JPS62193365A - Picture reader - Google Patents

Abstract

PURPOSE:To reduce the influence of noise by reducing the light unevenness on a film by an optical fiber and a light diffusing plate and reducing the noise component by the output of a CCD one-dimensional line sensor and the correlative double sampling method and providing a signal correcting circuit part with a means which adds and averages white-and-black correcting data. CONSTITUTION:A light diffusing plate 6 is inserted between the exit end of an optical fiber 5 and an X-ray film 7 to reduce the spatial light variance of the quantity of projected light. The light transmitted through and diffused by the X-ray film 7 passes an optical mirror 9 to focus a film image on a CCD one-dimensional line sensor 12 by a focusing lens 10 and is converted to an electric signal proportional to the intensity of light. An optical output monitor signal detected by a photodetector 18 passes an amplifier 19 and an AD converter 20 to obtain a digital signal REF. A signal correcting circuit 21 performs various corrections and logarithmic conversion with AD-converted digital signal SIG and REF to obtain correct X-ray film density information, and this information is stored in a picture memory and is outputted to a CRT display device 23 or the like as required.

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention relates to a facsimile machine and a film image reading device, and particularly to an optical system, a photodetection system, and a signal correction circuit of an image reading device using a C0D one-dimensional line sensor. Regarding the system.

[Prior art]

Reads images on paper or film in line sequential order,
Devices that convert electrical signals are widely used in facsimiles. Recently, devices that can accurately read extremely fine images with a wide dynamic range, such as medical X-ray film, are being considered. For devices that require such a high dynamic range, a configuration is used in which a laser beam is focused on the target image, one-dimensional light scanning is performed, and the surface reflected or transmitted light of the target image is converted into an electrical signal by a photodetector. is taken. in this case,
The disadvantages are that the optical system and optical scanning system are complicated, expensive, difficult to adjust, and the device is inevitably large.

On the other hand, related to image reading devices using C0D one-dimensional line sensors, for example, NEC Technical Report MoQ, 3B,
No, 1. Examples include those described in pp. 116-120 (1985).

In this device, instead of scanning light, a surface light source such as a fluorescent lamp is used as a light source, and a CCD one-dimensional line sensor is electrically scanned and read. Therefore, the mechanical drive unit only performs sub-scanning (-next feed) of the image surface (paper surface), and the optical system has a significantly simpler configuration, making maintenance and adjustment easier.

However, in devices equipped with a CCD that read and reproduce the intermediate density of an image, this non-uniformity appears as a sliver-like density unevenness in the reproduced image. It has level and white level correction means. In addition, means for controlling the power supply current of the AGC circuit fluorescent lamp is provided in order to compensate for temperature changes and temporal changes in the amount of light from the fluorescent lamp.

[Problems to be solved by the invention]However, with surface light sources such as fluorescent lamps, spatial brightness is
It cannot be absorbed by controlling the source current. On the other hand, the white level correction of the CC, D one-dimensional line sensor in the above-mentioned known example performs sensitivity correction by memorizing the output of each bit of the sensor when there is no film. It is not possible to correct the sensitivity unevenness caused by the spatial and temporal luminance fluctuations that occur in the image.

Furthermore, the dynamic range of the CCD one-dimensional line sensor is about 55 dB, which is insufficient especially for reading images of medical X-ray films.

For the above-mentioned reasons, it has heretofore been difficult to obtain an apparatus that does not require an optical scanning machine and has sufficient performance to effectively read images with high precision and high dynamic range.

SUMMARY OF THE INVENTION Accordingly, one object of the present invention is to provide an image reading device that does not require an optical scanning mechanism and that does not cause sensitivity unevenness due to spatial and temporal fluctuations in light source brightness.

Another object of the present invention is to provide an image reading device with a higher dynamic range.

[Means for Solving the Problems] A feature of the present invention is that the quantitative ends of the optical fiber bundle into which the light from the light source is input are uniformly distributed in one dimension on one surface of the light diffusing plate and coupled. It is equipped with a light illumination system that uses the other side of the light diffusing plate as the irradiation surface, and an image film is sent out to a part of this irradiation surface, and a one-dimensional image sensor detects the amount of light passing through the image film, and also measures the irradiation value. A monitoring photodetector is provided in the remaining portion of the image sensor, and the output of each bit of the one-dimensional image sensor is corrected by the output of the monitoring photodetector.

Another feature of the present invention is that the output of the one-dimensional image sensor is obtained by single-correlation double sampling.

Another feature of the present invention is that a CCD one-dimensional line sensor kept at a low temperature by a Peltz cooling means is used as the one-dimensional image sensor.

"Function" According to the above-mentioned optical illumination system, it is easy to make light of uniform illuminance enter the bundled optical fiber bundle, and since the optical fibers are homogeneous or easily obtained, the light can be directed in one dimension. It is possible to even out slight unevenness in the light intensity distribution due to uniformly dispersing the light, and it is possible to irradiate the detection position of the image film with uniform light. A monitoring photodetector is provided in the section, and the output of the one-dimensional image sensor for the line reading diagram is corrected based on the output thereof, so that it is possible to regularly compensate for temporal fluctuations in the amount of light. Thus, according to the configuration of the present invention, image reading without unevenness in sensitivity is possible.

On the other hand, regular ccop! The dynamic range of the tf-dimensional line sensor is about 55 dB due to the high noise level. Therefore, the C0D-dimensional sensor is kept at a low temperature using a Peltier cooling means to equalize dark current and reduce shot noise.

In addition, by performing double correlation sampling (CS), noise is further reduced and the dynamic range is increased to 70 to 80.
dB or more.

[Example] Here, an explanation will be given focusing on X-ray film image reading as an example.

In order to read medical X-ray film images, it is necessary to read densities of 0 to 3.2 or higher faithfully without being influenced by system noise. For this purpose, it is necessary for the image reading device to consider the following.

(1) There is little light unevenness on the film, and the light unevenness does not change over time. Even if there are fluctuations over time, it can be completely corrected.

(2) The dynamic range of the photodetection system is 3.5 to 4 orders of magnitude or more.

(3) Sensitivity unevenness in the photodetection system can be completely corrected, and no new noise or the like is introduced during each correction.

(4) The correction conditions in (3) should not change with the passage of time, and in particular should not change with temperature.

Among these items, for (1) conventional fluorescent lamps are used, and the spatial brightness unevenness varies over time. In order to correct this variation, it is necessary to install another light monitoring system that does not pass through the film and measure the temporal changes in spatial brightness unevenness. It is difficult to correct the uneven brightness of surface light sources such as fluorescent lights.

Regarding (2), 1. The dynamic range of a normal C0D one-dimensional line sensor is about 55 dB, and the above requirement is 7.
The principle of correlated double sampling (CDS) is well known as a noise reduction means for reducing the noise to 0 to 80 dB or more.

Regarding CDS, IEE, Journal of Solid State Circuits (IEE
E Journal of Solid-8tate
C1rcuits) Vol, S C-9, Na
1.

PP, 1-13 (1974). Regarding (3), it is necessary to use white correction, black correction, and light source light output correction methods that do not introduce new noise during correction, and for (4), temperature control of the photodetection system is used to reduce state fluctuations in the detection system. It is necessary to.

The present invention will be described in detail below based on the drawings.

11th! FIG. 1 is a block diagram illustrating the present invention in detail.

In the figure, 1 is a lamp light source, 2 is a DC power source for the light source, and 3 is a lamp light source.
is a lamp house whose inner surface has a high reflectance.

4 is a tapered light guide, 5 is an optical fiber guide,
6 is a light diffusion plate, 7 is X4! The film 8 is a roller for transporting the film. 9 is an optical mirror, 10 is an imaging lens, 11 is an optical shutter, 12 is a CCD one-dimensional line sensor, 13 is a CCD one-dimensional image sensor drive circuit, 14 is a Peltier cooling element, and 15 is a Peltier cooling element temperature control circuit. . 16 is a CDS @ path, and 17 is an AD converter. 18 is a light source light output monitor instep photodetector.

19 is an amplifier, 20 is an AD converter, 21 is a signal correction circuit section shown in FIG. 3, 22 is an image memory, 23 is a CRT display, 24 is a signal correction circuit control section, and 25 is an image reading control section. be. 26 in FIG. 2 is an optical shutter.

Explain the operation. Light from a lamp light source 1 is collected by a lamp house 3 and firstly guided by a tapered light guide 4 to the circularly bent input end of an optical fiber 5. A tapered light guide is inserted to efficiently guide light to the input end of the optical fiber. The output ends of the optical fibers 5 are uniformly arranged in one dimension, and the output light is guided to the light diffusing plate 6.

Light from the other side of the light diffusing plate 6 illuminates the XG film 7, and a portion of the contrast light is transmitted to the monitoring photodetector 1.
8. With such a configuration in which light is distributed using an optical fiber bundle, a fairly uniform light amount distribution can be easily obtained. However, since variations in the output angle and light transmittance of each optical fiber cannot be avoided, in this embodiment, a light diffusing plate 6 is inserted between the output end of the optical fiber 5 and the Xs film 7 to spatially change the amount of irradiated light. Light fluctuations are further reduced.

The light transmitted and diffused from the Xa film 7 passes through the optical mirror 9
Then, the film image is formed on the CCD one-dimensional line sensor 12 by the imaging lens 1o, and converted into an electric signal proportional to the light intensity. Conventional CCD one-dimensional line sensor electrical signal output has large reset noise and S/N ratio of approximately 5.
00: There is no IL, so a sufficient dynamic range cannot be obtained.

Therefore, in FIG. 1, the CDS circuit 16 is used to improve the SN. Finally, an AD converter 17 converts the analog signal into a digital signal. This digital signal is
Name it IG.

On the other hand, the optical output monitor signal detected by the photodetector 18 passes through an amplifier 19 and an AD converter 20 to obtain a digital signal REF.

The optical shutters 11 and 26 are for measuring the offset value of the photodetecting element. A when 11 is closed
The output of the D converter 17 indicates dark current non-uniformity data for each picture element of the C0D one-dimensional line sensor 14. Further, the output of the AD converter 20 when the switch 26 is closed becomes offset data of the photodetector 18.

Dark current non-uniformity changes depending on the temperature of the C0D one-dimensional line sensor, and shot noise due to dark current causes S
The N ratio also changes. By cooling the CCD element and controlling its temperature, dark current can be kept constant and shot noise can be reduced. Therefore, in the first session, the Peltier element 14. A Peltier cooling element temperature control circuit 15 is used.

The signal correction circuit 21 is a circuit that obtains correct X-ray film density information by performing various corrections and logarithmic conversion using the AD digital signals SIG and REF.The density information is stored in the image memory 22 and transferred to the CRT as necessary. Output to the display 23 or the like.

The entire system is controlled by an image reading control section 25 and a signal correction circuit 21.
This control is performed by the signal correction circuit control section 24.

Next, the signal correction circuit will be explained in detail with reference to FIG.

In FIG. 3, for the SIG signal system, 30 is an adder, 31 is a CCD element black level one-dimensional memory, and 32 is a register. 33 is a subtractor.

34 is an adder, 35 is a CCD element white level one-dimensional memory, 36 is a register, and 37 is a log conversion circuit.

38 is a white level one-dimensional log conversion memory.

40 is a log conversion circuit, 41 is a subtracter, and 42 is an adder.

Regarding the REF signal system, 5o is an adder, 51 is an offset average value memory, 52 is a register, 53 is a subtracter 54 is an adder, 55 is an optical monitor output average value memory, 56 is a register, 57 is a log conversion circuit, 58 is a light monitor log average value memory, 59 is a log conversion circuit, and 60 is a subtracter. 70 is a log conversion table, 71 is an address counter, 72 and 74 are down counters, 73 is addition scan number data for adding the black level and white level in the CCD element output, 75 is the optical monitor offset average value and the optical monitor output The data of the number of additions to obtain the average value memory, which are counted by the counters 72 and 74 at the operation start ST, respectively.
These data are set in . H8 is a scanning start signal for the CCD element, and CLK is a clock signal for driving the CCD element.

The density of an X-ray film is expressed as D=logl/T, where T is the film light transmittance. In reality, taking into account light intensity fluctuations of the light source and white level correction, the density D i K of the first pixel in the second scan of the one-dimensional scan is here, and Fwi is the CCD element output level (white) when there is no film. level), C when reading FpzK X-ray film
CD element output level, Rw is the output level of the optical output monitor when acquiring white level data, and RFIK is the output level of the optical output monitor when reading the xRIA film. It is necessary to take into consideration the fact that each picture element is different due to variations in sensitivity of the COD element itself, non-uniformity in black level, noise, and unevenness in the amount of light due to the optical system. Furthermore, the output level of the optical output monitor varies due to the offset of the photodetector and circuit system.

First, a method for correcting the black level of a COD element will be described. C.C.
Assuming that the black level value of the first picture element in the second scan of the D element output SIG is biK, and that n scans are added to reduce noise, the black level additive average value B1 of the first picture element is expressed as follows. This calculation is performed by setting the value of n in the down counter 72 and closing the optical shutter 11 in FIG.
i, SWt to prevent light from entering the CCD element 12.
t is turned ON and 30, 31, 3217) Add and average the n number of scans in the addition and averaging circuit system. As a result, black level data is obtained in a 31 black level one-dimensional memory.

The white level value including the optical system is obtained by opening the optical shutter 11 and calculating the value of the first pixel of the second scan of the 1CCD element output SIG from WiK and the black level additive average value B1 of the first pixel from WiK- It can be written as B i. In this case as well, in order to reduce noise, n scans are added as i white data FWI.

This calculation is performed by turning on SW 2 in Figure 3.
K-B i is performed by a subtracter 33, and addition for n scans is performed in the same manner as the calculation of the black level, using an addition average path using 34 and 35 degrees 36. The average value obtained here is log-converted by a 371-og converter 37 using a 701-og conversion table, and the white level value is stored in a white level one-dimensional log conversion memory 38.

For actual film data, the value of the first pixel of the second scan of the CCD element SIG can be written as fIK and the black level addition average value B1 of the i end picture element B1 to fix-B1, and this is FF1K. Turn on 5W13, input this data to the log conversion circuit 40, and log the log converted data.
Get Fpix. When this value is subtracted from the data in the white level one-dimensional log conversion memory 38 by the subtracter 41, the result of the subtraction (logFwtlogF p i ic) represents the film density value if there is no change in the light output of the light source. .

In the case of an actual lamp light source, there is a temporal change in the light output due to the passage of time between when white data is taken and when the film is being read. It is necessary to correct this when reading the film.

First, offset correction of the optical output monitor system REF will be described. If the offset value of the signal REF is averaged by ○j and i for m pixels to reduce noise, and it is set to 0, it can be expressed as . This calculation is performed by setting the value of m in the down counter 74, closing the optical shutter 26, and
Prevent light from entering the photodetector 18, and turn 5W21 to O.
In the N state, m pixels are averaged using an averaging circuit system of 50, 51 degrees and 52 degrees.

As a result, black level data is obtained in the offset average value memory 51.

Next, since the COD element signal takes white data, SW 1
Turn on SW2□ at the same timing as turning on *,
Optical shutter 2 while the COD element is searching for white data
Find the average value of the optical monitor output when 6 is open. The average value Rw of the optical monitor output after subtracting the black level by the subtracter 53 is determined by R1 being the optical monitor output and the number of additions 0 being determined by the fluctuation state of the optical monitor output. This calculation is performed by averaging Q times using a 54.55°56 averaging circuit system.

Once the average value is determined, it is converted into a log in a log conversion circuit 57 and stored in the optical monitor log average value memory.6 If the actual film data is loaded into the CCD element, S
W2. is turned on and the subtracter 53 obtains the value R, , K=R1K-0. i and k mean the first pixel in the second scan of the CCD element output SIG. This data is log-converted by the log conversion circuit 59, and finally subtracted from the output 1logRw of the 6Qlog average value memory 58 by the subtracter.
RW-logR,iK is obtained, and optical monitor output fluctuation correction data is obtained.

Finally, the output of the 60 subtracter 60, which is the optical monitor output fluctuation correction data, is subtracted by the subtracter 42 from the output of the subtracter 41, which is the film density data, to correct the fluctuation in the optical monitor output. Obtain film density data. This density data is stored in the image memory 22. If necessary, the data is displayed on the CRT display 23 after DA conversion.

For convenience of explanation, the description so far has been made with multiple components having the same function, but the adder,
The subtracter, log converter, etc. can be configured by combining those with common functions into one, and as a result, the overall configuration can be simplified.

The optical shutters 11 and 26 have been described as being installed in front of each photodetector, but by installing them at appropriate positions between the optical fiber 5 and the light source 1, the number of optical shutters can be reduced to one. It is possible. However, in this case, it is necessary to block light around the photodetector.

As for the light source, it is possible to replace it with a semiconductor laser, an LED, or other various laser devices.

Although a light diffusion plate 6 is inserted at the output end of the optical fiber for the purpose of uniformity of light intensity, this light diffusion disorder may be removed midway when uniformity is not required.

Since light sources generally rise in temperature, an infrared-cutting optical filter or ultraviolet-cutting filter is installed between the optical fiber and the light source to prevent materials such as optical fibers from deforming due to heat, and to prevent damage to organic materials due to ultraviolet rays. This can be countered by inserting an optical filter 80.

Black level data and white level data due to dark current are
When the temperature of the CCD element is controlled to a constant temperature using a Bertier element, these data do not change over time, so there is no need to take black and white data for each film. After measuring the offset data and the like of the optical monitor output system, each data is stored in the memories 31, 38, 51, and 58. In this way, it is not necessary to obtain correction data for each film, and processing time can be shortened.

In the explanation so far, we have described a one-dimensional CCD element, but it can also be applied to a two-dimensional CCD image sensor, and in this case, the configuration of the device is further simplified because a film transport system is not required. Ru.

Furthermore, although the explanation so far has been limited to film transmission images, it can also be applied to light reflection type image reading systems such as normal facsimile machines by making the light source-optical fiber system of the device configuration a reflection type. can.

[Effects of the Invention] According to the present invention, light unevenness on a film can be reduced by using optical fibers and a light diffusing plate, and spatial brightness unevenness, which has been a problem with conventional fluorescent lamps, can be reduced. or,
Luminance unevenness does not vary temporally or spatially.

The CCD one-dimensional line sensor output and correlated double sampling method reduce noise components, and by controlling the temperature of the element to a low temperature, it is possible to lower the dark current level and, as a result, reduce shot noise. By reducing the noise of the sensor output, we have achieved a wide dynamic range of the sensor.

Furthermore, the signal correction circuit section has a means for averaging the black and white correction data to reduce the influence of noise caused by problems when acquiring black and white data, and a logarithmic conversion circuit has been introduced to eliminate the conventional multiplication and division. The necessary correction calculations can be performed by simple addition and subtraction, shortening processing time and simplifying the circuit system, allowing high-quality X-ray film reading without density unevenness due to incomplete white correction to be performed using C0D one-dimensional technology. This can be achieved using an image sensor.

Furthermore, since the temperature of the C0D one-dimensional line sensor is controlled, temporal changes in white level, black level, etc. due to temperature fluctuations are small, and there is no need to perform the data acquisition process for each film, improving film reading speed. There are advantages that lead to

[Brief explanation of drawings]

FIG. 1 is a block diagram explaining the present invention in detail, FIG. 2 is a block diagram explaining a light source and an optical fiber section, and FIG. 3 is a signal correction circuit section. 1... Light source, 5... Optical fiber, 6... Diffusion plate,
11... Optical shutter, 12... CCD dimensional line sensor, 14... Peltier cooling element, 17 and 20...
- AD converter, 21... signal correction circuit section, 22...
Part of image memory, 23...CRT display? 15 no E2 ko J Fig. 2 ~/

Claims (1)

[Claims] 1. The output end of the optical fiber bundle that inputs the light from the light source is uniformly distributed and coupled in one dimension of one side of the light diffusing plate, and the output end is connected to the other side of the light diffusing plate. a light illumination system that forms an irradiation surface with a uniform amount of light; a film transport means that sends an image film to a part of the irradiation surface; a monitoring photodetector that detects the scene of the irradiation surface using the remaining portion of the irradiation surface; a one-dimensional image sensor that measures the amount of light passing through the image film; and a storage means that stores the output of the one-dimensional image sensor for each bit at a white level and a black level, and the image film is arranged on an irradiated surface. calculating an image density signal in which the sensitivity distribution of the one-dimensional image sensor is corrected using the output of each bit of the one-dimensional image sensor and the stored white level and black level data when the one-dimensional image sensor is An image reading device comprising: a signal correction circuit that further corrects a value calculated from the output of a photodetector to obtain an image density signal that compensates for fluctuations in light amount. 2. The image reading device according to claim 1, wherein the output data of the one-dimensional image sensor is obtained by correlated double sampling. 3. The image reading device according to claim 1, wherein the one-dimensional image sensor is a CCD image sensor controlled to a low temperature by a Peltier cooling means.