JPS60189372A - Picture input device - Google Patents

Abstract

PURPOSE:To improve an accuracy for correcting a picture, and an operating speed by bringing a picture signal inputted from each cell of an array sensor and logarithmically converted by eliminating a dark time component, to an addition/subtraction processing by a white level reference signal read out at every corresponding cell. CONSTITUTION:An output signal (vi) of a camera 1 provided with a CCD sensor is supplied to an A/D converter 3 and an adding and subtracting circuit 4 through a changeover switch (S)2. As for a dark time component, a position of each cell is made an address and the corresponding component is recorded in a memory 5. The circuit 4 eliminates the dark time component from the signal (vi) of the corresponding cell and inputs it to a logarithmic converting circuit 7. An output of the circuit 7 is connected to an adding and subtracting circuit 10 and 9 through an S8. A white level reference signal read out at every corresponding cell is stored in a memory 12 after a non-varied component is eliminated in the circuit 9. The circuit 10 corrects a sensitivity correcting value DELTACi and outputs a signal logR depending on onlyh a reflection factor or a transmission factor of an original. In this way, an accuracy for correcting a picture and an operating speed can be improved.

Description

Detailed Description of the Invention (Industrial Field of Application) The present invention is capable of producing a grayscale image with a wide range of densities, such as a color film, by arranging a plurality of photoelectric conversion elements (cells) in a linear or planar manner. The present invention relates to an image input device that can read images with rich gradations and with high fidelity using an array sensor made up of the following methods.

(Prior Art) When reading an original image using an array sensor, there are variations in the photoelectric conversion characteristics of individual cells, so it is necessary to correct them.

However, in conventional image input devices, for example,
As shown in the specification of No. 19566, data related to the light-receiving sensitivity of each cell is stored in advance, and when actually reading an image, the image signal from each cell is integrated with the signal related to the light-receiving sensitivity. , the input image was modified.

This correction method determines the correction coefficient by integrating the lighting conditions and the conversion characteristics of the cells, so under the same high-intensity lighting conditions as when the correction data was stored, the unevenness of the light receiving sensitivity of individual cells can be reduced. However, when reading grayscale images with a wide range of densities, such as color film, there are variations in the dark current of each sensor cell, so it is difficult to just correct the sensitivity difference in each cell at high illuminance. This method has the disadvantage that accurate image reading cannot be performed, especially in dark areas of a document.

Further, integration calculation for correction requires a D/A converter and an analog multiplier or a high-speed digital multiplier.
Even when the correction accuracy is set to about 1%, for example,
The memory capacity required for multiplication with a 3-digit correction coefficient such as 1.10 or 0.95 is 7 bits (0 to 2'-1), and the circuit configuration becomes complicated and expensive, so the correction accuracy is affected. There was a limit.

(Objective of the Invention) An object of the present invention is to provide an optimal image input device for correcting sensitivity characteristics in individual cells of a photosensor.

(Structure of the Invention) The device of the present invention comprises: an array sensor that performs photoelectric conversion based on the amount of light incident on each cell; 3- a first arithmetic circuit that removes a dark component from an output signal of the array sensor; a logarithmic conversion circuit that converts the output signal of the first arithmetic circuit into a logarithmic value; and a logarithmic conversion circuit that stores a logarithmically converted reference signal based on the output signal of each cell when the white level reference incident light is input to the array sensor. and a second arithmetic circuit that corrects the logarithmically converted image signal inputted from each cell of the array sensor by addition and subtraction using a white level reference signal read out for each corresponding cell from the storage means. It is characterized by the following:

(Effects of the Invention) With this configuration, the present invention can correct the conversion characteristics of each cell of the array sensor in the dark and when the white level reference incident light is input, thereby increasing the correction accuracy of the input image. At the same time, the white level reference signal is logarithmically transformed,
By performing the correction by addition and subtraction, the computation speed and reliability of the correction can be improved, and the structure can be relatively simple, and other great effects can be achieved.

(Example) Hereinafter, an example to which the present invention is applied will be described based on the accompanying drawings.

Here, a C0D (Character Couple Device) will be described as an example of a solid-state image sensor in which a plurality of photoelectric conversion elements, that is, cells are arranged in a 71-noise pattern.

In principle, a CCD sensor is considered to output an electrical signal proportional to the integrated amount of light within one scanning period, but as shown in the characteristic diagram of V for the incident light amount e and the output signal in Fig. 4, Since the sensitivity a1 and the offset value b□ differ slightly from cell to cell, even if the amount of incident light e is the same, the output signal V□ of each cell also differs slightly from each other.

However, as shown by the dotted line, the linearity of the conversion characteristic is usually maintained, so the relationship between e and Vl can be expressed by the following equation.

vl = a1e + bi (]) 7 However, when actually inputting an image to each cell and obtaining an output signal, illuminance conditions such as the illuminance to the image original itself and the illuminance distribution, the cos40 law and vignetting peculiar to the imaging optical system, etc. When development, etc. are added to the parameters related to the amount of incident light e, the situation becomes quite complicated.

Therefore, considering the parameters specific to each cell, the incident light amount e1 corresponds to the illuminance If for illuminating the original at the position corresponding to the i-th cell S, the reflectance of the original at that position, or the transmittance Ri corresponding to the cell 1. From the position, it can be considered that the angle θ1 between a line looking into the center of the lens and the optical axis, and the light amount integration time T of the cell S are determined by the following equation as parameters.

e1=■1・R4・C084θ, -T (2) Above (1
), from equation (2), if we solve for the reflectance or transmittance R1 that we are trying to find now, v□=a1・11・R1・C084θ1・T+b.

It can be seen that 201·R1+b□(3) (where cl = a141−cos4 θi·T). That is, C1 is a value that can be called a comprehensive gamma characteristic value that takes into account illumination conditions and integration time, and the output characteristics of each cell can be determined by this value c1 and the dark output b0.

As described above, in the present invention, in order to obtain accurate image input without impairing the tone of the density image and suppressing the occurrence of unevenness in light receiving sensitivity, the variation of C1rbj in the above equation (3) is , the output signal V□ is made to be a signal that depends only on the reflectance or transmittance R, of the document.

FIG. 1 is a circuit configuration diagram showing a first embodiment of the present invention.

The camera (1) equipped with a CCD sensor has a changeover switch (
2), and its output signal v1 is connected to an A/D converter (analog/digital) converter (3) or an analog addition/subtraction circuit (4) via a changeover switch (2).
The signal is switched to the (+) input terminal of the output terminal and output.

When storing the dark component, the selector switch (2)
The output of the A/D converter (3) is connected to the memory (5), and here, the position (cell number) i of each cell Sj, which will be described later, is set to address 7- As a step, the corresponding dark time component b1 is stored.

After the dark time component is stored, the changeover switch (2) is switched to the analog addition/subtraction circuit (4) side.

The output signal of the memory (5) is sent to the addition/subtraction circuit (4) via a D/A (digital/analog) converter (6).
The image input V□ of each cell S, which is input to the (-) input terminal of the cell S and read by the camera (1) in synchronization with this, is
(+) of the addition/subtraction circuit (4) via the changeover switch (2)
) connected to the input end.

As a result, the addition/subtraction circuit (4) outputs a signal vi from which the dark component b1 has been removed from the image input v1 of the corresponding cell S.
bl is output.

The output signal of the addition/subtraction circuit (4) is connected to the changeover switch (8) via the logarithmic conversion circuit (7), and the logarithmically converted output signal log(v i -b□) is connected to the addition/subtraction circuit ( (+) input terminal of 9) or addition/subtraction circuit (10)
It is switched and output to the (+) input terminal of.

When inputting the white level reference incident light from the camera (1),
The changeover switch (8) is switched to the addition/subtraction circuit (9) side.

The addition/subtraction circuit (9) is supplied with an input signal E of a constant value corresponding to a non-fluctuation value from the (-) input terminal in order to take out only the variation of the input signal, and its output terminal is supplied with a constant value input signal E corresponding to a non-fluctuation value.
□-bl)-E is output and connected to the memory (12) via the A/D conversion circuit (11).

When the white level reference incident light is input to the camera (1), this memory (12) uses a logarithmic conversion circuit (7) based on the output signal of each cell S, and the logarithmically converted reference signal is
When the entire surface R, = 1, and substituting equation (3), log (v 1 - b 1 ) = logc I R4 =
1ogc 1 +1ogR4=logc 1 ('
, 'log 1 = O), the value obtained by subtracting the constant value E from 1logcl in the addition/subtraction circuit (9) is log cm-
The output terminal of the memory (12) is connected to the addition/subtraction circuit (lO) via the D/A converter (13).
Connected to the (-) input terminal of

The above is the preparatory work necessary for the device of the present invention to work effectively.When inputting the image information R1 for each document from the camera (1), which is the original purpose, the changeover switch (2) ) side, the changeover switch (8) is switched to the addition/subtraction circuit (10) side.

In the addition/subtraction circuit (10), the signal log(
v1-b□) and (=) constant value E to the input terminal and memory (12) output signal read out from the memory (12) and sent via the D/A converter (13)], ogcm-E As is clear from equation (3) above, the signal at the (+) input terminal is 1, cl2(v4-J)=]ogcjR4=]ogc
i +], ogR1(4), and the signal at the (-) input terminal is log cm-E+E=
]oBc1, the signal from the output end of the addition/subtraction circuit (10) is configured such that the cl component is removed and becomes logRl.

Next, the overall operation of the apparatus of this embodiment and the input image correction operation will be briefly explained.

The image input correction operation is performed according to the following three steps.

1st step: Step to memorize the dark time component. The camera (1) is provided with, for example, a shutter, and the output No. 03 under the total light-shielding condition is sent to the A/
The dark time component 1 for each cell s1 of the camera (1) is supplied to the D converter (3) and converted into digital values.

is stored in the memory (5).

In other words, since the incident light to camera (1) is zero,
As is clear from equation (1) above, the dark output component is v
1=b4, and as shown in Figure 4, the variation in this value b□ is small, so the accuracy of the A/D converter (3) is several times the required dynamic range. Depends on whether or not variations occur.

For example, if the dynamic range is +000: ] and bj is always 16 or less, the A/D converter (3) will be 4
It can be seen that a bit is sufficient, and the memory (5) only needs to have 4 bits for each sensor.

Second step: white level reference setting step (own level correction value setting step).

Input the white level reference incident light under the blank paper condition, that is, the reflectance R1l- = 1, or the document-free condition, that is, the transmittance R=] to the camera (1), and then output the photoelectrically converted output signal of each cell at that time. , is supplied to the (+) input terminal of the addition/subtraction circuit (4) by the changeover switch (2). At this time, the dark time component b□ of the cell S1 at the corresponding position is sequentially read out from the memory (5) in synchronization with the camera (]), and it is D/A converted to the (-) of the addition/subtraction circuit (4). Each output signal V of each cell S1 of the camera (1) is
1, the dark time component b1 is removed.

That is, vl-v1=cI R4+b4-bl :c1R4=c, (where Rt=1) (5) becomes the output of the addition/subtraction circuit (4), and is converted to 1. by the logarithmic conversion circuit (7). og compressed.

By setting the changeover switch (8) to the side of the addition/subtraction circuit (9), the fluctuation range of the logarithmically converted reference signal 1ogcj is estimated in advance, and a constant level E below the fluctuation range is set to the addition/subtraction circuit ( 9) Subtract (]ogc1-E
)L/, only the change ΔQ112- (=], ogcm-E) is converted into A, /D,
Stored in memory (12).

For convenience of explanation, here, we have assumed that the sensitivity correction value formula cj is obtained under the condition of R = 1, but in this case, it is assumed that the input surface of the document is uniform over the entire surface and that R = 1. It is also possible to set up a reflective original or a transparent original that is close to the original and determine the sensitivity correction value.

As described above, each cell S is processed through the first and second steps.

The dark time component bi and the variation ΔC□ of the white level reference signal can be stored in memories (5) and (+2), respectively.

These first and second steps are preparatory steps, and the next third step is a thinning process in which the original image is scanned.

Step 3: Normal use step (image input correction operation step).

When the camera (1) reads the original image, the changeover switch (2) is placed on the addition/subtraction circuit (4) side.
8) is placed on the addition/subtraction circuit (■0) side.

The image signal vj from the camera (1) is synchronized with the
The dark time component at the corresponding position is read out from the memory (5), divided by the addition/subtraction circuit (4), and then the logarithmically transformed image signal log (v□-crime) is read out from the memory (12) in synchronization therewith. The variation Δ of the reference signal from the white level of the corresponding position
ci is read out and subtracted by the addition/subtraction circuit (10) via the D/A converter (13).

The above calculation process is as follows.

log(vi-v4)-ΔC1-E=log(c,
IRl +b1-b□)-(1ogcm-E) -E=
]ogci R, -]ogc.

=logcm+]ogR4-1ogcm=1ogR4(
6) In this way, the dark time component b□ for each individual cell S1
The variations in the synergistic value C□ of each parameter corresponding to the overall gamma characteristic value are removed, and a signal that depends only on the reflectance or transmittance R□ of the original is obtained at the output 1ogR4 of the addition/subtraction circuit (]0). It will be done.

In this embodiment, the image signal read in step 3 is processed as an analog signal, and the output 1ogR
Since □ can be obtained as an analog signal of the image density signal itself, reliability is extremely high when the input image is subjected to analog processing in a subsequent device.

FIG. 2 is a circuit diagram showing a second embodiment of the present invention.

This embodiment uses an addition/subtraction circuit (14
) is digitized, and an A/D converter (15) is installed between the logarithmic conversion circuit (7) and the changeover switch (8),
The D/A converter (13) in the first embodiment is removed.

With this configuration, the D/A converter (13)
can be omitted, and when the output 1ogR1 is obtained as a digital signal and the subsequent processing device is a digital device, there is an advantage that a suitable image input device can be obtained.

FIG. 3 is a circuit configuration diagram showing a third embodiment of the present invention.

In this embodiment, the photoelectrically converted output signal of the camera (1) is immediately converted into a digital signal by the A/D converter (15), and the D/A converter (6) in the first and second embodiments I am trying to remove it.

Therefore, the signal processing after the addition/subtraction circuit (4') is 15
- Everything is digital, and the logarithmic conversion circuit (16) is also composed of a digital log conversion table.

By configuring in this way, the memory (5).

The A/D converter and D/A converter before and after (12) are not required, and the circuit configuration becomes even simpler.

In addition to the first to third embodiments described above, the present invention also includes
The memory (5) provided for removing the dark component can be implemented with various modifications, for example, the variation in the dark component b1 can be ignored compared to the variation in the overall gamma characteristic value C1. If it is small enough, it can be removed by externally inputting a signal corresponding to the average value of bl, which further simplifies the circuit configuration, and also eliminates the need for the first step as a work step. Modifications can be made depending on the accuracy.

Furthermore, in each of the embodiments described above, the memory (5).

The signal bitΔc1 stored in (12) is both.

By scanning each cell multiple times and inputting it,
If these are stored as average values, A/D converters (3), (11,), (15)
) etc. can reduce quantization errors.

[Brief explanation of the drawing]

FIG. 1 is a circuit configuration diagram showing a first embodiment of the present invention. Fig. 2 is a circuit diagram showing a second embodiment of the present invention; Fig. 3 is a circuit diagram showing a third embodiment of the invention; Fig. 4 is a diagram showing the incident light amount e and the output signal of the CCD sensor. FIG. 3 is a characteristic diagram showing the relationship with V. (1) Camera (2) Changeover switch (3) A/D converter (4) Addition/subtraction circuit (first arithmetic circuit) (5) Memory (first storage means) (6) D/A converter (7 ) Logarithmic conversion circuit (8) Changeover switch (9) Addition/subtraction circuit (1 addition/subtraction circuit (second arithmetic circuit) (11
) A/D converter (12) Memory (second storage means) (
13) D/A converter

Claims (2)

[Claims] (1) An array sensor that performs photoelectric conversion based on the amount of light incident on each cell, a first arithmetic circuit that removes dark components from the output signal of this array sensor, and a pair of output signals of this first arithmetic circuit. a logarithmic conversion circuit for converting into a numerical value, a storage means for storing a logarithmically converted reference signal based on an output signal for each cell when white level reference incident light is input to the array sensor, and each of the array sensors. The logarithmically converted image signal inputted from the cell is combined with a white level reference signal successively outputted from the storage means for each corresponding cell,
An image input device comprising: a second arithmetic circuit that performs correction by addition and subtraction. (2) an array sensor that performs photoelectric conversion based on the amount of incident light of each cell; a first storage means that stores a dark time component corresponding to each cell of the array sensor; a first arithmetic circuit for removing a dark component stored in the storage means; a logarithmic conversion circuit for converting the output signal of the first arithmetic circuit into a logarithmic value; and a white level reference incident light to the array sensor. a second storage means for storing a logarithmically transformed reference signal based on the output signal of each cell when inputting the image signal; An image input device comprising: a second arithmetic circuit that performs correction by addition and subtraction using a white level reference signal read out for each corresponding cell from a storage means.