JP2002279414A - Signal processing method, image reading device, program and medium - Google Patents

Abstract

PROBLEM TO BE SOLVED: To accurately and quickly detect and correct positions of dirt or a flaw even if there is shifts in images, provided by two scans of visible light and invisible light when dirt or flaws of a visible light image provided by reading an original is removed. SOLUTION: A quantity of shift of an invisible light image and the visible light image is detected from an image, provided by the invisible light image and a first threshold. An optimal quantity of shift is detected, while shifting and detecting the quantity of shifts is thinned to make detection high speed. The positions of dirt or flaws are detected from the image provided by the invisible light image and a second threshold, to remove the dirt or the flaws from the visible light image.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a signal processing method, an image reading device program, and a medium, and more particularly, to a signal processing method for correcting a defective portion due to dust or scratches on a document, an image reading device, a program, and a medium. It is about.

[0002]

2. Description of the Related Art FIG. 13 shows a schematic configuration of a conventional image reading apparatus for a transparent original, for example. In FIG. 13, a transparent original 142 such as a positive or a negative placed on an original platen glass 141 is further illuminated by a transparent original illumination lamp 144 via a diffusion plate 143 provided above the original, and the transparent original 142 is illuminated. 147, the Hano mirror 1
48, sent to the CCD 150 via the imaging lens 149,
CCD with a large number of unit solid-state image sensors arranged in a line
At 150, an image in the main scanning direction is formed by converting the signal into an electric signal.

In this case, image formation in the sub-scanning direction is performed by mechanically moving the transmissive original illumination lamp 144 and the mirror 147 with respect to the transmissive original 142 in the sub-scanning direction while maintaining the same speed and the same phase. The Hano mirror 148 is caused to follow in the same direction at a scanning speed of 、, and
This process is performed while keeping the optical path length (conjugate relationship) up to D150 constant, and a two-dimensional image is formed in total together with the main scanning.

In the above-described image reading apparatus for a transparent original, a so-called reflective original of a type in which light is irradiated on an original written on an opaque sheet and light reflected from the sheet is processed. Is also possible. In that case, a reflection original is placed instead of the transmission original 142, the transmission original illumination lamp 144 is turned off, the reflection original illumination lamp 145 is turned on, and the direct light flux and the reflection shade 146 by the reflection original illumination lamp 145 are set. When the image is illuminated with the reflected light flux and the reflected light from the reflection original is read by the CCD 150, an image in the main scanning direction can be formed in the same manner as in the case of the transmission original.

In the color reading method, in particular, a lamp having white spectral characteristics is used as the reflection document illumination lamp 145, and a three-line CCD having RGB color filters is used as the CCD 150 for one scanning. There is generally known a three-line color image reading system in which image information of each color of RGB is read at the same time, and a color image is formed by superimposing signals of each color of RGB on the same line on an image processing circuit. .

In the above-described image reading apparatus for a transparent original, in order to correct a defective portion on the image due to dust or scratches on the transparent original, it is necessary to correct the retouch by image editing software after reading the image. There was no effective way. Therefore, it takes a very long time to correct a defective portion.

In recent years, in such an image reading apparatus for a transparent original, dust such as dust present on the transparent original and damage (scratch) on the film surface are detected (hereinafter, this detection is referred to as "dust / scratch detection"). )) From the scanned image,
An image reading apparatus having a function for removing so-called dust and scratches, which removes the influence of such dust and scratches by image processing, has been developed.

FIG. 14 is a view showing a conventional image reading apparatus 1 having a function of detecting dust and flaws. The same components as those of the image processing apparatus shown in FIG. Therefore, the description is omitted.

In FIG. 14, reference numeral 151 denotes a wavelength of about 880 n.
This is an infrared lamp made of an LED having a peak of emission intensity at m.

FIG. 15 is a block diagram showing a functional configuration of a dust / scratch remover 2 for removing dust / scratch using image data obtained by the image reading apparatus 1. In FIG. 15, reference numeral 21 denotes an interface (I) for inputting image data read by the image reading apparatus 1.
/ F), 22 is an image memory for storing an image (hereinafter referred to as a “normal image”) read using the transmissive original illumination lamp 144 or the reflective original illumination lamp 145, and 23 is an infrared light lamp An infrared light image memory for storing an image (hereinafter, referred to as an “infrared light image”) read using 151, a threshold holding unit 24 for holding a predetermined threshold, and a dust / scratch 25. The detection unit 26 is a dust / scratch correction unit.

FIG. 16 is a diagram showing the spectral intensity distributions of the transmissive original illumination lamp 144 and the infrared light lamp 151.
The characteristics of each lamp are shown by a solid line and an alternate long and short dash line, respectively. FIG. 17 shows spectral transmittance characteristics of cyan, yellow, and magenta dyes of general negative and positive color films, and a peak wavelength (about 880 mm) of a spectral intensity distribution of the infrared light lamp 151. It is a thing. As is clear from FIG. 17, in the case of a general color film, the transmittance at about 880 nm of any dye is very high, so that the luminous flux of the infrared lamp is independent of the image on the film. Almost through.

The operation of reading a transparent original when the dust / scratch removal operation is performed will now be described in detail with reference to the flowchart shown in FIG.

First, in step S10, the reflection document illumination lamp 145 and the infrared light lamp 151 shown in FIG. 14 are turned off, and the transmission document illumination lamp 144 is turned on. At this time, the illuminating light flux of the transmissive original illumination lamp 144 is diffused by the diffusion plate 143 without unevenness, and the diffused luminous flux transmits through the transmissive original 142. The transmitted light flux passes through the mirror 147 and the Hano mirror 148, further passes through the imaging lens 149, and is projected on the CCD 150. The image projected on the CCD 150 is converted into an electric signal, and the I / F of FIG.
The image data is temporarily stored in the image memory 22 through the memory 21. Here, when the transparent original is a negative film, a positive image is obtained by a reversal process and then temporarily stored in the image memory 22.

Next, in step S20, the lamp 145 for illuminating the reflection original and the lamp 1 for illuminating the transmission original shown in FIG.
44 and the infrared light lamp 151 is turned on. The illumination light beam of the infrared light lamp 151 having the characteristics shown in FIG. 16 is diffused without unevenness by the diffusion plate 143, and the diffused light beam passes through the transmission original 142, and further, the mirror 147.
The light passing through the Hano mirror 148 and the imaging lens 149 is projected on the CCD 150. Therefore, the illuminating light flux of the infrared lamp 151 transmitted through the transmissive document 142 is transmitted regardless of the image (photosensitive image) of the transmissive document 142 such as a negative or a positive as shown in FIG. 17 and physically blocks the optical path. An image such as dust, dirt, or a scratch is projected on the CCD 150 as a shadow. CC
The infrared light image projected on the D150 is converted into an electric signal, and the infrared light image memory 23 is transmitted via the I / F 21 in FIG.
Is temporarily stored.

Next, the detection and correction of dust and flaws are performed in the steps after step S30. The principle of detection of dust and flaws will now be described in detail.

FIG. 19 is a graph in which the gradation levels of the image read by the transmissive original illumination lamp 144 and the infrared light lamp 151 are plotted in the main scanning direction, and the relationship between dust and the like is clearly shown. In FIG. 19A,
Reference numeral 181 denotes a positive film, and 182 denotes dust on the positive film 181. FIG. 19B shows the gradation level when the portion shown in FIG. 19A is read by the transparent document illumination lamp 144. The darker the gradation, the lower the gradation level. The gradation level is naturally low regardless of the image on the positive film. FIG. 19C shows the gradation level when the portion of FIG. 19A is read by the infrared light lamp 151, and the gradation level of the portion of the dust 182 becomes lower because the infrared light does not pass therethrough. , Garbage 1
The portion other than 82 has a substantially constant level 183 because infrared light passes through. Therefore, the threshold value 184 is set to a gradation level lower than the level 183, and the defective area 185 due to dust can be detected by extracting a portion below the threshold value 184.

The threshold value 184 is stored in the threshold value storage unit 24 in advance. Therefore, in step S30, the dust / flaw detection unit 25 transmits the threshold value 184 from the threshold value storage unit 24.
Is read from the infrared light image memory 23, and the defective region 185 is detected by sequentially comparing the infrared light image data with the threshold value 184.

If the infrared light image data is smaller than the threshold value 184 (NO in step S30), the defective area 185 is subjected to interpolation processing from a normal area around the defective area 185 in step S40 to thereby remove dust. 182 is reduced. The above-described comparison operation is performed on all the infrared light image data, and when a defective area is detected, interpolation processing is performed on the data of the corresponding ordinary image (Step S).
50).

[0019]

However, in order to detect dust and flaws using an invisible light image, the same original is read twice, such as scanning with invisible light and scanning with visible light. Must-have. For this purpose, it is necessary to scan the original with a scanning unit including at least a part of the photoelectric conversion unit, the optical system, and the processing circuit. Then, a shift occurs due to a problem in operation accuracy of the scanning unit between an image obtained by invisible light scanning and reading for detecting dust and flaws and an image obtained by visible light scanning and reading for obtaining actual image information. As a result, there was a problem that removal of dust and scratches was not satisfactory.

The present invention has been made in view of the above-described problems, and when reading a document to correct dust and flaws, it is possible to obtain appropriate dust and flaws even if a shift of repetition of a plurality of scanning operations occurs. It is intended to stably perform detection and correction.

Further, since the wavelength characteristics of the illumination, the light sensor, and the original are not necessarily ideal characteristics, the invisible light image may be affected by the visible light image. This may affect the image, and there is a problem that a visible light image which is not a defect due to dust or scratch is erroneously corrected.

SUMMARY OF THE INVENTION The present invention has been made in view of the above problems, and has as its object to reduce the deviation of the wavelength characteristic from the ideal and the influence of the shadow of the film holder.

Also, when reading at a high designated resolution,
Since the amount of image data is large, there is a problem that it takes a very long time to detect dust / scratch positions between the normal image and the infrared light image.

The present invention has been made in view of the above problems, and has as its object to speed up the dust / scratch removal processing without reducing the dust / scratch repair effect.

[0025]

SUMMARY OF THE INVENTION The present invention has been made to solve the above problems, and is constituted as follows.

[Claim 1] An original is illuminated by a visible light irradiating means for irradiating visible light and an invisible light irradiating means for irradiating invisible light, respectively, and visible light obtained by photoelectrically converting an optical image of the original is obtained. A signal processing method for processing an optical image signal and an invisible light image signal, wherein a first dust / flaw image signal is generated from the invisible light image signal; and And performing a correlation operation with the dust / scratch image signal, and changing the correlation operation method according to whether the original is a negative image or a positive image,
A signal processing method comprising:

[Claim 2] In the correlation operation, in the visible light image signal, a pixel corresponding to a dust / scratch position of the first dust / scratch image signal is shifted by a predetermined pixel for each shift amount. 2. The signal processing method according to claim 1, further comprising a calculating step of calculating a sum of signal values of pixels corresponding to dust / flaw positions.

[Claim 3] A negative / positive inverting step of performing a negative / positive inverting process on the photoelectrically converted signal to obtain a visible light image signal when the original is a negative original. The signal processing method according to claim 1.

[4] In the changing step, when the original is a negative image, the shift amount at which the sum of the calculating steps is the maximum is calculated, and when the original is a positive image, the shift amount is calculated. 4. The signal processing method according to claim 2, further comprising a shift amount detecting step of obtaining a shift amount that minimizes the sum of the steps.

[5] The signal processing method according to [4], further comprising a correction step of correcting the visible light image signal using the deviation amount and the invisible light image signal.

[Claim 6] The correcting step includes a step of generating a second dust / flaw image signal from the invisible light image signal using a second threshold value. The signal processing method according to claim 5, wherein the visible light image signal is corrected using a dust / flaw image signal.

[Claim 7] A document is illuminated by a visible light irradiating means for irradiating visible light and an invisible light irradiating means for irradiating invisible light, respectively. A signal processing method for processing an optical image signal and an invisible light image signal, comprising: generating a first dust / flaw image signal from the invisible light image signal using a first threshold; Generating a second dust / flaw image signal from the light image signal using a second threshold, and using the visible light image signal and the first and second dust / flaw image signals to generate the visible light image Correcting the signal.

[8] The signal processing method according to [7], further comprising the step of detecting the amount of displacement between the visible light image signal and the first dust / flaw image signal.

[Claim 9] The signal processing method according to any one of claims 6 to 8, further comprising a threshold setting step of setting the second threshold higher than the first threshold.

[Claim 10] In the threshold setting step, the first threshold is set to a gradation level lower by a times the standard deviation of the invisible light image from the histogram average value of the invisible light image,
The signal processing method according to claim 9, wherein the second threshold value is set to a level lower than the average value of the histogram of the invisible light image by b times smaller than the standard deviation a of the invisible light image.

[Claim 11] In the threshold setting step, when the standard deviation exceeds a predetermined value, the standard deviation is replaced with a predetermined value.
2. The method according to claim 1, wherein the first and second thresholds are set.
0. The signal processing method according to 0.

[Claim 12] The signal processing method according to claim 1, wherein the resolution of the invisible light image signal is different from the resolution of the visible light image signal.

[Claim 13] Irradiation means for selectively irradiating visible light and invisible light, a visible light image signal obtained by photoelectrically converting an optical image of a document irradiated by the irradiation means, and the visible light An image signal is a signal processing method for processing an invisible light image signal having a different resolution, and a step of generating a first dust / flaw image signal using the first threshold from the invisible light image signal, After adjusting the resolution of the visible light image signal and the resolution of the first dust / flaw image signal, a shift amount detecting step of detecting a position shift amount between the visible light image signal and the invisible light image signal; A signal processing method comprising: correcting the visible light image signal using the amount and the invisible light image signal.

[14] In the displacement amount calculating step, in the visible light image signal, a pixel corresponding to a dust / scratch position of the first dust / scratch image signal is shifted by a predetermined pixel. A calculating step of calculating a sum of signal values of pixels corresponding to dust / flaw positions for each amount, and calculating a position of the visible light image signal and the invisible light image signal from the sum calculated for each shift amount. 14. The signal processing method according to claim 13, which is a shift amount detecting step of detecting a shift amount.

[15] The non-visible light image signal is scaled to generate a second non-visible light image signal having the same resolution as that of the visible light image signal. 15. The signal processing method according to claim 12, wherein a dust / flaw position image signal is generated.

[Claim 16] The predetermined pixel in the calculation step is a plurality of pixels.
The signal processing method according to 5, 6, 14, or 15.

(17) When the visible light image signal is M times the invisible light image signal, the predetermined pixel is set to M pixels. 3. The signal processing method according to 1.

[Claim 18] The method further comprises a calculating step of performing the predetermined pixel in the calculating step by one pixel, and a step of determining whether the predetermined pixel in the calculating step is performed by a plurality of pixels or by one pixel. The signal processing method according to claim 16 or 17, wherein

[Claim 19] The signal processing method according to claim 18, wherein the judging step judges each of a plurality of pixels when the resolution of the visible light image signal is higher than a predetermined resolution.

[Claim 20] An irradiating means for selectively irradiating visible light and invisible light, and an image forming optical system for irradiating a document with each light and forming an optical image of the document. An image reading apparatus comprising: a control unit that executes the signal processing method according to claim 1.

[Claim 21] A program for causing a computer to implement the signal processing method according to any one of claims 1 to 19.

[Claim 22] Claim 21 to the computer
A recording medium on which a program for realizing the signal processing method according to 1 is recorded.

[0048]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Preferred embodiments of the present invention will be described below in detail with reference to the accompanying drawings.

<First Embodiment> A first embodiment will be described with reference to FIGS. The configuration of the image reading apparatus is the same as that shown in FIG. FIG. 1 is a block diagram illustrating a functional configuration of a dust / scratch removal unit 3 that removes dust / scratch from an image signal output from the image reading apparatus 1 according to the first embodiment. In FIG. 1, the dust / flaw removing unit 3 is shown as a device different from the image reading device 1, but may be configured inside the image reading device 1.

In FIG. 1, reference numeral 21 denotes the image reading device 1.
(I / F) 22 for inputting image data read by the scanner, 22 is a lamp 14 for illuminating the transparent original
4 or an image memory for storing an image read using the reflective document illumination lamp 145, and a red memory 23 for storing an image read using the infrared lamp 151 or an image indicating a dust / scratch position. An external light image memory, 25 a dust / flaw detection unit, 26 a dust / flaw correction unit, 31 a histogram generation unit, 32 a threshold value determination / storage unit, and 33 a displacement detection unit.

Next, the transparent original reading operation in the case of performing the dust / scratch removing operation in the first embodiment will be described in detail with reference to the flowchart of FIG.

First, in step S201, FIG.
Reflective illumination lamp 145 and infrared lamp 151
Is turned off, and the transparent document illumination lamp 144 is turned on.
At this time, the illuminating light flux of the transmissive original illumination lamp 144 is diffused by the diffusion plate 143 without unevenness, and the diffused luminous flux transmits through the transmissive original 142. This transmitted light beam is mirror 147,
After passing through the Hano-shaped mirror 148, the image forming lens 149
And is projected on the CCD 150. The image projected on the CCD 150 is converted into an electric signal, and the I / F of FIG.
The image data is temporarily stored in the image memory 22 through the memory 21. Here, when the transparent original is a negative film, a reversal process is performed to obtain a positive image, and then the image is temporarily stored in the image memory 22 (hereinafter, referred to as a “normal image”). The user specifies in advance whether the transparent original is a negative film or a positive film.

Next, in step S202, FIG.
Are turned off, and the infrared lamp 151 is turned on.
The illumination light beam of the infrared light lamp 151 having the characteristics shown in FIG. 15 is diffused by the diffusion plate 143 without unevenness, and the diffused light beam is transmitted through the transmission original 142 and further transmitted to the mirror 14.
7. The light that has passed through the Hano mirror 148 and the imaging lens 149 is projected onto the CCD 150. Therefore, the transparent original 14
Illumination luminous flux of the infrared lamp 151 transmitted through the light source 2 is shown in FIG.
As shown in FIG. 3, the negative, positive or other transparent original 142 is transmitted irrespective of the image (photosensitive image) of the original 142 and physically blocks the optical path.
An image such as a flaw is projected on the CCD 150 as a shadow. C
The infrared light image projected on the CD 150 is converted into an electric signal, and the infrared light image memory 23 is transmitted through the I / F 21 in FIG.
(Hereinafter, referred to as an “infrared light image”).

Next, in step S203, using the infrared light image data temporarily stored in the infrared light image memory 23, the threshold value La used in step S204 is calculated by the thirty-two threshold value determining / holding unit and stored. I do. This calculation method will be described later.

Next, in step S204, dust
The scratch detection unit 25 reads out the threshold value La from the threshold value determination / holding unit 32, reads out the infrared light image data from the infrared light image memory 23, and sequentially compares the infrared light image data with the threshold value La to remove dust. A flaw is detected, a first dust / flaw position image is generated, and temporarily stored in the infrared light image memory 23.

Next, in step S205, the position shift amount detector 33 detects the position shift between the normal image stored in the image memory 22 and the first dust / scratch position image stored in the infrared image memory 23. Detect the amount. Details of the detection will be described later.

Next, in step 206, using the infrared light image data temporarily stored in the infrared light image memory 23, the threshold value Lb used in step S207 is calculated and held by the 32 threshold value determining / holding units. . The method of calculating Lb will also be described later.

Next, in step S207, the dust / flaw detection section 25 reads the threshold Lb from the threshold determination / holding section 32, reads the infrared image data from the infrared image memory 23, and sequentially reads the infrared image data. and by comparing the threshold value Lb, it detects dust and scratches, and generates a second dust, scratches position image, temporarily additionally stored in the infrared image memory 23.

Next, in step S208, the correction unit 2
A step 6 reads the dust / flaw position from the second dust / flaw position image stored in the infrared light image memory 23 and shifts the dust / flaw position by the positional shift amount detected in step 205. It is determined that the image data of the normal image stored in the image memory 22 corresponding to the shifted position is data of a defective area in which loss has occurred due to dust or scratches. By performing interpolation processing and the like on the data of the defective area of the normal image from a normal area around the defective area, the influence of dust and scratches is reduced. The effect of dust and flaws is obtained by correcting the image data of the normal image corresponding to the position shifted by the positional shift amount detected in step 205 for all of the dust and flaw positions of the second dust and flaw position image. Can be obtained.

<Second Embodiment> A second embodiment will be described with reference to FIGS. FIG. 8A shows a state in which dust 102 is present on the positive film 101. FIG. 8B shows the portion of FIG. 8A read by the transparent document illumination lamp 144 shown in FIG. This shows the gray level when the image is displayed. As for the gradation level on the dust, since the light does not transmit, a gradation level distribution which is convex downward is obtained.
FIG. 8C shows the gradation level when the portion of FIG. 8A is read by the infrared lamp 151 shown in FIG. 14, and particularly when the portion is read at a resolution much lower than the designated resolution. Is shown. In this case, since the dust / flaw information is read at a lower resolution, the gradation level gradually changes. Infrared light image A threshold value set in advance as a predetermined level ΔL12 (or a threshold value determined by histogram and gradation level analysis) with respect to the gradation level L1 not affected by dust and flaws in FIG. 8C. A gradation level L2 having only a difference is set, and binarization processing is performed at the gradation level L2 to determine dust / flaw detection information. In FIG. 8B, if the dust / scratch width affected by the dust / scratch is P1 and the dust / scratch width determined as dust / scratch detection information in FIG. 8C is P2, FIG. In (c), since the image is read at a lower resolution, it is detected with a width of P2 which is larger than P1, which means that it is insufficient to correctly specify the dust / flaw position.

FIG. 8 (d) shows the gradation level when the portion of FIG. 8 (a) is read by the infrared light lamp 151 shown in FIG. 14, and particularly when the portion is read at the designated resolution. . In this case, since the dust / flaw information has been read at a higher resolution, the gradation level has been precisely changed. In this case, the gradation level L which is not affected by dust and scratches
3, a gradation level L4 having a difference from the infrared light image by a predetermined threshold (or a threshold determined by histogram and gradation level analysis) as a predetermined level ΔL34 is set, and the gradation level L4 is set as the gradation level L4. Binarization processing to determine dust / flaw detection information. Assuming that the dust / flaw width determined as the dust / flaw detection information in FIG. 8D is P3, in FIG.
Is detected with a width of P3 substantially equal to the above, which means that the dust / flaw position can be specified correctly.

As described with reference to FIGS. 8C and 8D, the information to be obtained in the infrared light image is not the resolution and the gradation distribution but the dust / scratch width. In other words, it means that it is only necessary to be able to specify the dust / scratch width even at a resolution lower than the specified resolution. For example, a predetermined threshold value is set even when reading at a resolution lower than about 1/2 of the specified resolution. By doing so, the width of dust and scratches can be specified almost accurately.

FIG. 8 (e) shows the gradation level when the portion of FIG. 8 (a) is read by the infrared light lamp 151 shown in FIG. 14, and in particular, is read at 1/2 the specified resolution. Shows the case where Also in this case, the gradation level is minutely changed, very similar to the case where dust / flaw information is read at a specified resolution. In this case, the difference between the gradation level L5 which is not affected by dust and flaws by a threshold value (or a threshold value determined by histogram and gradation level analysis) using the infrared light image as a predetermined level ΔL56. Is set, and binarization processing is performed at the gradation level L6 to determine dust / flaw detection information. If the dust / scratch width determined as dust / scratch detection information in FIG. 8 (e) is P4, then in FIG. 8 (e), even if it is read at half the designated resolution, P4 is substantially equal to P1. The width is detected, which means that the dust / flaw position can be specified correctly.

That is, even if the infrared light image is read in FIG. 8E, the dust / scratch position can be correctly specified as an effect when the infrared light image is read at a half resolution from the designated resolution. However, the speed can be increased without lowering the dust / scratch removal performance.

FIG. 10 is a schematic diagram of a read image. As an example of an image read by the image reading apparatus 1, a normal image obtained by reading a positive film including dust and scratches is shown in FIG.
(A) and (c) show infrared light images in FIGS. 10 (b) and (d).
It was shown to. FIGS. 10A and 10B show images obtained by reading a normal image and an infrared light image at low resolution. FIG.
(C) and (d) show the case of reading at a specified resolution twice as high as the high resolution, and are spatially shown as a pixel width Rv 'and a sub-scanning pixel width RL. Further, FIG.
(B) is also used for the description as a figure enlarged twice.
In the ordinary images shown in FIGS. 10A and 10C, pixel information of a dark density level is obtained at the place where dust and flaws exist.
In the infrared light images (b) and (d), binarization is performed by the dust / flaw detection processing to clarify the position of the dust / flaw.

Hereinafter, the transparent original reading operation in the case of performing the dust / scratch removing operation will be described in detail with reference to the flowchart shown in FIG.

First, in step S301, FIG.
Reflective illumination lamp 145 and infrared lamp 151
Is turned off, and the transparent document illumination lamp 144 is turned on.
At this time, the illuminating light flux of the transmissive original illumination lamp 144 is diffused by the diffusion plate 143 without unevenness, and the diffused luminous flux transmits through the transmissive original 142. This transmitted light beam is mirror 147,
After passing through the Hano-shaped mirror 148, the image forming lens 149
And is projected on the CCD 150. The image projected on the CCD 150 is converted into an electric signal,
It is temporarily stored in the image memory 22 via F21. At this time, a normal image read at the designated resolution Rs is obtained.

Next, in step S302, FIG.
Are turned off, and the infrared lamp 151 is turned on.
The illumination light beam of the infrared light lamp 151 having the characteristics shown in FIG. 16 is diffused without unevenness by the diffusion plate 143, and the diffused light beam passes through the transmission original 142 and further passes through the mirror 14.
7. The light that has passed through the Hano mirror 148 and the imaging lens 149 is projected onto the CCD 150. Therefore, the transparent original 14
Illumination light flux of the infrared lamp 151 transmitted through the light source 2 is shown in FIG.
As shown in FIG. 7, dust, dust, etc., which are transmitted regardless of the image (photosensitive image) of the transparent original 142 such as a negative or a positive, and physically block the optical path,
An image such as a flaw is projected on the CCD 150 as a shadow. C
The infrared image projected on the CD 150 is converted into an electric signal, and the infrared signal is stored in the infrared image memory 2 via the I / F 21 in FIG.
3 is temporarily stored. At this time, an infrared light image read at a predetermined resolution Rn lower than the specified resolution is obtained.

Next, in step S303, step S3
The infrared light image obtained in step 02 is subjected to a magnification process of M times from a predetermined resolution Rn to a specified resolution Rs. M = Rs / R
n. Here, the infrared light image can handle the spatial distance in the same pixel unit as the designated resolution.

Next, in step S304, in order to detect the position of the dust and flaws in the normal image, dust is generated in the normal image and the infrared light image in accordance with the scanning accuracy of S301 and S302 as shown in FIG.・ If the scratch position is shifted,
The displacement amount is detected for each M pixels of the normal image. Details of the detection of the displacement will be described later. For example, FIG.
If the magnification ratio is twice as shown in FIG. 10D and the dust / flaw position image in FIG. 10B, the shift amount is detected every two pixels. M
The shift amount of the entire image is detected from the sum of the shift amount detection performed for each pixel.

Next, in step S305, the dust / scratch image in the normal image is restored. Here, in the ordinary image, the dust / scratch image corresponding to the position where the dust / scratch position obtained from the infrared light image is shifted by the shift amount obtained in step S304 is restored. Then, the dust / scratch removal processing ends. According to this flowchart, the reading speed of the infrared light image can be increased, and as a result, the speed can be increased without lowering the dust / scratch removal performance.

<Third Embodiment> A third embodiment will be described with reference to FIGS. FIG. 10 has been described in the second embodiment. The transparent original reading operation when the dust / scratch removal operation is performed will be described in detail with reference to the flowchart shown in FIG.

First, steps S101 and S102 are the same as steps S301 and S302 described in the second embodiment.
Is read at the same Rs. Here, steps S101 and S10
In the case where the resolution is extremely high, as described with reference to FIG. 8 of the first embodiment, it is not necessary to read the infrared light image at a higher resolution than a certain resolution to specify the dust / flaw position. There are cases. A similar effect can be obtained when performing alignment between the ordinary image and the infrared image. The resolution of this boundary is set as a predetermined resolution Rd.

Next, in step S103, step S1
02, it is determined whether the infrared light image obtained is lower than a predetermined resolution Rd.
Through S104 and step S106, the amount of misregistration of dust and flaws is detected for each pixel. If the resolution is equal to or higher than the predetermined resolution Rd in S103, the process goes through steps S105 and S106. At this time, when the designated resolution as shown in FIG. 2 is very high, the position shift in the case where the dust / scratch position accompanying the steps of S101 and S102 is obtained between the normal image and the infrared light image is shifted. Quantity detection by M (= Rs
/ Rn) It is possible to increase the speed by detecting the dust / flaw position shift amount for each pixel. The dust / scratch position in the normal image is restored by correcting the dust / scratch position based on the misregistration amount detected in step S106, and the dust / scratch removal process ends. According to this flowchart, when the designated resolution is very high, the dust / scratch alignment processing between the normal image and the infrared light image can be accelerated, and as a result, the dust / scratch removal performance can be reduced without decreasing the dust / scratch removal performance. Can be achieved.

Hereinafter, detection of the amount of displacement and detection of the threshold values La and Lb, which are common to the present embodiment, will be described.

<Detection of Positional Displacement Amount> Hereinafter, the detection of the amount of positional deviation between the normal image and the first dust / flaw position image by the positional deviation amount detection unit 33 will be described with reference to FIGS.

FIG. 3 is a flowchart showing the detection of the amount of displacement, and FIG. 4 is a schematic diagram for explaining the correction of the displacement according to the present embodiment. As shown in FIG. 4, the downward direction is the sub-scanning direction, and the horizontal direction is the main scanning direction. 4 in FIG.
Reference numerals 01 and 402 denote pixels detected as dust and flaws in the infrared light image of (a), and 403 and 40 in the normal image of (b).
The pixel coordinates of 4 correspond to the coordinates of 401 and 402, respectively. However, in fact, a positional shift has occurred between the infrared light image and the ordinary image as described above, and the dust / flaw pixel 401 may be shifted several pixels above and below the 403 pixel. The dust / scratch position image data is a normal image.
In the case of a negative film, it is inverted and expressed as a high density value in a normal image. Therefore, dust and scratch coordinates 4 on the infrared
From the visible dust / flaw coordinates 403 corresponding to 01, the density values of the pixels in the range of ± N pixels in the sub-scanning direction are obtained. Asked
Among the (2N + 1) pixels, the pixel with the lowest density value in the case of positive film and the pixel with the highest density value in the case of negative film are the pixels corresponding to dust and scratches on the normal image. And the amount of deviation between the infrared image and the infrared light image (step S30)
1). However, if the shift amount is determined with only one pixel, an erroneous position shift determination may be made, and detection failure may occur in dust / scratch pixels. Therefore, the accuracy of the position shift determination is improved by obtaining the total amount of the density values in the normal image for each shift amount. Fig. 5 shows an example. FIG. 5A shows an example of a positive film in a case where the range for detecting a deviation is N = 10. Pixels,... Are all the pixels determined to be dust and flaws in the infrared light image, and a density value in a range of ± 10 pixels is obtained for each pixel, and the sum of the density values is calculated for each shift amount. It is determined (step 302). FIG. 5 (a)
In the example of the above, since the total value of the shift amount +1 pixel is the minimum value, in this case, the shift amount between the infrared light image and the ordinary image is set to +1 pixel in the sub-scanning direction, and correction is performed at the time of dust / flaw detection. (Step 303). Similar results can be obtained by using an average value instead of a total value of density values. As shown in step 304 of FIG. 3, a similar deviation correction amount is determined in the main scanning direction. In FIG. 5B, the total amount of density values in the normal pixels for each shift amount of all pixels determined as dust and flaws in the infrared light image was calculated. In order to reduce the time, as shown in FIG.
It can also be performed for every two pixels as in. If M is 2 in step S105 in FIG. 11 of the third embodiment, position shift detection is performed for every two pixels as shown in FIGS. 10 (c) and 10 (d). In this case, since the displacement amount is detected every two pixels in both the main scanning direction and the sub-scanning direction, a calculation as shown in FIG. 5C is performed to detect the displacement amount. In FIG. 5C, the calculation amount is reduced to 1/4 as compared with FIG. 5A.

<Calculation of Thresholds La and Lb> The calculation of the thresholds La and Lb used by the threshold determination / storage section 32 to detect dust / flaw positions from an infrared light image will be described below.

When detecting the position of dust and flaws by reading a transparent original with infrared light, a threshold value is obtained from the histogram using the histogram of the infrared light image, and pixels less than the threshold value are determined as dust and flaws. , The pixel in the normal image corresponding to the pixel is corrected.

As a method of determining the threshold value, a value obtained by subtracting n times the value of the standard deviation from the average value calculated from the histogram of the infrared image is set as the threshold value. Equation (1) illustrates this method. Threshold = Average-SD xn (1) Threshold: Threshold Average: Average value of infrared image SD: Standard deviation of infrared image n: Coefficient n Figure 16 shows the spectral intensity of visible light and infrared light (peak wavelength 880 nm) FIG. 17 is a diagram showing the distribution, and FIG. 17 is a diagram showing the transmittance of yellow, magenta and cyan colors of a general negative / positive color film.

As is clear from FIG. 17, in the case of a general color film, the transmittance of infrared light is very high, regardless of the kind of dye, so that the luminous flux of the infrared lamp is almost transmitted. However, even if the transmittance is high, it does not reach 100%. In addition, some films have low transmittance near infrared light depending on the components of the film. For the above-mentioned reason, particularly when such a film is scanned with infrared light, not only dust and scratches but also image information originally obtained with visible light may be reflected in the infrared light image. When calculating the shift amount between the infrared light image and the normal image, if the normal image information is included, an error occurs in the calculation of the shift correction amount. Therefore, different threshold values are used for the first dust / flaw detection 204 for position shift correction and the second dust / flaw detection 207 for dust / flaw correction in FIG.

FIG. 6 shows an example of a histogram of an infrared light image, where Ta 602 represents a first threshold and Tb 603 represents a second threshold. The first threshold value Ta is used to measure the amount of deviation correction, and is set to a relatively low level such that there is no reflection other than dust and scratches. On the other hand, the second threshold
Tb is used to judge dust and flaws to be removed after misalignment correction, and the level is set so that dust and flaws are not detected even if reflections remain in the infrared image. .

In the above equation (1), the degree of dust / flaw detection can be adjusted by setting the coefficient n. If n is made too large, detection omission may occur. Conversely, if n is made small, detection omission decreases, but the possibility of occurrence of reflection increases.

The time coefficient n of the second threshold Tb for dust / flaw detection is set to n = b, and the coefficient n = a for the first threshold Ta is 2
By setting the relationship between the two coefficients as a> b, the threshold value of the positional deviation correction always becomes lower by a certain value, so that the influence of the reflection can be reduced.

FIG. 7 shows an example of a histogram when the frame of the film holder is reflected in the infrared light image. A film holder is used to fix a film when a transparent original is placed on an image reading apparatus. The film holder is usually formed of plastic or the like, and when the infrared image is read, if the film holder is reflected in the reading area, the film holder does not transmit light, so the film holder part (hereinafter referred to as holder shadow) is usually The density value is much lower than that of dust and scratches. Then, the density average value of the infrared light image is shown in FIG.
As shown in 501, it becomes lower than the case without the holder shadow, and at the same time, the standard deviation becomes larger. Then the above formula
Since (1) is expressed by the following equation (2), the threshold value becomes low, and the first threshold value Ta ′ and the second threshold value Tb ′ are also set to 70 in FIG.
2,703. Threshold (↓↓) = Average (↓) −SD (×) × n (2) In this case, depending on the size of the holder shadow entering, the threshold value becomes too low as shown at 702 and 703 in FIG. There is a possibility that an error may be made in the determination and the misalignment determination. Even if the holder shadow is not included, the same phenomenon may occur due to reflection. Therefore, if the standard deviation calculated by providing the maximum limit value to the standard deviation SD is larger than the limit, it is determined that there is an influence of the holder shadow or the reflection, and is replaced with a fixed value set in advance.

[Other Embodiments] The present invention may be applied to a system constituted by a plurality of devices or to an apparatus constituted by a single device.

Further, an object of the present invention is to supply a storage medium (or a recording medium) in which program codes of software for realizing the functions of the above-described embodiments are recorded to a system or an apparatus, and to provide a computer (a computer) of the system or the apparatus. Or a CPU or MPU) reads out and executes the program code stored in the storage medium,
It goes without saying that this is achieved. In this case, the program code itself read from the storage medium implements the functions of the above-described embodiment, and the storage medium storing the program code constitutes the present invention.
In addition, by executing the program code read by the computer, not only the functions of the above-described embodiments are realized, but also based on the instructions of the program code,
The operating system (OS) running on the computer performs part or all of the actual processing,
It goes without saying that a case where the function of the above-described embodiment is realized by the processing is also included.

Further, after the program code read from the storage medium is written into the memory provided in the function expansion card inserted into the computer or the function expansion unit connected to the computer, the program code is read based on the instruction of the program code. , The CPU provided in the function expansion card or the function expansion unit performs part or all of the actual processing,
It goes without saying that a case where the function of the above-described embodiment is realized by the processing is also included.

[0089]

As described above, in the image reading apparatus, even when a position shift occurs between the visible light image and the invisible light image due to the visible light, the shift is detected by detecting the shift amount, and the image reading apparatus does not cause any trouble. It is possible to perform dust and scratch correction without any problem.

Further, it is possible to remove the influence of visible light on the invisible light image and prevent unnecessary image correction.

Further, in the image reading apparatus, even when dust and scratches are removed from an image read with invisible light at a resolution different from the resolution of the image read with visible light, the performance of removing dust and scratches is reduced. Since the position of dust / scratch can be specified without any problem, the reading speed with invisible light can be optimized, and the speed of the dust / scratch removal function can be increased.

When the image is read at a very high resolution with respect to the size of the dust and flaws, the dust and flaw removal performance is reduced in order to identify the position of the dust and flaws in the image read by visible light. Pixel units for processing images read with invisible light and visible light can be thinned out without any need, so that the process of identifying the position of dust and flaws can be accelerated, and the speed of the dust and flaw removal function can be increased. .

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating a configuration of an image reading system according to a first embodiment of the present invention.

FIG. 2 is a flowchart illustrating processing in a dust / scratch removal unit according to the first embodiment of the present invention.

FIG. 3 is a flowchart illustrating a process performed by a displacement amount detection unit according to the first embodiment of the present invention.

FIG. 4 is a diagram for explaining detection of a displacement amount according to the first embodiment of the present invention.

FIG. 5 is a diagram illustrating an example of misregistration detection calculation in a misregistration amount detection unit according to the first embodiment of the present invention.

FIG. 6 is a diagram illustrating a threshold value determination and a threshold value determination according to the first embodiment of the present invention.
FIG. 4 is a diagram illustrating setting of two thresholds in a storage unit.

FIG. 7 is a diagram illustrating a threshold value determination and a threshold value determination in the first embodiment of the present invention
FIG. 8 is a diagram illustrating a histogram and threshold settings when a holder shadow is reflected in a storage unit.

FIG. 8 is a schematic diagram illustrating a second embodiment of the present invention.

FIG. 9 is a flowchart illustrating processing in the image reading apparatus according to the second embodiment of the present invention.

FIG. 10 is a schematic diagram illustrating second and third embodiments of the present invention.

FIG. 11 is a flowchart illustrating processing in the image reading device according to the third embodiment of the present invention.

FIG. 12 is a schematic diagram when the resolution is low in the third embodiment of the present invention.

FIG. 13 is a configuration diagram of a conventional image reading apparatus.

FIG. 14 is a configuration diagram of a conventional image reading apparatus that detects a defective area due to dust and scratches on a transparent original.

FIG. 15 is a block diagram illustrating a configuration of a conventional image reading system.

FIG. 16 is a diagram illustrating spectral intensity distributions of a transmission original illumination lamp and an infrared light lamp.

FIG. 17 is a diagram showing spectral transmittance characteristics of three color pigments in a general color film and a peak wavelength of a spectral intensity distribution of an infrared light lamp.

FIG. 18 is a flowchart showing a conventional process in a dust / flaw removing unit.

FIG. 19 is a diagram showing a relationship between dust on a film and a gradation level obtained by reading the film with a transmissive original illumination lamp and an infrared light lamp in a conventional example.

 ──────────────────────────────────────────────────続 き Continuing on the front page (72) Inventor Tsutomu Takayama 3-30-2 Shimomaruko, Ota-ku, Tokyo Canon Inc. (72) Inventor Atsuko Kashiwazaki 3- 30-2 Shimomaruko, Ota-ku, Tokyo F term in reference (reference) 5B047 AA01 AA05 BA02 BB02 BC01 CA19 CA21 CB04 CB09 CB12 DA06 DC04 5C072 AA01 CA02 EA05 UA05 UA11 UA20 VA03 5C077 LL02 MM03 PP05 PP43 PQ08 SS01 TT06

Claims (22)

[Claims] 1. A visible light image signal obtained by irradiating an original with a visible light irradiating means for irradiating visible light and an invisible light irradiating means for irradiating invisible light, and photoelectrically converting an optical image of the original. And a signal processing method for processing an invisible light image signal, wherein a first dust / flaw image signal is generated from the invisible light image signal; and A change operation for performing a correlation operation with a flaw image signal and changing the correlation operation method according to whether the original is a negative image or a positive image. 2. The correlation operation according to claim 1, wherein the pixel corresponding to the dust / scratch position of the first dust / scratch image signal is shifted by a predetermined pixel in the visible light image signal, and the dust / scratch amount is shifted for each shift amount. The signal processing method according to claim 1, further comprising a calculating step of calculating a sum of signal values of pixels corresponding to the flaw positions. 3. A negative / positive inverting step of performing a negative / positive inversion process on the photoelectrically converted signal to obtain a visible light image signal when the original is a negative original. 2. The signal processing method according to 1. 4. In the changing step, when the original is a negative image, the shift amount at which the sum of the calculating steps is maximum is determined by the sum of the calculating steps when the original is a positive image. 4. The signal processing method according to claim 2, further comprising a shift amount detecting step of obtaining a shift amount that minimizes the shift amount. 5. A signal processing method according to claim 4, characterized in that it comprises a correction step of correcting the visible light image signal using the said shift amount invisible light image signal. 6. The correction step includes a step of generating a second dust / flaw image signal from the invisible light image signal using a second threshold value, wherein the deviation amount and the second dust / flaw image signal are generated. The signal processing method according to claim 5, wherein the visible light image signal is corrected using an image signal. 7. A visible light image signal obtained by irradiating an original with a visible light irradiating means for irradiating visible light and an invisible light irradiating means for irradiating invisible light, and photoelectrically converting an optical image of the original. And a signal processing method for processing an invisible light image signal, wherein a first dust / flaw image signal is generated from the invisible light image signal using a first threshold value, and the invisible light image signal Generating a second dust / flaw image signal using a second threshold value from; and correcting the visible light image signal using the visible light image signal and the first and second dust / flaw image signals. A signal processing method. 8. The signal processing method according to claim 7, further comprising a step of detecting a displacement between the visible light image signal and the first dust / flaw image signal. 9. The signal processing method according to claim 6, further comprising a threshold setting step of setting the second threshold higher than the first threshold. 10. The threshold setting step sets the first threshold to a gray level lower than the average value of the histogram of the invisible light image by a times the standard deviation of the invisible light image, and sets the second threshold to a non-visible level. 10. The signal processing method according to claim 9, wherein the level is set to a level lower by b times smaller than a of the standard deviation of the invisible light image from the average value of the histogram of the visible light image. 11. The signal processing according to claim 10, wherein the threshold setting step sets first and second thresholds by replacing the standard deviation with a predetermined value when the standard deviation exceeds a predetermined value. Method. 12. The signal processing method according to claim 1, wherein a resolution of the invisible light image signal is different from a resolution of the visible light image signal. 13. An irradiating means for selectively irradiating visible light and invisible light, and a visible light image signal and a visible light image signal obtained by photoelectrically converting an optical image of a document irradiated by the irradiating means. A signal processing method for processing invisible light image signals having different resolutions, wherein a first dust / flaw image signal is generated from the invisible light image signal using a first threshold value; After adjusting the resolution of the image signal and the first dust / flaw image signal, a shift amount detecting step of detecting a position shift amount between the visible light image signal and the invisible light image signal, and the shift amount and the shift amount A correcting step of correcting the visible light image signal using an invisible light image signal. 14. The displacement amount calculating step includes, in the visible light image signal, shifting a pixel corresponding to a dust / scratch position of the first dust / scratch image signal by a predetermined pixel.
It has a calculation step of calculating the sum of the signal values of the pixels corresponding to the dust and scratch positions for each shift amount, from the sum calculated for each shift amount, the visible light image signal and the invisible light image signal 14. The signal processing method according to claim 13, further comprising a shift amount detecting step of detecting the shift amount of the position. 15. The non-visible light image signal is scaled to generate a second non-visible light image signal having the same resolution as the visible light image signal. 13. A flaw position image signal is generated.
15. The signal processing method according to any one of claims 14 to 14. 16. The method according to claim 2, wherein the predetermined pixels in the calculating step are a plurality of pixels.
16. The signal processing method according to 4 or 15. 17. The method according to claim 12, wherein when the visible light image signal is M times the invisible light image signal, the predetermined pixel is set to M pixels. Signal processing method. 18. The method according to claim 1, further comprising: a calculating step of performing the predetermined pixel in the calculating step with one pixel; and a step of determining whether the predetermined pixel in the calculating step is performed by a plurality of pixels or by one pixel. Item 18. The signal processing method according to Item 16 or 17. 19. The signal processing method according to claim 18, wherein, in the determining step, when the resolution of the visible light image signal is higher than a predetermined resolution, the determination is made every plural pixels. 20. An image reading apparatus comprising: irradiating means for selectively irradiating visible light and invisible light; and an image forming optical system for irradiating a document with each light to form an optical image of the document. An image reading apparatus, comprising: a control unit that executes the signal processing method according to claim 1. 21. A program for causing a computer to implement the signal processing method according to claim 1. Description: 22. A recording medium on which a program for causing a computer to implement the signal processing method according to claim 21 is recorded.