JP2002033906A - Image-correcting method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a method for correcting an image for allowing, even if an operator is one who does not have sufficient knowledge or experience to easily and quickly perform proper image correction, and to perform efficient correcting work, corresponding to the personal sensibility of an operator. SOLUTION: Language expression, indicating the state of an image or the correcting direction of this image and an image correction condition corresponding to the language expression are set preliminarily, and the language expression is inputted as a correction instruction according to an image, and the image is corrected by using the image correction condition corresponding to the inputted language expression.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention belongs to the technical field of image processing, and more particularly, to an image correction method capable of easily and satisfactorily correcting an image in a test or the like performed in a laboratory.

[0002]

2. Description of the Related Art At present, an image photographed on a photographic film (hereinafter, referred to as a film) such as a negative film or a reversal film is printed on a photographic paper (photosensitive material) by projecting the image of the film onto the photosensitive material. Exposure, so-called direct exposure, is the mainstream.

In recent years, digital photo printers using digital exposure have been put to practical use. Digital photo printers basically include a scanner (image reading device) that photoelectrically reads an image recorded on a film by reading light incident on the film and reading the projected light, and reproduces an appropriate image. In order to output a printed image (appropriate print), input image data read by a scanner or input image data supplied from a digital camera or the like is subjected to a predetermined process, and the image is output as image data for image recording. In accordance with the processing device and the image data output from the image processing device, the photosensitive material is scanned and exposed by, for example, light beam scanning to record a latent image, subjected to a development process, and a reproduced image is reproduced (photograph). (Image recording apparatus).

According to such a digital photo printer, image processing (optimization) can be performed by processing image data. Therefore, gradation adjustment, color balance adjustment, color / density adjustment, and the like can be suitably performed. Thus, high-quality prints that cannot be obtained by conventional direct exposure can be obtained.

[0005]

As described above, in a digital photo printer, image processing is performed by processing input image data in order to output a proper print. On the other hand, in a normal direct exposure photo printer, image processing is similarly performed by adjusting the amount of projection light for exposing a photosensitive material (photographic paper) and inserting a color filter in order to output a proper print. . here,
The photo printer checks whether the image processing is proper, that is, whether a proper print can be output, and if necessary, corrects the image (that is, adjusts the image processing conditions). It is configured to be able to do it.

[0006] Regardless of the digital or direct exposure photo printer, image processing conditions are generally such that an image photographed on a film is photoelectrically read by a CCD sensor or the like, and an image to be printed is formed. It is set by obtaining data and analyzing the obtained image data (image analysis). In addition, the verification is usually performed by processing image data of an image read by the CCD sensor or the like in accordance with the set image processing conditions and reproducing the processed image data as a visible image on a display. This is performed using images.

When the operator determines that the reproduced simulation image is inappropriate, the image is corrected. However, in order to quickly correct the image, it is necessary to use specialized knowledge or the like. Experience is required.

For example, the correction of the color of an image is generally performed by C
(Cyan), M (magenta) and Y (yellow)
Various color adjustment keys, plus (increase) key and minus (decrease) key (both are collectively referred to as increase / decrease keys)
And for each color, the -7 key to
The adjustment of about 15 steps of the +7 key is set. However, it is very difficult to judge how the image changes (corrects) when you increase or decrease how much the color is, and it takes a long experience and knowledge to be able to judge this appropriately. is necessary. Therefore, for an operator with insufficient experience and skill, it takes a very long time to properly correct an image by the verification.

[0009] In addition to the color adjustment keys, the photo printer includes a gradation (γ) adjustment key, a contrast adjustment key, a density adjustment key, and a sharpness enhancement (sharpness).
Various adjustment keys such as keys are set. As with color correction, corrections using these keys are used in combination with the increase / decrease keys to correct the image, but it is also necessary to make adjustments and how much to correct the image. In order to make an appropriate image correction, a high level of knowledge and experience is required.

SUMMARY OF THE INVENTION An object of the present invention is to solve the above-mentioned problems of the prior art, and even if an operator does not have sufficient knowledge and experience in correcting an image in a photo printer verification or the like, a simple operation is possible. Another object of the present invention is to provide an image correction method which can correct an image quickly and appropriately, and can perform an efficient correction operation according to the sensitivity of the operator.

[0011]

SUMMARY OF THE INVENTION In order to achieve the above object, according to the present invention, a linguistic expression indicating a state of an image or a correction direction of the image, and an image correction condition corresponding to the linguistic expression are set in advance. Providing an image correction method, comprising: inputting the linguistic expression as a correction instruction according to the image; and correcting the image using the image correction condition corresponding to the input linguistic expression. Is what you do.

Here, a plurality of image correction conditions having different intensities are set for the linguistic expression, and a plurality of corrected images corrected by the respective image correction conditions are set according to the input linguistic expression. Reproduction is preferred. In addition, the relationship between the language expression first input to the image and the final image correction is totaled, and the image correction condition corresponding to the language expression is updated according to the totaling result. Is preferred.

Preferably, the image scene of the image is classified using the image feature amount of the image, and the totaling is performed for each classified image scene. Further, the image is classified according to at least one of a type of printing method, a type of print paper, a type of printer, or a type of printer, a printer operator, and a photo shop for reproducing the image as a photographic print. Thereafter, the relationship between the first input linguistic expression and the final image correction is totaled for each classification, and the image correction condition corresponding to the linguistic expression is updated according to the total result. Is preferred.

A plurality of image correction conditions having different image correction algorithms are set for the language expression. The image correction is performed according to the selected image correction conditions, and the number of times the image correction conditions are selected is reduced. It is preferable that the totaling is performed, and the priority of each of the plurality of image correction conditions is updated according to the result. In addition, it is preferable to update a condition setting algorithm of the image processing according to the totaling result.

[0015] Preferably, the image processing includes density control according to a result of extracting the main part, and the image correction according to the language expression preferably includes recalculation of a density control amount according to the result of extracting the main part. . In the image correction processing, it is preferable to switch between a language input mode for inputting the language expression and a numerical value input mode as the correction instruction.

[0016]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An image correcting method according to the present invention will be described below in detail with reference to a preferred embodiment shown in the accompanying drawings.

FIG. 1 is a block diagram showing one embodiment of a digital photo printer for implementing the image correcting method of the present invention. The digital photo printer shown in FIG.
A photo printer 10 basically includes a scanner 12 that photoelectrically reads an image photographed on a (photographic) film F, image processing of image data read by the scanner 12, and control of the entire photo printer 10. And the like, and a photosensitive material (photographic paper) is exposed (printed) with a light beam modulated in accordance with the image data processed by the image processing device 14, developed and output as a (photographic) print. And a printer 16.

The image processing apparatus 14 includes an operation system 18 having a keyboard 18a and a mouse 18b for inputting various processing selections and instructions, instructions for color / density correction, etc., and a simulation for verification. A display 20 for displaying an image or the like is connected. In a preferred embodiment, the photo printer 10 shown in the drawing is capable of inputting a voice for an image correction instruction, and the operation system 18 has a microphone 18c for this.

The scanner 12 is a device that photoelectrically reads an image of each frame photographed on the film F.
A variable aperture 24, a color filter plate 26 for sequentially applying R (red), G (green) and B (blue) color filters to the optical path;
A diffusion box 28 for making the reading light incident on the film F uniform, an imaging lens unit 32, a (area) CCD sensor 34 for reading one frame image of the film, an amplifier (amplifier) 36, and an A / D (analog). / Digital) converter 38.

In the illustrated example of the photo printer 10,
Scanner 1 according to the type and size of film such as new photo system (Advanced PhotoSystem) and 135 size
A dedicated carrier (not shown) that can be mounted on the main body 2 is prepared, and by changing the carrier, it is possible to cope with various films and processes. Each frame (image) photographed on the film F is sequentially conveyed to a predetermined reading position by a carrier when reading.

In such a scanner 12, the reading light emitted from the light source 22, the light amount is adjusted by the variable diaphragm 24, the color is adjusted by passing through the color filter plate 26, and the reading light diffused by the diffusion box 28 is applied to the film F. Incident,
By transmitting the light, projection light carrying an image of the frame captured on the film F is obtained. The projection light of the film F is transmitted to the CCD sensor 34 by the imaging lens unit 32.
An image is formed on the light receiving surface of the CCD, is photoelectrically read by the CCD sensor 34, and the output signal is amplified by the amplifier 36.
The signal is converted into a digital signal by the A / D converter 38 and sent to the image processing device 14.

In the scanner 12, such an image reading is performed three times by sequentially inserting the respective color filters of the color filter plate 26, so that one frame of image is read by R, G and B.
And read out. The scanner 12 reads an image photographed on the film F by two image readings: a pre-scan for reading at a low resolution and a main scan for obtaining image data of an output image. The pre-scan is performed under the pre-scan reading conditions set so that the images of all the films targeted by the scanner 12 can be read without the CCD sensor 34 being saturated. On the other hand, the main scan is performed under the main scan reading conditions set for each frame from the pre-scan data so that the CCD sensor 34 is saturated at a density slightly lower than the minimum density of the image (frame). Therefore, for the same image, the output signals of the pre-scan and the main scan are
Resolution and output level are different.

It should be noted that a photo printer 1 utilizing the present invention
In the case of 0, the scanner 12 is not limited to the above-described configuration, and in addition, photoelectrically reads an image captured on a film by slit scanning using a line CCD sensor corresponding to reading of R, G, and B images. Without using a scanner and a color filter plate, each of R, G and B lights is sequentially incident on the film, and the image photographed on the film is converted to 3
Various scanners, such as a scanner that separates into primary colors and reads photoelectrically, are available. Further, in the photo printer 10 utilizing the present invention, not only the image photographed on the film F but also a CD-ROM which records the image data of the print output together with the print by the (digital) photo printer.
Print creation from image data recorded on a recording medium such as R, print creation from image data taken by various imaging devices such as digital cameras, print creation from image data supplied from a computer or communication means, etc. It is possible.

As described above, the output signal (image data) from the scanner 12 is output to the image processing device 14.
As shown in FIG. 2, the image processing device 14 (hereinafter, referred to as the processing device 14) includes a data processing unit 46, a Log converter 48, a prescan (frame) memory 50, a main scan (frame) memory 52, a press Can processing unit 54,
It has a main scan processing section 56 and a condition setting section 58. FIG. 2 mainly shows parts related to image processing. The image processing apparatus 14 further includes a CPU and the like for controlling and managing the entire photo printer 10 including the processing apparatus 14. Is done.

First, the R, G, and B output signals output from the scanner 12 are first processed by the data processing unit 46.
Predetermined processes such as DC offset correction, dark time correction, shading correction, etc. are performed, converted by the Log converter 48 into digital image data, and pre-scanned (image)
The data is stored (stored) in the pre-scan memory 50 and the main scan (image) data is stored in the main scan memory 52.

The pre-scan data stored in the pre-scan memory 50 is processed by the pre-scan processing section 54, and the main scan data stored in the main scan memory 52 is processed by the main scan processing section 56. The pre-scan processing unit 54 includes an image processing unit 62 and a signal conversion unit 64. On the other hand, the main scan processing section 56 has an image processing section 66 and a signal conversion section 68.

Image processing unit 62 of prescan processing unit 54
And the image processing unit 66 of the main scan processing unit 56,
It is a part that performs image processing on an image (image data) read by the scanner 12 according to the setting of the condition setting unit 58. Both perform basically the same processing except that the pixel densities of the image data to be processed are different.

The image processing in the image processing section 62 and the image processing section 66 are various kinds of known image processing, and include, for example, gray balance adjustment, gradation adjustment, density adjustment, electronic scaling processing, sharpness (sharpening). ) Processing, graining suppression processing, smoothing processing, dodging processing (giving a dodging effect in a direct-exposure photo printer by compressing image data while maintaining halftones), and red-eye correction.

Each of these processes may be performed by a known method, such as a process operation (algorithm), a process by an adder or a subtractor, a process by an LUT (lookup table), a matrix (MTX) operation, a process by a filter, and the like. Are appropriately combined. For example, gray balance adjustment, density adjustment, and gradation adjustment are performed by a method using an LUT created according to an image feature amount.
In the method using the X operation, the sharpness processing is performed by dividing an image into frequency components, multiplying a luminance signal obtained from middle and high frequency components by a sharpness gain (sharpness correction coefficient), and obtaining luminance information. Is added to the low frequency component, and the smoothing process is a method of averaging image data using a mask.

The signal conversion section 64 of the prescan processing section 54
Represents the image data processed by the image processing unit 62,
3D (three-dimensional) -LUT conversion and display 2
This is a part to be image data corresponding to the display by 0.
On the other hand, the signal conversion section 68 of the main scan processing section 56 converts the image data processed by the image processing section 66 into a 3D-LU.
It is a part that converts the image data into image data corresponding to the image recording by the printer 16 by conversion with T or the like.

The image processing and processing conditions performed by the pre-scan processing section 54 and the main scan processing section 56 are set by a condition setting section 58. The condition setting unit 58 includes a setup unit 70, a key correction unit 74, and a parameter integration unit 76.

The setup section 70 determines the reading conditions of the main scan, the image processing conditions of the image processing section 62 of the pre-scan processing section 54, and the image processing section 66 of the main scan processing section 56, and the like. Specifically, the setup unit 70 creates a density histogram from the pre-scan data, and calculates the average density, highlight (minimum density)
Calculation of image features such as image and shadow (maximum density), extraction of main subjects, etc., setting of scanning conditions for main scan as described above, and furthermore, density histogram, image feature, operator's instructions, etc. The image processing unit 62
And the image processing conditions at 66 are determined and supplied to the parameter integration unit 76.

The key correction unit 74 selects an image correction condition according to an image correction instruction such as color correction, density correction, or contrast correction input from the microphone 18c or the like of the operation system 18 at the time of verification or the like. It is a part to be supplied to the parameter integration unit 76. Here, the key correction unit 74 determines the state of the image, specifically, an improper point of the image (for example, “Dark”) for the expected finish image (simulation image) displayed on the display 20 for performing the verification. , “Thin”, etc.), and the degree (for example, “somewhat”, “very”, “very”, etc.) are set in linguistic expressions. Further, an image correction condition for correcting each language expression is set. This will be described in detail later.
Further, as described above, the photo printer 10 in the illustrated example is
The key correction unit 74 has a voice input function of an image correction instruction, and recognizes image correction for the simulation-input simulation image using a known voice recognition unit.

The parameter integration section 76 receives the image processing conditions calculated by the setup section 70 and sets them in predetermined portions of the image processing sections 62 and 66. Also, as a result of the test, when an instruction for image correction is input by the operator, calculation of image data processing conditions for performing the image correction and calculation of image data processing conditions are performed in accordance with the image correction conditions supplied from the key correction unit 74. The settings for the image processing units 62 and 66 and the correction of the image processing conditions previously set for the image processing units 62 and 66 are performed.

As described above, the image data processed by the pre-scan processing unit 54 of the processing device 14 is displayed on the display 2.
The image data processed by the main scan processing unit 56 is sent to the printer 16. In the photo printer 10 using the present invention, the display 20 is not particularly limited, and various known displays such as a CRT (Cathode Ray Tube), an LCD (Liquid Crystal Display), and a plasma display can be used.

On the other hand, the printer 16 exposes a photosensitive material (printing paper) in accordance with the supplied image data to record a latent image, and performs a predetermined process on the exposed photosensitive material. And a processor (developing device) that outputs the image as a (finished) print. In the printer, for example, after cutting a photosensitive material into a print length, a back print is recorded, and then, three types of light beams of R exposure, G exposure, and B exposure are emitted in accordance with image data output from the processing device 14. By modulating and deflecting the light in the main scanning direction, and conveying the photosensitive material in a sub-scanning direction orthogonal to the main scanning direction, the photosensitive material is two-dimensionally scanned and exposed to record a latent image and supplied to a processor. . After receiving the photosensitive material, the processor performs a predetermined wet development process such as color development, bleach-fixing, washing with water, etc., and forms a print by drying.
Sort and accumulate in predetermined units such as main units.

Hereinafter, the operation of the photo printer 10 will be described, and the image correction method of the present invention will be described in more detail.

The operator mounts the carrier corresponding to the film F on the scanner 12, sets the film F on the carrier 30, inputs the print size to be created, and inputs various necessary instructions and information. Instruct to start. In response to this, the scanner 1
2 is determined to be in a predetermined state, and further, after the scanner 12 is adjusted to the prescan reading conditions such as the aperture value of the variable aperture 24, the prescan is started, and the first frame of the film F is moved to the carrier. To a predetermined reading position.

When the first frame of the film F is conveyed to the reading position, the reading light emitted from the light source 22, modulated by the variable aperture 24 and the color filter plate 26, and diffused by the diffusion box 28 is transmitted to the reading position. Projection light, which enters and transmits the first frame and carries an image captured in this frame, forms an image on a CCD sensor 34 by an imaging lens unit 32, is photoelectrically converted, is amplified by an amplifier 36, and is processed. It is sent to the device 14. As described above, in the scanner 12, such image reading is performed three times by sequentially inserting each color filter of the color filter plate 26, whereby the film F is read.
Are sequentially decomposed into three primary colors of R, G and B.

The pre-scan and the main scan are
The prescan and the main scan may be performed one frame at a time, or continuously for all frames or for a predetermined plurality of frames. In the following example, for clarity of explanation, 1
The case where the pre-scan and the main scan are performed frame by frame will be described as an example.

The output from the CCD sensor 34 is supplied to the amplifier 3
6, the digital signal is converted to a digital signal by the A / D converter 38, sent to the processing device 14, and subjected to predetermined processing such as offset correction by the data processing unit 46.
The digital image data is converted into digital image data at 8 and stored in the prescan memory 50.

When the pre-scan data is stored in the pre-scan memory 50, the setup section 70 reads it out, cuts out an image area for each frame, and further, as described above, sequentially obtains a density histogram for each frame. Is created and the image feature amount is calculated.
The main scanning conditions for each frame, such as the aperture value, are set and sent to the scanner 12. The setup unit 70
Further, image processing conditions are set according to the density histogram, the image feature amount, and the like, and are sent to the parameter integration unit 76. The parameter integration unit 76 sets the set image processing conditions at predetermined positions of the image processing unit 62 of the pre-scan processing unit 54 and the image processing unit 66 of the main scan processing unit 56.

The image processing conditions are determined by the image processing units 62 and 66.
Is set, the pre-scan data is read from the pre-scan memory 50, image-processed in accordance with the image processing conditions set in the image processing unit 62, and then displayed on the display 20 by the signal conversion unit 64. The converted image data is displayed on the display 20. This image has been subjected to image processing under the same image processing conditions as those of the image processing unit 66 of the main scan processing unit 56, and is a predicted image of the finish of an image reproduced in print, that is, a simulation image.

The operator looks at the simulation image displayed on the display 20 and conducts a test. When the operator determines that the image is appropriate, the operator issues an instruction of test OK and determines that the image needs to be corrected. If so, the user inputs an instruction to correct the image.

Here, the photo printer 10 in the illustrated example is
As a preferred embodiment, the instruction to correct the image can be input by voice input using the microphone 18c as described above. However, in the present invention, the input of the image correction instruction is not limited to the voice input, and various known methods can be used. For example, the keyboard 18a
A key corresponding to a linguistic expression described later may be provided and key input may be performed. Alternatively, a key may be input in a linguistic expression described later in words using a keyboard 18a, or a GUI (Graphical User Interface) using a mouse 18b and a display 20. )
, A linguistic expression to be described later may be input. Further, using a digital pen or the like, a GUI may be used to write a correction instruction in a linguistic expression described later. In this case, the correction portion may be cut out with an electronic pen or the like, a correction instruction may be input using characters or marks, and the characters or marks may be recognized by known recognition means, and the image may be corrected accordingly. Also, a plurality of input means may be selected and made available,
Further, a correction instruction may be input using a plurality of input means in combination.

Further, in the present invention, the correction instruction is not limited to the instruction using the simulation image.
For example, when a print instruction is issued from a customer to correct the color or density of a previously output print, the print is recreated. A correction instruction may be input using an image.

As described above, the key correction unit 74 of the photo printer 10 utilizing the present invention uses an improper point (hereinafter referred to as an expression) as a linguistic expression (hereinafter referred to as an expression) indicating the image state of a simulation image. , Points to be pointed out), and expressions expressing the degree of inappropriateness (hereinafter, referred to as the degree) are appropriately combined and set as an image state in advance. In the key correction section 74, image correction conditions for the expression of the set image state are also set. In the photo printer 10 shown in the figure, an instruction to correct an image is issued by inputting the expression of the image state.

The pointed item is not particularly limited, and various words expressing an inappropriate point of the image can be used. For example, expressions relating to density such as "dark" / "light",
Expressions relating to the color of the image such as “blue” / “red” / “yellow” are exemplified. Further, the indication may be a part of the image, such as “red eye” or “the center person has a bad complexion”. Also, as a degree,
A three-level intensity expression of “very” and “very” is illustrated. Note that the degree is not limited to such three levels, and may be two levels or four or more levels of intensity expression. That is, in this example, as a correction instruction according to the image state, an expression such as “(the image) is considerably dark” or “slightly yellow” is input.

As described above, in the key correction section 74, image correction conditions according to the set image state are set. As an example, an image correction condition that reduces the density when an indication of “dark” is input, and reduces the bluish when an indication of “blue” is input (for example, an increase in yellow ) Are set respectively. In addition, when a pointed out item “the center person has a bad complexion” is input according to the face extraction result (if face extraction is not set as the default in image processing, face extraction is performed) , Correct the skin color of the face area. The face area at this time may be specified by an operator using a GUI or the like. Also, if "Slightly" is entered as the degree, enter "Very" so as to make a weaker correction, and if "Pretty" is entered, enter "Very" so as to make a slightly stronger correction. In this case, an image correction condition is set for each pointed item so as to perform high-intensity correction.

In this embodiment, as a preferable mode, both the pointed item and the degree are input as the image state.
The present invention is not limited to this, and the image correction may be performed by inputting the image state only with the pointed-out item, and selecting the image correction condition by the key correction unit 74 accordingly. At this time, for example, as described above, when it is determined that the image is dark, "dark" is input, and when it is determined that the image is still dark after the correction, "dark" may be input again.

Hereinafter, a specific example will be described with reference to the input of the item indicated as “thin” as a representative example. However, this description can also be used for inputting (correction instructions) for other points (other points to be described later) regarding colors and densities.

With respect to a point indicated as "thin", image correction is performed to increase the image density, as a matter of course. In the apparatus of the illustrated example, as the concentration adjustment,
15 steps [−7 (key) to 0 (key) to +7 (key)]
Is set. (-) Indicates a decrease and (+) indicates an increase. Therefore, as shown in Table 1 below, for the pointed out item “thin”, seven steps (+1 to +7)
As shown by the underline, +1 is set for "slightly", +4 is set for "very", and +6 is set for "very". Is set as a typical default. That is, when "slightly light" is input as the image state, +1
When the density of the minute is input as "very light", a density increase of +4 minutes is performed, and when the input is "very light", a density increase of +6 minutes is performed as an image correction.

[0053]

[Table 1]

As will be described later in detail, since the image processing conditions are adjusted according to the input image correction conditions, the simulation image displayed on the display 20 changes accordingly. For example, when a correction instruction of “very light” is input, a simulation image in which the density is improved by +4 is displayed. When the operator determines that the density has become appropriate by the correction, the operator issues an instruction of verification OK (or inputs an image correction other than the density), and conversely,
If it is determined that it is not appropriate, "Slightly thin" again
And the like are further corrected until the image density becomes appropriate. If you want to make further corrections after correcting the simulation image,
The correction instruction may be input in a linguistic expression such as “more” or “excessive”.

In the above example, one simulation image is displayed, the correction is made, and the correction is repeated as necessary. In the present invention,
In response to the input of the image state, a plurality of simulation images having different correction intensities (that is, correction images) may be displayed, and the operator may select an appropriate image and correct the image accordingly. According to this, the time required for the verification and the amount of work of the operator can be reduced, and more efficient print creation can be performed.

For example, in the above-described example, in the image evaluation of “very thin”, the density increase of the +4 key is set by default, so that the density increase is performed together with the simulation image in which the +4 key is corrected. The display 20 also displays a simulation image obtained by performing the +3 key and the +5 key. Note that the correction strength and the number may be selectable. The operator looks at the displayed simulation images having different correction intensities, selects the image determined to be most preferable, and determines the correction condition for the density.

In this example, the simulation image is preferably displayed on the display 20 all at the same time, but if this is difficult due to the calculation speed, the size of the display 20, etc., one frame is required. Alternatively, a simulation image may be sequentially displayed for each of a plurality of frames. At this time, the simulation image is switched, for example, by N
It may be performed according to the instruction of G or the like.

As is clear from the above examples, in the present invention, the operator specifically corrects an image by an abstract language instruction without needing to specify which correction and what key. Can be.

Here, in the present invention, as a preferred embodiment, for each pointed item (correction instruction) input for image correction, the image state initially input for the simulation image and the final It is preferable to sum up the relationship with the corrected image and update the image correction condition corresponding to each image state. For example, it is assumed that the correction instruction input first is “very light”, and after the simulation image is corrected in accordance with the correction instruction, the operator further determines that the test is OK as a correction instruction “slightly light”. In this case, the final correction of the image density is “(+4)
+ (+ 1) = (+ 5) "
In, the first input “pretty”, +5
Is counted up by one (when displaying the plurality of simulation images, the column of the selected key is counted up). Note that, in Table 1, only the basic default neighborhood is shown for clarity of explanation, but in the present invention, there is no limitation on the key to be totalized in each instruction.

Such totaling is performed, for example, for a predetermined number of frames. The total amount is not particularly limited, and may be set to an arbitrary amount, for example, 1000 frames, one month, one season, one year, etc., according to the processing amount and scale of the lab store. Based on the result of this aggregation, the defaults for each of "Slight", "Pretty" and "Very" in the "Thin" indications are updated to best match the final corrections made I do. For example, in Table 1, “Somewhat” has a +1 key, “Pretty” has a +3 key,
“Very” has the largest number of +7 keys. Therefore, as a result of counting the predetermined number of frames, "Slight" updates the default to the +1 key as it is, "Pretty" updates the default to the +3 key, and "Very" updates the default to the +7 key.

This counting may be performed for each photo printer 10 (lab shop), or
May be performed for each operator who operates. With such a configuration, the tendency of image correction according to the policy or preference of the lab store or the operator can be set as a default. Therefore, after performing one image correction according to the simulation image, Again, the work of making corrections and the number of repeated corrections can be reduced, that is, the processing efficiency of print creation can be improved. Further, as will be described in detail later, such tallying may be performed for each region.

Such counting may be performed for each scene of the image, and the image correction condition corresponding to the image state may be updated for each image scene. The method of classifying the image scene is not particularly limited. For example, a normal scene / overexposure using at least one of a density distribution pattern obtained from a density histogram and a density difference between an image center portion and a peripheral portion. Classification of (over) scene / underexposure (under) scene is exemplified. In addition, scene classification based on scene estimation performed using a face extraction result or a subject extraction result (for example,
Portrait and landscape classification) are also suitable. Furthermore, scene classification such as person / landscape / night scene / under scene / high contrast scene may be performed by an operator's input. These may be used in combination of two or more.

Further, such classification and aggregation are performed in accordance with at least one of the following cases, that is, any one or a combination of any two or more of them, and the association between the linguistic expression and the image correction condition is performed. You may give feedback. That is, when an image is reproduced as a photographic print, at least one of the following for each printing method type, print paper type, printer type or printer type, printer operator, and lab store (photo store)
After classifying the images according to the type, the relationship between the linguistic expression that was input first and the image correction that was finally made is tabulated for each classified image, and the linguistic expression is supported according to the tabulated result. May be updated. Here, examples of the printing method type include a silver halide photographic printing method and an ink jet printing method, and examples of the printing paper type include a paper type and a surface type. The reason why the classification and the tabulation are performed according to the type of the printer or the type of the printer is that countermeasures can be taken for the device difference.

Further, the algorithm for setting the image processing conditions performed by the setup unit 70 may be updated using the total result of the image correction. Note that the total amount may be based on the above-described example.

For example, as shown in Table 2, the finally applied corrections (in the above example, +5
Increase in concentration). Here, in Table 2, 0
Is the case where the concentration correction by the test is not performed, that is, the case where the concentration adjustment condition set by the setup unit 70 is appropriate. As described above, in the illustrated example, a total of 15 levels of concentration correction from -7 to +7 are set, but correction exceeding ± 3 is performed by the test.
Since it is very small, it is omitted in Table 2. According to the tabulation in Table 2, the photo printer 10 (or the operator) often makes a correction of +1 minute with respect to the density rather than 0 which is uncorrected at the time of verification. That is, under the density adjustment conditions set by the setup unit 70, the density is often lower by +1 minute than the appropriate image.

[0066]

[Table 2]

Therefore, the setup section 70 performs the processing so that the image (image before correction by the test) which has been subjected to the image processing under the conditions set by the setup section 70 has a density that is +1 minute higher than the conventional one. Update the density adjustment condition setting algorithm. As a result, the frequency of image correction at the time of verification can be reduced, and the processing efficiency of print creation can be improved.

The above example relates to the correction of the color and the density of the image. However, it is also possible to correct the gradation and the image structure of the image according to the above-mentioned method. By summing up the correction results, it is possible to update the image correction conditions set by default and to update the algorithm for setting the image processing conditions set by the setup unit 70. For example, the points to be pointed out are "blurred", "tight", "sleepy", "hard", and the like, and the degree is the above-mentioned "slightly", "pretty" and "very". Is exemplified.

Regarding the pointed out point of "blurred",
Image correction that applies sharpness processing more strongly, but for the pointed out point of "tightness" Image correction that applies smoothing processing more strongly, and image correction that for the pointed out of "neutral" increases the contrast more However, for the item indicated as “hard”, an image modification for reducing the contrast is exemplified. Also, each modification
Changing the intensity according to the input degree is the same as in the previous example.

Here, the contrast can be adjusted, for example, by adjusting a characteristic curve of gradation adjustment (LUT corresponding thereto). As indicated by the arrow in FIG. 3, the image may be corrected by increasing the gradient at the intermediate value of the characteristic curve as the contrast is increased. The strength of the sharpness processing and the smoothing processing is N
It can be adjusted with the × N mask filter coefficient (sharpness gain). Depending on the level of image correction,
What is necessary is just to set this coefficient.

In these various image corrections, image correction conditions may be set by combining a plurality of processes. For example, for the pointed out item "blurred",
In addition to the above-described enhancement of the sharpness processing, image correction for enhancing the contrast may be performed. Correspondingly, as shown in Tables 3 and 4, "blurred"
In response to the pointed out item, M (for example, seven) correction patterns having different correction intensities in which levels of sharpness intensity and contrast enhancement intensity are set are set. At the same time, as shown by the underline in Table 4, for example, the image state “slightly blurred” is corrected by the pattern 1 and the image state “slightly blurred” is corrected by the pattern 4. For the image state of "very blurred", correction by pattern 6 is set by default. Then, according to the input image state, image correction is performed using a corresponding correction pattern. Also, in this example, the above-described aggregation may be performed.

[0072]

[Table 3]

[0073]

[Table 4]

As described above, in the present invention,
As an indication, input an image state such as "red eye"
Red-eye correction may be performed. Here, like red eye correction,
A plurality of types of correction methods (algorithms) are known, and if it is better to change the algorithm to be used according to the image state or the like, the algorithm adopted for a predetermined number of frames for each image correction is classified into a scene classification. And the priority of the adopted algorithm may be updated according to the result of the aggregation. In addition, the same may be performed by setting parameters of the algorithm.

For example, with regard to red-eye correction, an algorithm (algorithm 1) for extracting a red-eye area by judging color and shape, determining red-eye and correcting the red-eye, and a similar correction to algorithm 1 using a wider red-eye determination parameter The algorithm has three red-eye correction algorithms, namely, an algorithm (Algorithm 2) to perform the red-eye correction by extracting a red-eye area based on the spatial distribution shape of lightness (Algorithm 3). The scene is classified into three, large, medium and small, according to the ratio of the number of red-eye pixels). At this time, for the predetermined number of frames on which the red-eye correction processing has been performed, for each of the above scenes, finally, which algorithm was adopted is totaled, and as shown in Table 5 below, the priority order is set for each of the scenes. A red-eye correction process is performed by setting an algorithm (the priority is given to the left) and adopting an algorithm for each scene in descending order of priority.

[0076]

[Table 5]

Further, in the present invention, when the main subject is extracted and the density is adjusted, if the image is incorrect due to erroneous extraction, the fact is input as a correction instruction. The image correction may be performed by recalculating the density adjustment conditions in accordance with the correction. That is, when the main subject is extracted, the density adjustment condition is set using the average density of the candidate region of the main subject extracted by the processing. Here, when the density adjustment becomes inappropriate due to the erroneous extraction of the candidate area, the image state is input as a correction instruction, and the average density of the candidate area is input according to the correction instruction. Find the value that should be, and that value and more than a predetermined,
The density adjustment is performed using the average density of the remaining area, excluding the candidate areas with different densities.

For example, in a photograph as shown in FIG. 4, it is assumed that three areas indicated by dotted lines are extracted as face candidate areas as main subjects. That is, in this example, the closet is erroneously extracted as a face candidate area. The closet concentration is higher than the face. Therefore, the average density of the face candidate area increases, and the setup unit 70 sets image processing conditions (density adjustment conditions) so as to reduce the density of the face area. As a result, the density of the face region properly extracted becomes lower than that of the proper image.

The operator who sees such a simulation image responds to this by inputting the fact that the face candidate area is erroneously extracted as needed, and then, for example, corrects the image state of “very thin”. Enter as Thereby, by the relationship between the density of the extraction area and the ideal average density,
An erroneously extracted area that has an adverse effect on the density adjustment can be determined. In an image state in which the density of the face is “light”, it is necessary to perform correction to increase the density of the face area. Here, the density of the closet is usually higher than the ideal average density of the face, and if this area is a face, if the area is corrected to be darker, the image becomes more inappropriate. In other words, the correction applied to this area contradicts the instruction from the operator, and it can be determined that the closet area extracted as the face area candidate is erroneously extracted. In response to this, using the other face candidate extraction results, the density adjustment condition is calculated again and the image is corrected.

In the above example, by inputting the image state, an instruction to correct the image in the verification is input. However, the present invention is not limited to this, and the direction of image correction may be input in (language) expression as an instruction for image correction in the test. Also in this example, the image can be corrected in exactly the same manner as in the examples described above, except that the expression method is different.

There is no particular limitation on the expression of the correction direction. For example, the correction direction to be implemented, such as "slightly reduce bluish color", "increase the density considerably", "extremely sharpen the sharpness", and the like, are expressed. The correction instruction may be given by inputting only the correction to be performed, such as “suppress blueness” or “emphasize sharpness”. The image correction condition may be in accordance with the input of the image state described above. For example, if the correction instruction is “slightly reduce bluish color”, Y is set to +
If the value is “increase the density considerably”, increase the density by +4.
do it.

In the conventional image correction method, for example, when the image is entirely blue, yellow is increased by several keys,
To reduce the density, a specific operation corresponding to the image is required, such as lowering the density by several keys, and a high level of experience and knowledge is required to quickly perform appropriate correction. On the other hand, according to the present invention, the operator does not need to input a specific correction instruction at all, and inputs an image state such as “very blue” and a correction direction such as “suppress blue”. In this way, appropriate image correction can be performed. Therefore,
According to the present invention, even an operator who does not have sufficient knowledge and experience can quickly and appropriately perform image correction (verification) and perform print creation with high productivity.

As described above, when the operator inputs an image correction (input of an image state or
When the correction direction is input, the correction signal and the adjustment signal are sent to the key correction unit 74. It should be noted that the correction of the image at the time of the verification is not limited to one image state (correction direction). For example, if the above-mentioned “slightly light” correction is input and the density becomes appropriate, Then, a correction such as "quite sleepy" may be entered. The key correction unit 74 selects an image correction condition according to the input of the image correction, and sends it to the parameter integration unit 76. The parameter integration unit 76 sets processing conditions for performing the correction, corrects the previously set image processing conditions, and the like for the image processing unit 62 and the image processing unit 66 in accordance with the transmitted image correction conditions. I do. Therefore, the image displayed on the display 20 changes according to the input by the operator.

As a result of the test, the operator determines that the image displayed on the display 20 is appropriate (test OK).
Then, an output instruction is performed using the keyboard 18a or the like.

Thus, the image processing to be performed on the frame (image) is determined, and the main scan is started. When the main scan starts, the scanner 12 adjusts the aperture value of the variable aperture 24 to the reading condition of the main scan in which the set value is set. The main scan is performed in the same manner as the pre-scan except that the reading conditions and the resolution are different, and the output signal of the CCD sensor 34 is processed by the amplifier 36 and the A / D converter 38.
Processed by the data processing unit 46 of the processing device 14
The data is converted into main scan data by the converter 48 and sent to the main scan memory 52.

The main scan data is stored in the main scan memory 52.
The image data is read out by the main scan processing unit 56, subjected to image processing under the image processing conditions determined by the image processing unit 66, and then converted by the signal conversion unit 68 into image data for output. And a print is created by reproducing the image data.

In the various image correction processes described above,
The correction instruction input by the operator is performed by inputting a linguistic expression by voice using the microphone 18c of the operation system 18 or by inputting a linguistic expression using the keyboard 18a or the mouse 18b, but the present invention is not limited to this. A language input mode for inputting a language expression and a conventional numerical input mode for directly inputting a correction amount as a numerical value are provided as correction instructions by an input operator, and a function of switching between these two modes is provided. Is preferred. As such a mode switching function, for example, a method in which the operator manually switches the mode using the keyboard 18a or the mouse 18b of the operation system 18 or an ID card such as an employee card of the operator or a registration card such as a skill may be used. When the photo printer 10 is set in the photo printer 10, a mode set in advance according to the ID and skill of the operator is automatically read, and the photo printer 10 may be set to the read mode.

By doing so, a skilled operator (experienced) can use the numerical value input mode, and if the operator (newcomer) does not have sufficient knowledge and experience, input a linguistic expression. In this manner, the language input mode for performing the operation can be used, and thus the two modes can be properly used by a newcomer and a veteran. Of course, even for newcomers, if the level of skill (level) increases, the numerical input mode becomes easier to work. Alternatively, when detailed settings are desired, the numerical value input mode is more suitable for operation. Therefore, the two modes can be selectively used according to the knowledge, experience, and skill level of the operator, or according to the type of image to be corrected, the processing content, the operation, and the like.

In the above example, the results of image correction are tabulated for each photo printer 10 or each operator, so that the image correction conditions can be updated and the image can be updated according to the policies and preferences of the lab shop and the operator. The setting parameters (algorithms) of the processing conditions are updated. However, the present invention is not limited to this. For each region or season, the same calculation is performed, and the image correction conditions are updated. Good.

For example, if an algorithm for image processing and image correction (hereinafter simply referred to as an algorithm) is set in the photo printer 10 so that the print will be finished in the skin color desired by the Japanese, a printer of the same model is used. If you use the same algorithm in Europe, it may not be the color you like for Europeans. Also, if the algorithm is set based on shooting in Japan with sunlight, if you use the same algorithm with the same model of photo printer in a region where the latitude is significantly different from Japan, the sunlight will be different from Japan in this region Therefore, it is not possible to obtain prints of the same quality as in Japan.

In order to eliminate such inconvenience, the tendency of image correction is accumulated for each area and the algorithm is updated. For example, when an operator performs an image correction by a test, after classifying a scene, for a certain area, the operator's image correction tendency prefers to emphasize the contrast or prefers to see the contrast suppressed. If the tendency is divided, the parameters for that area are corrected according to the tendency.
Scene classification may be performed in the same manner as in the above-described example.

In addition, as an example of updating the algorithm,
Examples include how to apply contrast enhancement and resolution (sharpness), or change the target values of the color tone and density of the human skin color. This is because the difference between the person and the preferred skin color is considered to be different depending on the region, and the difference between the regions is considered. In Europe, compared to skin color based on Oriental people, it is preferable to reduce the density of human faces, and the operator tends to subtract the density. The algorithm is automatically updated so that the density becomes thinner.

As described above, the image correction by the operator is classified according to the number of frames to be processed, and the relevance to a specific scene classification is examined. Thereafter, in a predetermined scene classification, the statistical image correction tendency is reflected for each parameter for each parameter. Then, the algorithm is updated and optimized with a predetermined number of frames. At this time, it is preferable that the stored data is also updated to the latest data by a predetermined number of frames.

In the above processing, the type of scene classification and the data accumulation period may be set for each season. For example, in the vicinity of a ski resort, a “snow scene” classification that is determined according to a highlight ratio in a scene only in winter may be added, or a gradation characteristic that emphasizes whiteness may be applied to “snow scene”. It may be.

Further, in a certain area, each lab shop in the area is connected to a lab management center in the area by a network, and correction information by an operator in each lab shop is collected by the lab management center. The trend of the whole area is grasped, the image processing algorithm of the lab shop in the whole area is corrected at a certain timing, the updated algorithm is distributed to each lab shop, and the lab management center updates the algorithm of the area. May be performed on a regional basis.

When performing the above processing, data recording is performed in combination with a customer ID, and data is managed for each customer.
The parameters may be optimized for each customer.
It should be noted that the above processing may be performed on a home printer of each customer instead of the photo printer.

According to the above-described example, it is possible to appropriately reproduce an image in response to a change in a favorite tendency for each scene depending on a customer, a regional characteristic, a season, or the like. Further, at the time of manufacturing a photo printer, it is not necessary to set parameters one by one according to the tendency of each region where the product is to be exported, which facilitates the production and shipping of the product. Furthermore, the contents of the request may be divided between the print processing and the monitor display processing, or the registration data processing may be stopped or a specific request processing may be specified for a specific frame.

The image correction method of the present invention has been described in detail above. However, the present invention is not limited to the above embodiment, and various improvements and changes may be made without departing from the gist of the present invention. Of course. For example, in the above example, the image correction method of the present invention was applied to a digital photo printer, but the present invention is applied to a normal direct exposure photo printer that exposes a photosensitive material with projection light of an image photographed on a film. Is also available.

[0099]

As described above in detail, according to the present invention, even in an operator who does not have sufficient knowledge and experience in image correction in photo printer verification, etc., simple operations can be performed. In addition, it is possible to quickly and appropriately correct an image, and to perform an efficient correction operation according to the sensitivity of an individual operator, thereby improving the production efficiency of (photo) prints and the like. .

[Brief description of the drawings]

FIG. 1 is a block diagram of an embodiment of a digital photo printer that implements an image processing apparatus of the present invention.

FIG. 2 is a block diagram of an embodiment of the image processing apparatus of the digital photo printer shown in FIG.

FIG. 3 is a diagram illustrating an example of an image correction method according to the present invention.

FIG. 4 is a diagram for explaining another example of the image correction method of the present invention.

[Explanation of symbols]

 Reference Signs List 10 (Digital) photo printer 12 Scanner 14 Image processing device 16 Printer 18 Operation system 20 Display 22 Light source 24 Variable aperture 28 Diffusion box 32 Imaging lens unit 34 CCD sensor 36 Amplifier 38 A / D converter 46 Data processing unit 48 Log conversion Unit 50 prescan (frame) memory 52 main scan (frame) memory 54 prescan processing unit 56 main scan processing unit 58 condition setting unit 62, 66 image processing unit 64, 68 signal conversion unit 70 setup unit 74 key correction unit 76 parameters Integration department

Claims (9)

[Claims] 1. A linguistic expression indicating a state of an image or a correction direction of the image and an image correction condition corresponding to the linguistic expression are set in advance, and the linguistic expression is input as a correction instruction according to the image. And correcting the image using the image correction condition corresponding to the input linguistic expression. 2. A plurality of image correction conditions having different intensities are set for the linguistic expression, and a plurality of corrected images corrected by the respective image correcting conditions according to the input linguistic expression. The image correction method according to claim 1, wherein the image is reproduced. 3. The image processing apparatus according to claim 1, further comprising: summing up a relationship between a language expression first input to the image and a finally applied image correction; 3. The method according to claim 1, wherein the condition is updated. 4. The image correction method according to claim 3, wherein the image scene of the image is classified using the image feature amount of the image, and the totaling is performed for each of the classified image scenes. 5. A printing method, a printing paper type, and a printing method for reproducing the image as a photographic print.
After classifying the images according to at least one of a printer model or a printer, a printer operator, and a photo shop, the relationship between the first input linguistic expression and the final image correction is determined. 4. The image correction method according to claim 3, wherein the image correction conditions are totalized for each classification, and the image correction condition corresponding to the linguistic expression is updated according to the totalization result. 6. A plurality of image correction conditions having different image correction algorithms are set for the language expression, and image correction is performed according to the selected image correction condition.
The image correction method according to claim 1, wherein the number of times the image correction conditions are selected is totaled, and the priority of each of the plurality of image correction conditions is updated according to the result. 7. The image correction method according to claim 3, wherein an algorithm for setting a condition for the image processing is updated according to the totaling result. 8. The method according to claim 1, wherein the image processing includes density control in accordance with a result of extracting the main part, and the image correction in accordance with the linguistic expression includes recalculation of a density control amount in accordance with the result of extracting the main part. An image correction method according to any one of claims 1 to 7, 9. The image correction method according to claim 1, wherein, in the image correction processing, a language input mode for inputting the language expression and a numerical value input mode are switched as the correction instruction. .