangement of the image sensor 10 and the number of pixels. As the charge amplifier array, for example, there are a staggered array with odd and even pixels as shown in FIG. 1, a staggered array with multiple pixels (for example, three stages) (see FIG. 2), a composite array thereof, and the like, and charging with the photodiode 13. There is a difference in the wiring length with the amplifier 14. Since the wiring 17 has an electric resistance (R), when a charge flows, a loss (loss = I ² R) proportional to the square of the charge flow (current I) occurs. There is a difference.

FIG. 2 is a diagram showing a configuration of an image sensor 10A having a three-stage staggered arrangement. As shown in FIG. 2, the charge amplifiers 14 ₁ , 14 ₃ , ₁₄₅ ... Connected to the photodiodes 13 ₁ , 13 ₃ , ₁₃₅ ... Are on one side (odd side) of the photodiode array 11. ), And the charge amplifiers 142, ₁₄₄ , ₁₄₆ ..., which are arranged in _three stages and connected to the photodiodes 132, ₁₃₄ , ₁₃₆ ..., are _on the other side (even number) of the photodiode array 11. It is arranged in three stages on the side). As can be seen from FIG. 2, the wiring length between the _third -stage charge amplifier 141 and the photodiode 13 ₁ on the odd side is L1, and the wiring length between the second-stage charge amplifier 14 ₃ and the photodiode 13 ₃ . L3, the wiring length _L5 between the first _- stage charge amplifier 145 and the photodiode 135 has a relationship of L1>L3> L5. In this case, the stored charge amount (pixel output value) _{Q1 of the pixel 12 1 having} the _third -stage charge amplifier 141, the stored charge amount Q3 of the pixel 123 having the second-stage charge amplifier 14 ₃ The accumulated charge amount Q5 of the pixel ₁₂₅ having the charge amplifier ₁₄₅ is Q1 <Q3 <Q5. The same applies to the even-numbered side.

Further, since the charge generated by the photodiode 13 is proportional to the amount of input light to the photodiode 13, the larger the input light amount, the larger the loss, and the difference in the accumulated charge amount (pixel output value) increases (see FIG. 3). .. FIG. 3 is a diagram showing variations in pixel output values depending on the amount of input light when the image sensors have a three-stage staggered arrangement. As described above, the amount of accumulated charge (pixel output value) for each pixel varies due to the two fluctuation factors of the wiring length difference between the pixels and the amount of input light to the photodiode.

Further, since the accumulated charge amount (pixel output value) is proportional to the accumulation time, it is necessary to adjust the correction value according to the accumulation time even with the same accumulated charge amount (pixel output value) (see FIG. 4). FIG. 4 is a diagram showing the variation in the pixel output value due to the accumulation time when the image sensor has a three-stage staggered arrangement. As described above, the amount of accumulated charge (pixel output value) for each pixel varies due to the two fluctuation factors of the wiring length difference between the pixels and the accumulation time.

As described above, the variation in the pixel output value due to the difference in the wiring length of each pixel of the image sensor varies depending on the amount of input light to each pixel and the charge accumulation time in each pixel, which is described in Patent Document 1. The reading device did not take such fluctuations into consideration.

The present invention has been made to solve the above-mentioned problems, and is an image reading device and an image capable of accurately correcting the variation of the pixel output value for each pixel even if the input light amount and the storage time fluctuate. The purpose is to provide a reading method.

In order to achieve the above object, the image reading device of the present invention has an image sensor (10) having a plurality of pixels and an analog digital that acquires a digital pixel output value by analog-digital conversion of an electric signal obtained from each of the pixels. The converter (21), the correction data acquisition unit (30) for acquiring the correction data used for correcting the interpixel variation of the pixel output value, and the light used when the correction data acquisition unit acquires the correction data. A white light source (2) for outputting the above, a light amount setting unit (42) for adjusting the output light amount of the white light source, and a correction unit (23) for correcting the interpixel variation of the pixel output value based on the correction data. , And the correction data acquisition unit acquires the light amount characteristic as the correction data, which defines the change in the variation between the pixels of the pixel output value due to the change in the input light amount to each pixel (38). ), And the light amount characteristic acquisition unit includes a first measurement value storage unit (31) that stores the pixel output value of each pixel acquired each time the input light amount is changed as a first measurement value. The first correction value calculation unit (32) for calculating the first correction value for correcting the inter-pixel variation of the pixel output value based on the first measurement value stored in the first measurement value storage unit, and the first correction value calculation unit (32). A first approximation function determination unit (33) for determining an approximation function representing the relationship between one correction value and the pixel output value is provided, and the first correction value calculation unit includes the maximum value of the pixel output value and the said. It is characterized in that the difference between each pixel and the pixel output value is calculated as the first correction value .

With this configuration, the light amount characteristic acquisition unit acquires the light amount characteristic as correction data, which defines the change in the variation between pixels of the pixel output value due to the change in the input light amount to each pixel, and the correction unit acquires the pixel based on the correction data. Since the variation in the output value between pixels is corrected, it is possible to accurately correct the variation in the pixel output value for each pixel even if the amount of input light to each pixel fluctuates.

With this configuration, the first approximation function determination unit determines an approximation function representing the relationship between the first correction value and the pixel output value depending on the input light amount, and the correction unit determines the pixel of the pixel output value based on the approximation function. Since the intervening variation is corrected, even if the measurement data related to the first measured value is small, the variation of the pixel output value for each pixel can be accurately corrected in consideration of the variation of the input light amount to each pixel.

With this configuration, the first correction value can be easily calculated, and the correction value of the inter-pixel variation of the pixel output value can be efficiently calculated in consideration of the input light amount and the accumulation time.

In the image reading device of the present invention, each pixel has a photodiode that converts light into a charge and a charge amplifier that stores the converted charge, and the first correction value calculation unit is the photodiode. An approximation function is determined for the pixel output value of the pixel group having the shortest wiring length between the and the charge amplifier, and the maximum pixel output value obtained by the approximation function corresponding to each pixel and the said. The configuration may be such that the difference between the pixel output value of each pixel and the pixel output value is calculated as the first correction value.

With this configuration, the first correction value can be calculated accurately, especially when the group can be grouped according to the wiring length such as in a multi-stage staggered arrangement. Therefore, the pixel output value between pixels can be calculated in consideration of the input light amount and the accumulation time. The correction value of the variation can be calculated with high accuracy.

In the image reading device of the present invention, the approximation function determined by the first approximation function determination unit is
Corr = (Pmes2 / Prefmes2) × Corrref
Here, Prefmes may be a reference pixel output value, Corrref may be a reference correction value, Pmes may be a pixel output value to be corrected, and Corr may be the first correction value of the pixel output value to be corrected.

With this configuration, the approximation function can be easily calculated without using, for example, the least squares method.

In the image reading device of the present invention, the correction data acquisition unit acquires, as the correction data, the accumulation time characteristic that defines the change in the variation between the pixels of the pixel output value due to the change in the charge accumulation time in each pixel. It may be configured to further include an accumulation time characteristic acquisition unit (39).

With this configuration, the accumulation time characteristic acquisition unit acquires the accumulation time characteristic that defines the change in the variation between pixels of the pixel output value due to the change in the charge accumulation time in each pixel as correction data, and the correction unit obtains the correction. Since the variation between pixels of the pixel output value is corrected based on the data, it is possible to accurately correct the variation of the pixel output value for each pixel even if the charge accumulation time in each pixel fluctuates.

The image reading device of the present invention further includes a storage time setting unit (43) for setting the charge storage time in each pixel, and the storage time characteristic acquisition unit is acquired every time the storage time is changed. The pixel output value is based on a second measured value storage unit (34) that stores the pixel output value of each pixel as a second measured value and the second measured value stored in the second measured value storage unit. The second correction value calculation unit (35) for calculating the second correction value for correcting the variation between the pixels, and the second correction value obtained at the specific accumulation time were obtained at each accumulation time. A correction value ratio calculation unit (36) that calculates the ratio of the second correction value, and a second approximation function determination unit (37) that determines an approximation function that represents the relationship between the ratio of the second correction value and the accumulation time. ) May be provided.

With this configuration, the second approximation function determination unit determines an approximation function representing the relationship between the ratio of the second correction value and the accumulation time, and the correction unit corrects the interpixel variation of the pixel output value based on the approximation function. Therefore, even if the measurement data regarding the second measured value is small, it is possible to accurately correct the variation in the pixel output value for each pixel in consideration of the variation in the charge accumulation time in each pixel.

The image reading device of the present invention may be configured to further include a spectroscopic unit (5) in front of the image sensor.

With this configuration, even if the amount of input light and the accumulation time fluctuate, the variation of the pixel output value for each pixel is accurately corrected, and the light transmitted from the object to be measured is separated by the spectroscope and incident on the image sensor. , High-precision spectroscopic measurement is possible.

The image reading method of the present invention is an image reading method for reading image data from an image sensor (10) having a plurality of pixels, and an electric signal obtained from each of the pixels is analog-digitally converted by an analog-digital converter (21). The analog digital conversion step for acquiring the digital pixel output value, the correction data acquisition step for acquiring the correction data used for correcting the interpixel variation of the pixel output value, and the correction data by the correction data acquisition step. A light output step for outputting the light used at the time of acquisition from the white light source, a light amount setting step for adjusting the output light amount of the white light source, and a correction step for correcting the interpixel variation of the pixel output value based on the correction data. Including The light quantity characteristic includes a storage time characteristic acquisition step of acquiring the storage time characteristic as the correction data, which defines the change in the variation of the pixel output value between the pixels due to the change in the charge accumulation time in each pixel. The acquisition step is stored in the first measured value storage step for storing the pixel output value of each pixel acquired each time the input light amount is changed as the first measured value, and the first measured value storage step . The relationship between the first correction value calculation step for calculating the first correction value for correcting the inter-pixel variation of the pixel output value based on the first measurement value, and the first correction value and the pixel output value is shown. The first correction value calculation step includes a first approximation function determination step for determining an approximation function, and the first correction value calculation step sets the difference between the maximum value of the pixel output value and the pixel output value of each pixel as the first correction value. It is characterized by calculating as.

With this configuration, in the light amount characteristic acquisition step, the light amount characteristic that defines the change in the variation between pixels of the pixel output value due to the change in the input light amount to each pixel is acquired as correction data, and in the accumulation time characteristic acquisition step, each pixel The accumulation time characteristic that defines the change in the pixel output value between pixels due to the change in the charge accumulation time is acquired as correction data, and the correction unit corrects the pixel output value between pixel variations based on these correction data. Even if the amount of input light to each pixel and the accumulation time of the charge in each pixel fluctuate, the variation in the pixel output value for each pixel can be corrected with high accuracy.

With this configuration, the first approximation function determination step determines an approximation function representing the relationship between the first correction value and the pixel output value depending on the amount of input light, and the correction step determines the pixel of the pixel output value based on the approximation function. Since the intervening variation is corrected, even if the measurement data related to the first measured value is small, the variation of the pixel output value for each pixel can be accurately corrected in consideration of the variation of the input light amount to each pixel.

With this configuration, the first correction value can be easily calculated, and the correction value of the inter-pixel variation of the pixel output value can be efficiently calculated in consideration of the input light amount and the accumulation time.

INDUSTRIAL APPLICABILITY According to the present invention, it is possible to provide an image reading device and an image reading method that can accurately correct the variation of the pixel output value for each pixel even if the input light amount and the storage time fluctuate.

It is a figure which shows the schematic structure of an image sensor. It is a figure which shows the schematic structure of the image sensor of a three-stage staggered structure. It is a figure which shows the relationship between the pixel number and the pixel output value when the input light amount changes in the image sensor of a three-stage staggered structure. It is a figure which shows the relationship between the pixel number and the pixel output value when the accumulation time changes in the image sensor of a three-stage staggered structure. It is a figure which shows the schematic structure of the image reading apparatus which concerns on 1st Embodiment of this invention. It is a figure which shows the correction (wiring length correction) of the variation of a pixel output value due to the difference in the wiring length of each pixel in comparison with the conventional moving average correction in the image sensor of a three-stage staggered structure. It is a figure which shows the wiring length correction when the input light amount is changed in the image sensor of a three-stage staggered structure. It is a figure which shows the approximate curve (approximate function) which shows the relationship between the pixel output value and the correction value when the input light amount is changed. It is a figure which shows the approximate curve (approximate function) which shows the relationship between the accumulation time and the correction value ratio. It is a figure which shows the schematic structure of the image reading apparatus which concerns on 2nd Embodiment of this invention. It is a figure explaining the correction of the variation of the pixel output value due to the difference in the wiring length of each pixel when the input light is different for each pixel in the image sensor of a three-stage staggered structure. It is a flowchart which shows the outline of the image reading method which concerns on one Embodiment of this invention (at the time of correction data acquisition). It is a flowchart which shows the outline of the image reading method which concerns on one Embodiment of this invention (at the time of correction data acquisition). It is a flowchart which shows the outline of the image reading method which concerns on one Embodiment of this invention (at the time of measurement).

Hereinafter, embodiments of the present invention will be described with reference to the drawings.

(First Embodiment)
FIG. 5 is a diagram showing a schematic configuration of an image reading device 1 according to the first embodiment of the present invention. As shown in FIG. 5, the image reading device 1 is a device that reads image data of the object 4 to be measured from the image sensor 10, a white light source 2 used at the time of acquiring correction data, a light source 3 used at the time of measurement, and an image. It includes a sensor 10, a measurement data acquisition unit 20, a correction data acquisition unit 30, and a device control unit 40. A white light source 2 may be used as the light source 3.

(Image sensor)
As shown in FIG. 1, in general, the image sensor 10 has a pixel 12 composed of a photodiode 13 that converts light into an electric charge, a charge amplifier 14 that stores the converted electric charge, and an electric charge accumulated from a plurality of pixels 12. The shift register 15 for reading the above and the shift register reading circuit 16 are provided. As described above, the image sensor 10 may have various configurations such as a staggered arrangement consisting of odd-numbered and even-numbered pixels, a staggered arrangement consisting of a plurality of pixels (for example, three stages), and a composite arrangement thereof. It is assumed that the image sensor 10A has such a three-stage staggered configuration.

(Measurement data acquisition unit)
The measurement data acquisition unit 20 reads an analog voltage signal corresponding to the accumulated charge amount of each pixel 12 output from the shift register reading circuit 16 of the image sensor 10, and switches between the ADC (analog digital converter) 21 and the ADC (analog digital converter) 21. It includes a switch 22, a correction unit 23, and a correction data storage unit 24.

The ADC 21 is adapted to acquire a digital pixel output value by analog-digital conversion of an electric signal obtained from each pixel 12. The changeover switch 22 switches the output destination of the ADC 21 between the correction unit 23 and the correction data acquisition unit 30 under the control of the device control unit 40. Specifically, when the correction data is acquired, the output destination of the ADC 21 is set to the correction data acquisition unit 30 by the changeover switch 22, and when the image of the object 4 is measured, the output destination of the ADC 21 is set to the correction unit 23 by the changeover switch 22. Set.

The correction data storage unit 24 stores the correction data used by the correction unit 23 to correct the inter-pixel variation of the pixel output value. The correction unit 23 corrects the inter-pixel variation of the pixel output value based on the correction data.

(Correction data acquisition unit)
The correction data acquisition unit 30 is adapted to acquire correction data used for correcting variations in pixel output values between pixels, and includes a light quantity characteristic acquisition unit 38 and an accumulation time characteristic acquisition unit 39.

The light amount characteristic acquisition unit 38 acquires the light amount characteristic that defines the change in the variation between pixels of the pixel output value due to the change in the input light amount to each pixel 12 as correction data, and the first measurement value storage unit 31. And a first correction value calculation unit 32 and a first approximation function determination unit 33.

The first measured value storage unit 31 stores the pixel output value of each pixel 12 acquired each time the input light amount is changed as the first measured value. Hereinafter, the first measured value may be simply referred to as a measured value.

The first correction value calculation unit 32 calculates the first correction value for correcting the inter-pixel variation of the pixel output value based on the first measurement value stored in the first measurement value storage unit 31. Hereinafter, the first correction value may be simply referred to as a correction value.

The first approximation function determination unit 33 determines an approximation function (approximate curve) representing the relationship between the first correction value and the pixel output value depending on the amount of input light.

The accumulation time characteristic acquisition unit 39 acquires, as correction data, the accumulation time characteristic that defines the change in the variation between pixels of the pixel output value due to the change in the charge accumulation time in each pixel 12. It includes a value storage unit 34, a second correction value calculation unit 35, a correction value ratio calculation unit 36, and a second approximation function determination unit 37.

The second measured value storage unit 34 stores the pixel output value of each pixel 12 acquired each time the storage time is changed as the second measured value. Hereinafter, the second measured value may be simply referred to as a measured value.

The second correction value calculation unit 35 calculates the second correction value for correcting the inter-pixel variation of the pixel output value based on the second measurement value stored in the second measurement value storage unit 34. Hereinafter, the second correction value may be simply referred to as a correction value.

The correction value ratio calculation unit 36 is adapted to calculate the ratio of the second correction value obtained at each storage time to the second correction value obtained at a specific storage time.

The second approximate function determination unit 37 determines an approximate function (approximate curve) representing the relationship between the ratio of the second correction value calculated by the correction value ratio calculation unit 36 and the accumulation time.

The image reading device 1 may be configured to include only one of the light quantity characteristic acquisition unit 38 and the accumulation time characteristic acquisition unit 39. When the image reading device 1 includes only the accumulation time characteristic acquisition unit 39, the accumulation time characteristic acquisition unit 39 does not have to include the correction value ratio calculation unit 36, and the second approximation function determination unit 37 is the second. An approximate function (approximate curve) representing the relationship between the correction value and the accumulation time may be determined.

(Device control unit)
The device control unit 40 controls the image reading device 1, and includes a control unit 41, a light amount setting unit 42, and a storage time setting unit 43.

The control unit 41 controls the switching of the changeover switch 22 and adjusts the amount of light input to each pixel 12 and the charge accumulation time in each pixel 12.

The light amount setting unit 42 sets (adjusts) the input light amount to each pixel 12 by adjusting the output light amount of the white light source 2 at the time of acquiring the correction data. The storage time setting unit 43 sets (adjusts) the charge storage time in each pixel 12 by adjusting the read timing of the shift register read circuit 16 of the image sensor 10.

A part or all of the measurement data acquisition unit 20, the correction data acquisition unit 30, and the device control unit 40 may be, for example, a CPU (Central Processing Unit), a ROM (Read Only Memory), a RAM (Random Access Memory), and input / output. It may be configured by using a computer having a storage device such as an interface and a hard disk, and a part or all of its functions are read from a ROM or various processing programs stored in the storage device into a RAM and executed by a CPU. Can be realized by.

Next, the operation of the image reading device 1 according to the present embodiment will be described.
12 to 14 are flowcharts showing an outline of the image reading method according to the present embodiment. In the following, first, the operation at the time of acquiring the correction data will be described, and then the operation at the time of measurement will be described.

[At the time of correction data acquisition]
(Acquisition of light intensity characteristics)
First, the white light source 2 is installed so that the input light amount is the same for all the pixels 12 of the image sensor 10A (step S1).

The light amount setting unit 42 sets the output light amount of the white light source 2 to a predetermined light amount under the control of the control unit 41 (step S2). The output light amount of the white light source 2 may be set manually.

Under the control of the control unit 41, the storage time setting unit 43 sets the charge storage time in each pixel 12 of the image sensor 10A to a predetermined value (step S3). Specifically, the storage time setting unit 43 sets the storage time to, for example, 1 ms by adjusting the read cycle of the shift register read circuit 16 of the image sensor 10A.

The changeover switch 22 switches the output destination of the ADC 21 to the correction data acquisition unit 30 under the control of the control unit 41 (step S4).

The ADC 21 converts the output voltage of each pixel 12 ₁ , 12 ₂ , ..., ₁₂₂₅₆ read by the shift register read circuit 16 into analog-digital conversion, and digital pixel output values P ₁ , P ₂ , ... _-Obtain P256 (step S5).

The first measured value storage unit 31 uses the pixel output values P ₁ , P ₂ , ..., P ₂₅₆ of each pixel 12 ₁ , 12 ₂ , ..., 12 ₂₅₆ acquired by the ADC 21 as the first measured value. Store (step S6) (see the measured value column in Table 1 below).

Next, the control unit 41 determines whether or not the pixel output value of each pixel 12 has been acquired for all the preset input light amounts (step S7).

If NO in step S7, the light amount setting unit 42 adjusts the white light source 2 under the control of the control unit 41, changes the input light amount to each pixel 12 (step S8), and returns to step S5. Since the pixel output value P of each pixel 12 is proportional to the amount of input light to each pixel 12, the light amount setting unit 42 may set the input light amount using the maximum value of the pixel output value P as an index. In Table 1, the pixel output value P of each pixel 12 is acquired as the first measured value for the three cases in which the input light amount is set so that the maximum pixel output value Pmax on the odd-numbered side is 1000, 500, 250. ..

If YES in step S7, the first correction value calculation unit 32 is based on the pixel output value P stored in the first measurement value storage unit 31, between pixels in the pixel output value P caused by the wiring length difference. The first correction value C1 for correcting the variation is calculated (step S9). Specifically, as shown in FIG. 6, the first correction value calculation unit 32 calculates the difference (Pmax-P) of the pixel output value P from the maximum pixel output value Pmax as the first correction value C1 (Table). Refer to the column of correction value (measured value) of 1). FIG. 7 is a diagram showing wiring length correction when the input light amount is changed in an image sensor having a three-stage staggered configuration, and as the input light amount increases, the difference (Pmax-P) from the maximum pixel output value Pmax is used. It shows how the first correction value C1 becomes large.

Next, the first approximation function determination unit 33 defines the relationship between the first correction value C1 calculated by the first correction value calculation unit 32 and the pixel output value P depending on the input light amount for each pixel 12. Is determined (step S10). Specifically, the first approximation function determination unit 33 is an Nth-order function (first-order function, second-order function, 2nd-order function) using the minimum square method or the like by interpolation or extrapolation from data of two or more points on the PC1 plane. Determines an approximate function represented by an order function, a cubic function, etc.) or a spline function. For example, as shown in FIG. 8, when approximating by a quadratic function (y = ax ² + bx + c), the coefficients a, b, c are obtained by the least squares method for each pixel (correction coefficients a, b in Table 1). , See column c). The first approximation function determination unit 33 sends the correction coefficients a, b, and c to the correction data storage unit 24.

The correction data storage unit 24 uses the values of the coefficients a, b, and c indicating the approximation function determined by the first approximation function determination unit 33 for each pixel 12 from the first approximation function determination unit 33 as correction data to be used at the time of correction. Receive and store (step S11). The same applies to the even-numbered side.

(Acquisition of accumulation time characteristics)
Next, the light amount setting unit 42 sets the input light amount to each pixel 12 to a predetermined light amount by adjusting the output light amount of the white light source 2 under the control of the control unit 41 (step S21). In the example shown in Table 2, the light amount setting unit 42 adjusts the output light amount of the white light source 2 so that the maximum pixel output value Pmax becomes 1000. The output light amount of the white light source 2 may be set manually.

Under the control of the control unit 41, the storage time setting unit 43 sets the charge storage time in each pixel 12 of the image sensor 10A to a predetermined value (step S22). In the example shown in Table 2, the storage time setting unit 43 adjusts the read cycle of the shift register read circuit 16 so that the storage time is either 1 ms, 2 ms, or 4 ms.

The ADC 21 converts the output voltage of each pixel 12 ₁ , 12 ₂ , ..., ₁₂₂₅₆ read by the shift register read circuit 16 into analog-digital conversion, and digital pixel output values P ₁ , P ₂ , ... _-Obtain P256 (step S23).

The second measured value storage unit 34 uses the pixel output values P ₁ , P ₂ , ..., P ₂₅₆ of each pixel 12 ₁ , 12 ₂ , ..., 12 ₂₅₆ acquired by the ADC 21 as the second measured value. Store (step S24) (see the measured value column in Table 2 below).

Next, the control unit 41 determines whether or not the pixel output value of each pixel has been acquired for all the preset accumulation times (for example, 1 ms, 2 ms, 4 ms) (step S25).

If NO in step S25, the storage time setting unit 43 adjusts the shift register reading circuit 16 under the control of the control unit 41 to change the charge storage time in each pixel 12 (step S26). Then, the light amount setting unit 42 adjusts the input light amount so that the maximum pixel output value Pmax becomes constant regardless of the change in the accumulation time (step S28), and returns to step S23.

If YES in step S25, the second correction value calculation unit 35 is based on the pixel output value P stored in the second measurement value storage unit 34, and is generated between pixels in the pixel output value P caused by the wiring length difference. The second correction value C2 for correcting the variation is calculated (step S27). Specifically, the second correction value calculation unit 35 calculates the difference (Pmax-P) of the pixel output value P from the maximum pixel output value Pmax as the second correction value C2 (correction value (measured value) in Table 2). )).

The correction value ratio calculation unit 36 calculates the correction value ratio R of the second correction value obtained at each storage time with respect to the second correction value obtained at the reference storage time (1 ms in this example) for each pixel 12. (Step S29) (see the column of correction value ratio (measured value) in Table 2).

Next, the second approximation function determination unit 37 determines an approximation function representing the relationship between the correction value ratio R calculated by the correction value ratio calculation unit 36 and the accumulation time T for each pixel 12 (step S30). Specifically, the second approximate function determination unit 37 uses the least square method or the like from data of two or more points on the TR plane by interpolation or extrapolation to obtain an Nth-order function (first-order function, second-order function, 2). Determines an approximate function represented by an order function, a cubic function, etc.) or a spline function. For example, as shown in FIG. 9, when approximating by y = a'(1 / x) ² , the coefficient a'is obtained by the least squares method for each pixel. The second approximation function determination unit 37 sends the value of the correction coefficient a'to the correction data storage unit 24.

The correction data storage unit 24 receives and stores the value of the coefficient a'indicating the approximation function determined by the second approximation function determination unit 37 for each pixel 12 from the second approximation function determination unit 37 as correction data used at the time of correction. (Step S31). Execute both odd and even sides.

[At the time of measurement]
Next, the operation of the image reading device 1 at the time of measurement will be described.
First, the light source 3 is set at a predetermined position (step S41), and the object to be measured 4 is set at a predetermined position (step S42).

The storage time setting unit 43 sets the storage time by adjusting the read cycle of the shift register read circuit 16 under the control of the control unit 41 (step S43).

The changeover switch 22 switches the output destination of the ADC 21 to the correction unit 23 under the control of the control unit 41 (step S44).

The ADC 21 converts the output voltage of each pixel 12 ₁ , 12 ₂ , ..., ₁₂₂₅₆ read by the shift register read circuit 16 into analog-digital conversion, and digital pixel output values P ₁ , P ₂ , ... _-Obtain P256 (step S45).

Next, the correction unit 23 corrects the first correction value C1, which is the difference (= Pmax−P) between the pixel output value P and the maximum pixel output value Pmax, for each pixel 12 based on the light amount characteristic. Acquire the completed correction value C3 (step S46). Specifically, the correction value C3 corresponding to the pixel output value P is calculated based on the approximation function of each pixel 12. For example, when the approximation function is represented by y = ax ² + bx + c, the correction value C3 for which the amount of light has been corrected for the pixel output value P can be calculated by C3 = a × P ² + b × P + c. When considering only the light quantity characteristic, the correction unit 23 may output the value of P + C3 as the corrected pixel output value Pc of each pixel 12.

Next, the correction unit 23 corrects the correction value C3 for which the amount of light has been corrected based on the accumulation time characteristic, and acquires the correction value C4 for which the accumulation time has been corrected (step S47). Specifically, the correction value ratio R corresponding to the accumulation time T is calculated based on the approximate function of each pixel 12. For example, when the approximation function is represented by y = a'(1 / x) ² , the correction value ratio R with respect to the accumulation time T can be calculated by R = a'(1 / T) ² . The correction value C4 for which the accumulation time has been corrected is calculated by C4 = C3 × R.

Then, the correction unit 23 acquires the corrected pixel output value Pc (= P + C4) by adding the correction value C4 for which the accumulation time of each pixel 12 has been corrected to the pixel output value P of each pixel 12. , Output as measurement data.

Next, the action and effect of this embodiment will be described.

In the image reading device 1 of the present embodiment, the light amount characteristic acquisition unit 38 acquires the light amount characteristic as correction data, which defines the change in the variation between pixels of the pixel output value due to the change in the input light amount to each pixel 12, and is the correction unit. Since 23 corrects the variation between pixels of the pixel output value based on the correction data, it is possible to accurately correct the variation of the pixel output value for each pixel even if the amount of input light to each pixel 12 fluctuates.

In the image reading device 1 of the present embodiment, the accumulation time characteristic acquisition unit 39 acquires the accumulation time characteristic as correction data, which defines the change in the pixel output value between pixels due to the change in the charge accumulation time in each pixel 12. Then, since the correction unit 23 corrects the inter-pixel variation of the pixel output value based on the correction data, the variation of the pixel output value for each pixel is accurately corrected even if the charge accumulation time in each pixel 12 fluctuates. can do.

In the image reading device 1 of the present embodiment, the first approximation function determination unit 33 determines an approximation function representing the relationship between the first correction value and the pixel output value depending on the amount of input light, and the correction unit 23 determines the approximation function. Since the variation between pixels of the pixel output value is corrected based on the function, even if the measurement data related to the first measurement value is small, the variation of the pixel output value for each pixel can be accurately adjusted in consideration of the fluctuation of the input light amount to each pixel. It can be corrected.

In the image reading device 1 of the present embodiment, the second approximation function determination unit 37 determines an approximation function representing the relationship between the ratio of the second correction value and the accumulation time, and the correction unit 23 determines the pixels based on the approximation function. Since the variation between pixels of the output value is corrected, even if the measurement data related to the second measurement value is small, the variation of the pixel output value for each pixel is accurately corrected in consideration of the fluctuation of the charge accumulation time in each pixel. be able to.

(Second embodiment)
Next, the image reading device 1A according to the second embodiment of the present invention will be described with reference to the drawings.

The second embodiment is different from the first embodiment in that the spectroscopic unit 5 is provided in the front stage of the image sensor 10A, in the rear stage of the white light source 2, or in the rear stage of the light source 3 and the object 4 to be measured. Other configurations are the same for both. The same components as those in the first embodiment are designated by the same reference numerals, and detailed description thereof will be omitted as appropriate.

FIG. 10 is a diagram showing a schematic configuration of an image reading device 1A according to a second embodiment of the present invention. As shown in FIG. 10, the image reading device 1A is a device that reads image data of the object 4 to be measured from the image sensor 10A, and has a white light source 2 used at the time of acquiring correction data, a light source 3 used at the time of measurement, and spectroscopy. It includes a unit 5, an image sensor 10, a measurement data acquisition unit 20, a correction data acquisition unit 30, and a device control unit 40. A white light source 2 may be used as the light source 3. The image reading device 1A according to the present embodiment can be suitably used for spectroscopic analysis.

At the time of acquisition of the correction data, the light emitted from the white light source 2 passes through the spectroscopic unit 5 and enters the image sensor 10A, and the ADC 21 reads out the pixels 12 ₁ , 12 ₂ , ... , 12 ₂₅₆ is converted into analog-to-digital to obtain digital pixel output values P ₁ , P ₂ , ..., P ₂₅₆ . Therefore, the pixel output value includes the wavelength characteristic of the white light source 2 and the variation for each pixel of the image sensor 10A.

The operation of the image reading device 1A according to the second embodiment may be the same as the operation of the image reading device 1 according to the first embodiment, and the image reading method using the image reading device 1A is described above. It may be the same as the image reading method of the first embodiment described.

The image reading device 1A according to the present embodiment has a configuration in which the spectroscopic unit 5 is further provided in front of the image sensor 10A, so that even if the input light amount and the storage time fluctuate, the variation in the pixel output value for each pixel is accurately corrected. Since the light reflected from the object 4 to be measured or the light transmitted through the object 4 to be measured is separated by the spectroscopic unit 5 and incident on the image sensor 10A, highly accurate spectroscopic measurement is possible.

<Modification 1>
Next, modification 1 of the first and second embodiments will be described.

The first correction value calculation unit 32 determines an approximation function for the pixel output value of the pixel group having the shortest wiring length, and corresponds to each pixel, the maximum pixel output value obtained by the approximation function and the pixel output of each pixel. The configuration may be such that the difference from the value is calculated as the first correction value.

Specifically, as shown in FIG. 11, the pixels 12 are grouped according to the structure of the image sensor 10A based on the wiring length 7. For example, as shown in FIG. 2, in the case of a three-stage staggered structure, pixel numbers 1, 7, 13, 19 ... 9, 15, 21 ... Are groups with intermediate wiring lengths (that is, pixel output values are intermediate), and pixel numbers 5, 11, 17, 23 ... Are groups with short wiring lengths (that is, pixel output values are large). ) Classified into groups. Do the same for the even side.

Then, the first correction value calculation unit 32 first defines the pixel output values of the pixel groups 5, 11, 17, 23 ... Having the shortest wiring length on the odd-numbered side by the pixel numbers and the pixel output values. The approximation function f is determined on the plane. Specifically, the first correction value calculation unit 32 uses the least squares method or the like from data of two or more points on the plane by interpolation or extrapolation to obtain an Nth-order function (first-order function or second-order function). 3. Determine the approximation function f represented by a cubic function ...) or a spline function. Then, the first correction value calculation unit 32 calculates the difference between the maximum pixel output value (interpolation value) obtained by the approximation function f corresponding to each pixel 12 and the pixel output value of each pixel as the first correction value. do. Do the same for the even side.

With this configuration, the first correction value can be calculated accurately, especially when the group can be grouped according to the wiring length such as in a multi-stage staggered arrangement. Therefore, the pixel output value between pixels can be calculated in consideration of the input light amount and the accumulation time. The correction value of the variation can be calculated with high accuracy.

<Modification 2>
Next, the second modification of the first and second embodiments will be described.

In the first and second embodiments, the first approximation function determination unit 33 determines the approximation function representing the relationship between the "pixel output value" and the "first correction value" by using the least squares method or the like. The approximation function can be determined using the square proportional relationship as described below.

Since the first correction value is proportional to the square of the current generated by the photodiode 13, it is based on the pixel output value (reference measurement value) Pref _mes and its first correction value (reference correction value) Corr _ref at the reference light amount. In addition, the first correction value Corr of each pixel output value (measured value to be corrected) P _mes can be calculated by using the square proportional formula shown by the following formula.

Square proportional formula:
Pref _mes ² : Corr _ref = P _mes ² : Corr ... (1)
Here, Pref _mes is a reference measurement value, Corr _ref is a reference correction value, P _mes is a measurement value to be corrected, and Corr is a first correction value of a pixel output value to be corrected.

From the above equation (1), the following equation can be derived.
Corr = (P _mes ² / Pref _mes ² ) × Corr _ref ... (2)

That is, the first approximate function determination unit 33 determines the approximate function represented by the above equation (2), and sends the information of the approximate function to the correction data storage unit 24. The correction unit 23 obtains a correction value of the pixel output value of each pixel 12 based on the information of the approximation function stored in the correction data storage unit 24.

For example, from Table 3, in the light amount setting in which the maximum pixel output value of a pixel is 1000, the measured value of the pixel output of pixel number 1 (referred to as “reference measurement value”) is 968, and the maximum pixel output value of the pixel is 968. In the light amount setting of 500, the measured value (referred to as “corrected measured value”) of the pixel output of pixel number 1 is 492. In the light amount setting in which the maximum pixel output value of the pixel is 1000, the first correction value (referred to as “reference correction value”) of the pixel output of pixel number 1 is 1000-968 = 32.

That is,
Reference measured value Pref _mes = 968
Reference correction value Corr _ref = 32
Measured value to be corrected P _mes = 492
In this case, the first correction value Corr is calculated from Eq. (2).
Corr = (492 ^{2/968 2 )} x 32 = 8.266648
(Refer to the column of correction value (calculated value) in Table 2 below).

Similarly, in the light amount setting in which the maximum pixel output value of the pixel is 250, the measured value of the pixel output of pixel number 1 (referred to as “corrected measured value”) is 248. The first correction value Corr in this case is calculated from the equation (2).
Corr = (248 ^{2/968 2 )} x 32 = 2.100403
(Refer to the column of correction value (calculated value) in Table 2 below).

For the pixel numbers 2 to 256, the correction value of the pixel output value can be obtained by performing the same calculation as described above.

As described above, the approximation function can be easily calculated without using, for example, the least squares method, by the configuration in which the first approximation function determination unit 33 determines the approximation function represented by the above equation (2). ..

In the description of the first and second embodiments, the image sensors 10 and 10A are one-dimensional line sensors in which the photodiode 13 is linearly arranged, but the image sensors 10 and 10A are not limited to the one-dimensional area sensor. It may be a two-dimensional area sensor. Further, the present invention can be applied to all devices using an image sensor, and can be applied to, for example, an image measuring device (image scanner, CCD camera, etc.) and a spectroscopic measuring device (spectrometer, spectro camera, etc.). ..

Further, the object 4 to be measured and the image sensors 10 and 10A may be measured in a fixed state, or may be measured while moving one of them.

As described above, the present invention has the effect of being able to accurately correct the variation of the pixel output value for each pixel even if the input light amount and the storage time fluctuate, and is used in the image reading device and the image reading method. It is useful in general.

1, 1A image reader 2 white light source 3 light source 4 object to be measured 5 spectroscopic unit 10, 10A image sensor 11 photodiode array 12 pixels 13 photodiode 14 charge amplifier 15 shift register 16 shift register read circuit 17 wiring 20 measurement data acquisition unit 21 ADC (Analog-to-Digital Converter)
22 Changeover switch 23 Correction unit 30 Correction data acquisition unit 31 First measurement value storage unit 32 First correction value calculation unit 33 First approximation function determination unit 34 Second measurement value storage unit 35 Second correction value calculation unit 36 Correction value ratio Calculation unit 37 Second approximate function determination unit 38 Light quantity characteristic acquisition unit 39 Accumulation time characteristic acquisition unit 40 Device control unit 41 Control unit 42 Light quantity setting unit 43 Accumulation time setting unit

Claims (7)

An image sensor (10) having a plurality of pixels and
An analog-to-digital converter (21) that acquires a digital pixel output value by analog-digital conversion of an electric signal obtained from each of the pixels.
A correction data acquisition unit (30) for acquiring correction data used for correcting variations in pixel output values between pixels, and a correction data acquisition unit (30).
A white light source (2) that outputs light used when the correction data is acquired by the correction data acquisition unit, and a white light source (2).
A light amount setting unit (42) for adjusting the output light amount of the white light source, and
A correction unit (23) for correcting the inter-pixel variation of the pixel output value based on the correction data is provided.
The correction data acquisition unit includes a light amount characteristic acquisition unit (38) that acquires the light amount characteristic as the correction data, which defines the change in the variation between the pixels of the pixel output value due to the change in the input light amount to each pixel.
The light quantity characteristic acquisition unit is
A first measured value storage unit (31) that stores the pixel output value of each pixel acquired each time the input light amount is changed as the first measured value.
The first correction value calculation unit (32) for calculating the first correction value for correcting the inter-pixel variation of the pixel output value based on the first measurement value stored in the first measurement value storage unit.
A first approximation function determination unit (33) for determining an approximation function representing the relationship between the first correction value and the pixel output value, and
Equipped with
The first correction value calculation unit is an image reading device that calculates the difference between the maximum value of the pixel output value and the pixel output value of each pixel as the first correction value. Each pixel has a photodiode that converts light into an electric charge and a charge amplifier that stores the converted electric charge.
The first correction value calculation unit determines an approximate function for the pixel output value of the pixel group having the shortest wiring length between the photodiode and the charge amplifier, and corresponds to each pixel. The image reading device according to claim 1, wherein the difference between the maximum pixel output value obtained by the approximation function and the pixel output value of each pixel is calculated as the first correction value. The approximate function determined by the first approximate function determination unit is
Corr = (Pmes2 / Prefmes2) × Corrref
Here, Prefmes is a reference pixel output value, Corrref is a reference correction value, Pmes is a pixel output value to be corrected, and Corr is the first correction value of the pixel output value to be corrected, claim 1 or claim. 2. The image reading device according to 2. The correction data acquisition unit
Accumulation time characteristic acquisition unit (39) that acquires the accumulation time characteristic that defines the change in the variation between the pixels of the pixel output value due to the change in the charge accumulation time in each pixel as the correction data.
The image reading device according to any one of claims 1-3, further comprising. A storage time setting unit (43) for setting the charge storage time in each pixel is further provided.
The accumulation time characteristic acquisition unit is
A second measured value storage unit (34) that stores the pixel output value of each pixel acquired each time the accumulation time is changed as a second measured value.
A second correction value calculation unit (35) that calculates a second correction value for correcting the inter-pixel variation of the pixel output value based on the second measurement value stored in the second measurement value storage unit.
A correction value ratio calculation unit (36) that calculates the ratio of the second correction value obtained at each storage time to the second correction value obtained at the specific storage time.
The image reading device according to claim 4, further comprising a second approximate function determination unit (37) for determining an approximate function representing the relationship between the ratio of the second correction value and the accumulation time. The image reading device according to any one of claims 1 to 5, further comprising a spectroscopic unit (5) in front of the image sensor. An image reading method for reading image data from an image sensor (10) having a plurality of pixels.
An analog-to-digital conversion step of acquiring a digital pixel output value by analog-digital conversion of an electric signal obtained from each of the pixels by an analog-to-digital converter (21).
The correction data acquisition step of acquiring the correction data used when correcting the inter-pixel variation of the pixel output value, and the correction data acquisition step.
An optical output step that outputs light used when acquiring the correction data by the correction data acquisition step from a white light source, and
The light amount setting step for adjusting the output light amount of the white light source, and
A correction step for correcting the inter-pixel variation of the pixel output value based on the correction data is included.
The correction data acquisition step is
A light quantity characteristic acquisition step of acquiring the light quantity characteristic that defines the change in the variation between the pixels of the pixel output value due to the change in the input light quantity to each pixel as the correction data.
The accumulation time characteristic acquisition step of acquiring the accumulation time characteristic as the correction data, which defines the change in the variation between the pixels of the pixel output value due to the change in the charge accumulation time in each pixel, and the step of acquiring the accumulation time characteristic.
Including
The light quantity characteristic acquisition step is
A first measured value storage step of storing the pixel output value of each pixel acquired each time the input light amount is changed as a first measured value, and
The first correction value calculation step for calculating the first correction value for correcting the inter-pixel variation of the pixel output value based on the first measurement value stored in the first measurement value storage step .
A first approximation function determination step for determining an approximation function representing the relationship between the first correction value and the pixel output value, and
Including
The first correction value calculation step is an image reading method in which the difference between the maximum value of the pixel output value and the pixel output value of each pixel is calculated as the first correction value.

Images