JP2003348305A - Image processing apparatus and image processing method - Google Patents

Abstract

(57) [Problem] To eliminate erroneous correction due to erroneous detection and uncorrected due to non-detection in correction of dust and scratches in an image reading device,
Improve processing speed and corrected image. An image reading apparatus includes a detection unit that detects dust and scratch information from an invisible light image, and a correction unit that corrects dust and scratch from a visible light image and dust and scratch information. The determination of dust and scratches is performed using a threshold, and the threshold includes a threshold changing unit that changes according to the reading resolution.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image reading apparatus having correction means for correcting missing or degraded image information of a digital image obtained due to dust or scratches on a document surface. The present invention relates to a scanner that reads a developed film and corrects missing or deteriorated information due to dust or scratches on the film.

[0002]

2. Description of the Related Art Conventionally, image information recorded on a film is read by a line sensor and the RG of the line sensor is read.
The B output voltage is amplified by an analog amplifier, then converted into a digital signal by an A / D converter, and subjected to various digital image processing and output. In the film scanner, the resolution of the film scanner itself increases as the resolution of the film scanner itself increases. On the other hand, it has become possible to obtain more detailed information, but on the other hand, there has been a problem that dust and flaws existing on a film as an original have a large effect on an output image.

As a method for solving this problem, Japanese Patent No. 2559970 discloses a method for correcting the effect of a defect of a storage medium on an image stored in the storage medium.
Applying infrared energy and visible light energy to the storage medium, and detecting an infrared energy distribution corresponding to the defect caused by applying the infrared energy to the storage medium in association with each storage location of the storage medium. Detecting the visible light energy distribution generated by applying the visible light energy to the storage medium in association with each of the storage locations of the storage medium; and detecting, for each of the storage locations of the storage medium, If the infrared energy component intensity is greater than a predetermined threshold, the visible light energy distribution intensity at the storage location is increased to a level that cancels the infrared energy distribution intensity at the storage location, and the detected infrared energy distribution intensity is equal to or less than the predetermined threshold. In the case of, the visible light energy of the storage location The distribution intensities presents a method for correcting the influence of defects in storage media including a step of correcting by interpolation.

It is described that the processing content is changed depending on the magnitude of data obtained with respect to a threshold.

In Japanese Patent Application Laid-Open No. 2000-341473, a film flaw or dust on a film is detected by invisible light, and when the detected flaw or dust is corrected by image processing, it is determined based on information unique to the film. A flaw-corrected image processing method that controls the threshold value of the flaw or dust detection so as to surely detect the flaw or flaw but does not erroneously detect anything other than flaw or flaw as a flaw or dust. A flaw-corrected image processing apparatus to which this method is applied is presented.

Here, it is described that the threshold value is controlled by information unique to the film, ie, the type of the film.

[0007]

However, such control is effective only when reading is performed at a fixed resolution, such as a so-called laboratory, and as a consumer device that allows the user to change the resolution arbitrarily. In an image reading apparatus such as a flatbed scanner or a film scanner, there is a problem that the detection of dust or flaw cannot be reliably detected due to a change in the output of dust or flaw in an infrared image due to a change in resolution.

[0008]

In order to solve the above-mentioned problems, an image reading apparatus according to the present invention comprises means for reading an image by visible light and an image by invisible light, and dust and scratches from the means for reading by invisible light. Image reading comprising: detecting means for detecting the information of the image and image information obtained from the means for reading the visible light, and correcting means for correcting dust and scratches on the document surface based on the information of dust and scratches by the detecting means. In the apparatus, the determination of dust and scratches by the detection means is determined by a threshold value, and the threshold value has means for changing the resolution according to the resolution of the reading means. And then amplifies the RGB output voltage of the line sensor by an analog amplifier, and then converts and outputs the digital signal by an A / D converter. It is to have a further threshold is a threshold 1, constructed from two threshold 2, at least one threshold value
One changes discontinuously with respect to the reading resolution, or at least one of the thresholds is represented by a function having the reading resolution as an argument. In order to change the value of the pixel, an averaging means using an average value of a plurality of adjacent pixels is provided.

A step of reading an image by visible light and an image by invisible light; a step of detecting dust or flaw information from the step of reading by invisible light; and a step of detecting image information and detection information obtained by means of reading visible light. The step of determining dust and scratches by the correction step for correcting dust and scratches on the document surface based on the information of dust and scratches by the detection step and the step of determining dust and scratches by the threshold value, and the threshold value changes according to the resolution of the reading step And an image processing method comprising the steps of:

Accordingly, it is an object of the present invention to reliably detect a defective image such as dust or a scratch regardless of a change in resolution, reduce erroneous detection, and improve processing speed and image quality.

[0011]

(Embodiment 1) The structure of a film scanner to which the present invention is applied will be described with reference to FIG. In particular, a case where a negative film is used as a document will be described.

Reference numeral 101 denotes a lamp for illuminating the film, which comprises a lamp a for visible light and a lamp b for infrared light.
It is controlled by a lamp drive circuit 110 receiving 16 commands. Reference numeral 102 denotes a film folder for holding a film, which can be moved in the direction of arrow A by a sub-scanning motor 107, and is controlled by a motor drive circuit 109 which receives a command from the CPU 116. Further, the position of the film folder 102 is detected by the sub-scanning position detection sensor 106, and the initial position of the film folder 102 is obtained. 1
03 is an infrared light cut filter. Here, the sub-scanning motor 107 uses a stepping motor, and the sub-scanning position detection sensor 106 uses a photo interrupter. 10
Reference numeral 4 denotes an imaging lens which forms an image of light radiated from the lamp 101 and transmitted through the film on a CCD line sensor (hereinafter referred to as CCD) 105.

The CCD 105 is controlled by a CCD driving circuit 108 which receives a command from the CPU 116, and has an electronic shutter function capable of controlling the timing of discharging the accumulated charges. The CCD 105 is mounted so as to be perpendicular to the moving direction A of the film folder 102. Due to this positional relationship, the main scanning direction, which is the longitudinal direction of the CCD 105, and the sub-scanning direction, which is the moving direction of the film folder 102, are perpendicular (orthogonal). ).

Reference numeral 111 denotes an analog processing circuit which performs gain setting and clamp processing of an analog image signal output from the CCD 105, and can independently amplify each of RGB signals by an analog amplifier. An A / D converter 112 converts an analog signal output from the analog processing circuit 111 into a digital signal. 113 is A /
A digital image processing unit that processes an image signal digitized by the D converter 112. The digital image processing unit performs a negative-positive conversion or a white balance adjustment of a not-shown histogram creation unit and a gradation conversion curve for adjusting an input signal to an appropriate gradation. It consists of a gradation conversion curve creation unit to be created. A line buffer 114 temporarily stores image data.
Reference numeral 15 denotes an interface unit for connecting to an external device such as a personal computer. Reference numeral 116 denotes a system controller that stores the sequence of the entire film scanner and performs various operations in accordance with commands from an external device.

In such a configuration, software in the system controller 116 (hereinafter referred to as firmware software) and software for scanning a film scanner from an external device such as a personal computer (hereinafter referred to as driver software).
The image data is output to an external device by communicating with the external device.

Next, dust correction performed by the image processing section 113 will be described in detail with reference to FIG.

In S201 and S202, the visible image and the infrared image respectively read from the film by the switching operation of the irradiation lamp by the lamp driving circuit are output as digital signals by the A / D converter 112 and input to the image processing unit 113. You. Subsequently, in S203, a threshold value for detecting dust and scratches from the infrared image is calculated.

A method of detecting dust and scratches on a film by infrared light will be described with reference to FIGS. FIG. 3 (A),
(B) shows an output signal for one line of the CCD 105 when the film is irradiated with visible light and infrared light. When the lamp for visible light is illuminated by the lamp driving circuit 110, it is difficult for the output of the CCD 105 to discriminate between a change in the output of the CCD 105 due to dust or a scratch and a change due to an image signal as shown in FIG. On the other hand, in the case of (B) irradiated with infrared light, the CCD 10 based on the image signal is used.
Since there is almost no change in the output of No. 5, an output change due to dust or a flaw appears, and it is possible to detect the position of the dust or flaw on the film by using a threshold.

This utilizes the characteristics of the transmittance with respect to the wavelength of the film, and a general film has a high transmittance for infrared light. Thus, by irradiating the film with infrared light, signal change information due to dust or flaws on the film can be obtained irrespective of image information recorded on the film.

FIG. 4 shows a change in the read pixel due to a change in the read resolution. When there is dust or a flaw having a size of about one pixel, and when reading is performed at the highest resolution, one pixel is read. The information is reflected as it is. In the case of low resolution of 1/4 of the maximum resolution, 1 with dust and scratch information
Since the average of the pixel and three normal pixels around the pixel is obtained, the information of the pixel of the dust or flaw is weakly output. FIG. 5A,
(B) shows output signals for one line of the CCD 105 at the time of high resolution and at the time of low resolution. As described above, when the resolution becomes low, the CC of dust and scratches and normal image
The output difference of D105 becomes smaller. Therefore, when the resolution is changed, it becomes necessary to change the threshold value accordingly.

[0021]

[Table 1]

As shown in Table 1, when the maximum resolution is R, data in which threshold values are set discontinuously for each resolution such as 1/2, 1/3, and 1/4 of the resolution are stored in advance in the storage means. When the resolution is designated, this data is called and used. For other resolutions, once processed at the closest resolution,
Finally, the image is expanded or compressed by image processing such as the nearest neighbor method and output at a desired resolution. For example, if the maximum resolution is 4000 dpi, 4000 dpi, 2
Thresholds are set for each of the resolutions of 000 dpi, 1000 dpi, and 500 dpi, and when the reading resolution instruction is 800 dpi, a threshold of 1000 dpi is used, and all processes such as dust and scratch correction and gamma are performed.
Finally, the resolution is converted to 800 dpi.

Here, the threshold 1 and the threshold 2 are set when correction processing is selected according to the amount of attenuation of infrared light due to dust or scratches on the film, or when a normal image but a very bright image is used. This is for determining whether or not the image is reflected in an infrared image. If the reflection described later is an infrared output between the threshold 1 and the threshold 2 and is brighter than a predetermined value as compared with the output of the visible image that is reflected, it is regarded as dust or a scratch. If it is darker, it is determined that the visible image is reflected.

As a method of calculating the threshold value, for example, there is a method of calculating the maximum value of the infrared signal level, and setting the threshold value 1 at a position of a predetermined ratio and the threshold value 2 at a position of a smaller ratio. . Here, the threshold value is calculated based on the maximum value of the infrared signal level, but needless to say, an average value or a median value may be used, and there is no particular limitation.

Next, in step S204, a process for correcting a shift between the visible image and the infrared image is performed. Although the infrared image and the visible image are read by scanning the film folder 102 with the sub-scanning motor 107, in this embodiment, since the visible image and the infrared image are separately acquired in a reciprocating manner, an error due to feeding accuracy may occur. is expected. Since dust correction is performed on a pixel-by-pixel basis, deviations of several pixels between the visible image and the infrared image due to these errors must be corrected as much as possible. The shift correction process will be described with reference to FIG.

First, a predetermined number of defective pixel areas are detected in the sub-scanning direction using an infrared image. Since this is a correction process for detecting a shift between the visible image and the infrared image due to the feed accuracy, it is based on the assumption that the shift direction is only the sub-scanning direction. Next, the coordinates of the detected defective pixel area in the infrared image are acquired. Then, while shifting the visible image and the infrared image by a predetermined number of pixels, the sum of the signal values of the visible image and the infrared image corresponding to the coordinates of the defective pixel area is calculated. When the sum of the calculated signal values is the minimum, it is considered that the defective pixel areas of the visible image and the infrared image match, and the number of shifted pixels is stored as the number of shifted pixels.

This is based on the idea that the defective pixel area of the visible image will have a lower signal value than the peripheral pixels. Further, the number of defective pixel areas to be detected and the number of pixels to be shifted can be determined depending on whether correction accuracy or required calculation time is prioritized.

After the shift between the visible image and the infrared image in the sub-scanning direction has been corrected by the shift correction process, a correction process 1 as a first correction process is performed in S205. Correction processing 1 finds a continuous defective pixel region from the defective pixel, detects normal pixels that do not need to be corrected from both ends of the defective pixel region, finds a normal pixel region (normal pixel pattern) continuous from the normal pixels, and detects the defective pixel region. Is replaced with the normal pixel area (normal pixel pattern).

After the relatively uniform image portion has been corrected by the correction processing 1 in S205 in this manner, in the correction processing 2 in S206, an image having a large luminance level difference (edge image portion) that could not be completely corrected by the correction processing 1 With respect to the defective pixel region of (1), the edge direction is detected, and correction is performed to replace the defective pixel by using normal pixels located in the periphery in a direction substantially parallel to the edge direction. Thus, the correction processing 2 of S206
After the edge image portion is corrected by the above, correction process 3 is performed to further eliminate the remaining correction. In the correction process 3, a four-point internal interpolation is performed to find and replace the four pixels of the upper, lower, left, and right from the detected defective pixel (S207). Is averaged (S208), and an image is output without discomfort from surrounding pixels (S209).

(Embodiment 2) A second embodiment of the present invention will be described with reference to FIG. Since the main flow of the second embodiment is the same as that of the first embodiment, the description is omitted.

FIG. 7 shows a change in the read pixel due to a change in the read resolution. When reading at the highest resolution, as in the first embodiment, the information of one pixel is directly reflected. If the resolution is not an integer fraction of the highest resolution, one output pixel is the average of one pixel with dust or scratch information and two normal pixels around it, and the next pixel is 1
Since the pixels are doubled and the average of the three pixels is obtained, the information on the pixels of dust and scratches is weakly output. Equation 1 is an equation used to determine the threshold 1, and Equation 2 is an equation used to determine the threshold 2 similarly. In this embodiment, the output resolution r is a polynomial having an argument. This makes it possible to deal with fine dpi reading.

Each of the above embodiments is merely an example of the embodiment for carrying out the present invention, and the technical scope of the present invention should not be interpreted in a limited manner. It is. That is, the present invention can be implemented in various forms without departing from the technical idea or the main features. For example, the threshold value may be set after the CCD output of the infrared image is expanded by electrical processing.

[0033]

As described above, according to the present invention, erroneous detection of dust and scratches can be reduced by changing the threshold value in accordance with the change in the reading resolution, and the processing speed and the corrected image can be improved. Become.

[Brief description of the drawings]

FIG. 1 is a configuration diagram of a film scanner according to an embodiment of the present invention.

FIG. 2 is a flowchart of a trash correction process.

FIG. 3 is a diagram showing detection of dust on a film by an infrared light image.

FIG. 4 is a diagram illustrating a method for outputting image data according to a difference in resolution according to the first embodiment.

FIG. 5 is a diagram showing an output of an infrared light image depending on a difference in resolution.

FIG. 6 is a diagram illustrating shift correction.

FIG. 7 is a diagram illustrating an output method of image data according to a difference in resolution according to the second embodiment.

[Explanation of symbols]

101a Lamp for visible light 101b Infrared lamp 102 Film folder 103 infrared cut filter 104 imaging lens 105 CCD line sensor 106 Sub-scanning position detection sensor 107 Sub-scanning motor 108 CCD drive circuit 109 Motor drive circuit 110 Lamp drive circuit 111 Analog processing circuit 112 A / D converter 113 Digital Image Processing Unit 114 line buffer 115 Interface 116 CPU

   ────────────────────────────────────────────────── ─── Continuation of front page    F term (reference) 5B047 AA05 BA01 BB02 CB10 CB12                       CB22 DA06 DC09                 5C072 AA01 BA20 DA02 DA09 EA05                       QA10 TA04 UA06 UA07 VA03                 5C077 LL02 PP02 PP20 PQ18 RR04                       RR15 RR16 SS01 TT06

Claims (7)

[Claims] 1. A means for reading an image using visible light and an image using invisible light, a detecting means for detecting information on dust and scratches from the reading means using invisible light, and image information obtained from the means for reading visible light. In an image reading apparatus having correction means for correcting dust and scratches on a document surface based on information on dust and scratches by the detection means, the determination of dust and scratches by the detection means is determined by a threshold, and the threshold is An image reading apparatus comprising: a threshold changing unit that changes according to a resolution of the reading unit. 2. The apparatus according to claim 1, wherein said reading means includes means for reading image information by a line sensor, amplifying an RGB output voltage of said line sensor by an analog amplifier, and converting the digital output to a digital signal by an A / D converter. The image reading device according to claim 1, wherein: 3. An image reading apparatus according to claim 1, wherein said threshold value comprises two threshold values, a threshold value 1 and a threshold value 2. 4. The image reading apparatus according to claim 1, wherein at least one of the threshold values changes discontinuously with respect to a reading resolution. 5. The image reading apparatus according to claim 1, wherein at least one of the thresholds is represented by a function having a reading resolution as an argument. 6. The image reading apparatus according to claim 1, further comprising an averaging unit that uses an average value of a plurality of adjacent pixels as a method of changing the resolution to a lower resolution in the invisible light reading unit. . 7. A step of reading an image by visible light and an image by invisible light, a detecting step of detecting dust or flaw information from the step of reading by invisible light, and an image obtained from the means for reading visible light. A correction step for correcting dust and scratches on the document surface based on the information and the information on dust and scratches by the detection step; a step of determining the determination of dust and scratches by the detection step by using a threshold value; A threshold changing step that changes according to the resolution of the step.