JP2001222710A - Device and method for image processing - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an image processor capable of obtaining a high quality image more stably for images photographed in various scenes such as stably and appropriately reproducing the skin part of a person even in the case of photographing the person by using flash. SOLUTION: This image processor has a scene discriminating part discriminating an image scene from image information, a setting part which analyzes the image information and sets a compression processing condition with which image gradation is compressed through different intensities or characteristics in accordance with this analysis results and scene discrimination results by the scene discriminating part, and a processing part processing the image information in accordance with the compression processing condition set by the setting part to solve the above problem.

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 digital image processing used in digital photographic printers and the like, and in particular, reduces blurring of bright and dark areas even in flash photography scenes and backlight scenes. And an image processing method capable of reproducing a high-quality image.

[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 photosensitive material (photographic paper) by projecting an image of the film onto the photosensitive material. Exposure, so-called direct exposure, is the mainstream.

On the other hand, in recent years, a printing apparatus using digital exposure, that is, an image recorded on a film is photoelectrically read, the read image is converted into a digital signal, and then various image processing is performed. 2. Description of the Related Art Digital photo printers have been put to practical use, in which image data (latent images) are recorded by scanning and exposing a photosensitive material with recording light modulated in accordance with the image data, and the resulting images are printed.

A digital photo printer basically includes a scanner (image reading apparatus) that photoelectrically reads an image recorded on a film by reading light incident on the film and reading the projected light. A predetermined process is performed on the read image data or the image data supplied from the digital camera or the like, and the image data for image recording, that is, an image processing device that is used as an exposure condition, and an image data output from the image processing device. For example, a printer (image recording apparatus) that scans and exposes a photosensitive material by light beam scanning to record a latent image and develops the photosensitive material exposed by the printer to reproduce an image (finished) (Developing device).

According to such a digital photo printer, the image can be processed (optimized) by processing the image data as digital image data, so that gradation adjustment, color balance adjustment, color / color adjustment can be performed. Density adjustment,
By suitably performing a sharpness (sharpening) process or the like, a high-quality print that cannot be obtained by conventional direct exposure can be obtained. Moreover, according to the digital photo printer, an image photographed by a digital camera or the like can be output as a print. As an example of image processing by processing such image data, giving a dodging effect in printing by direct exposure is given.

[0006]

The photographing conditions of a photograph are various. For example, as in the case of flash photography or a backlight scene, the difference in brightness of an image, that is, from the minimum density to the maximum density of an image photographed on a film. (The difference between the highest density and the lowest density = dynamic range of image density) is often very wide. However, photosensitive materials (photographic paper)
Has a narrower reproducible density range than a film. Therefore, when a print is created by exposing (printing) a photosensitive material by a normal method using a film image having a wide dynamic range, the image may not be properly reproduced. For example, when a person is photographed in backlight, if a person is exposed so as to be a suitable image, a bright portion such as the sky will fly white in the image reproduced in the print, and conversely, the sky is preferable. When exposure is performed to form an image, a person is blackened.

Therefore, in a conventional printing apparatus using surface exposure, when a print is created from a film image having a wide dynamic range, so-called dodging is performed. Dodging means using a method such as inserting a light-shielding plate or an ND filter in the exposure light path. For example, if a print is made from a negative film, a bright portion where an image is likely to fly increases the exposure amount, Conversely, this technique is to obtain a print in which the entire density region of an image photographed on a film is properly reproduced by reducing the exposure amount in a dark portion where the image is likely to be crushed.

The present applicant has developed an image processing method and an image processing apparatus in a digital photo printer or the like which enable reproduction of an image having the same effect as dodging in such direct exposure by processing image data. It has been previously proposed (JP-A-10-13680).

In this image processing (hereinafter referred to as dodging processing for convenience), an image (image data) supplied from a scanner or the like is analyzed to determine an image state. A device that compresses the gradation of an image so that the entire density area photographed in the printer can be properly reproduced on an output device such as a printer, and compresses the dynamic range of the image so that the range is in accordance with the output device. It is.

Specifically, according to the determined image state,
The bright part (low-density part) and the dark part (high-density part) are compressed independently, linearly or non-linearly, without changing the intermediate density part of the image. For example, in a system in which the image data of a bright portion (a high-density portion of a negative film = a low-density portion of a reproduced image) in a shooting scene has a large numerical value, image data of a bright portion (large image data) in which an image is likely to fly Is reduced, and the image data of the dark part where the image is likely to be crushed is raised to increase the gradation. Therefore, according to this dodging process, an image of an extremely wide density region photographed on a film can be stably and appropriately reproduced on a photosensitive material.

In this dodging process, the lightest part (that is, white) and the darkest part (that is, black), which can be a photographic scene, are basically determined by analyzing the image so that white and black are appropriate in the reproduced image. The gradation of the image is compressed so as to be black. However, even if such a dodging process is performed, a necessary density region of an image may be skipped or crushed depending on a shooting scene. For example, when a person is photographed using a flash, the person's skin may be the brightest area in the photographed scene. In this case, even if the dodging process is performed, the person's skin is In rare cases, the image will fly white and become an inappropriate image.

SUMMARY OF THE INVENTION An object of the present invention is to solve the above-mentioned problems of the prior art, and to use the present invention in a digital photo printer or the like which gives the effect of dodging in a conventional direct exposure photo printer by processing image data. Image processing device, for example, even in the case of shooting a person using a flash, it is possible to stably reproduce the skin portion of the person properly, such as images shot in various scenes On the other hand, an object of the present invention is to provide an image processing method capable of more stably obtaining a high-quality image, and an image processing apparatus for implementing the image processing method.

[0013]

According to an aspect of the present invention, there is provided an image processing apparatus comprising: a scene determining unit configured to determine, from image information, a scene of an image carried by the image information; A setting unit configured to analyze and set a compression processing condition for compressing a gradation of an image with different intensities or characteristics according to the analysis result and the scene determination result by the scene determination unit; and a compression set by the setting unit. An image processing apparatus comprising: a processing unit configured to process image information according to a processing condition.

In the image processing apparatus according to the present invention, the scene discriminating section discriminates a scene of the image into a scene which does not require a minimum reproducible density and a necessary scene in a reproduced image, and the setting section comprises: The scene where the minimum density is unnecessary, it is preferable to set the compression processing conditions so as to compress the bright part of the image more strongly than the scene where the minimum density is required,
Further, it is preferable that the compression of the gradation of the image is performed separately for the bright part and the dark part of the image depending on the frequency component of the characteristic.

Further, according to the image processing method of the present invention, a scene of an image carried by the image information is discriminated from the image information, and different intensities or characteristics are determined according to the analysis result of the image information and the discrimination result of the scene. And an image processing method for compressing the gradation of an image.

Further, in the image processing method of the present invention, the scene of the image is discriminated into a scene which does not require the minimum density and a scene which does not need to be reproduced in the reproduced image. It is preferable to compress the gradation of the bright part of the image more strongly than the necessary scene.

Further, in the image processing apparatus and the image processing method of the present invention, preferably, the gradation of the image is compressed with a different intensity or characteristic for each image scene.

[0018]

DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, an image processing apparatus and an image processing method according to the present invention will be described in detail based on preferred embodiments shown in the accompanying drawings.

FIG. 1 is a block diagram showing an example of a digital photo printer using an image processing apparatus of the present invention for implementing the image processing method of the present invention. The digital photo printer (hereinafter referred to as a photo printer 10) shown in FIG. 1 basically includes a scanner (image reading device) 12 and
It has an image processing device 14 and a printer 16. The image processing apparatus 14 includes a keyboard 18a and a mouse 18 for inputting (setting) various conditions, selecting and instructing processing, and inputting instructions such as color / density correction.
The operation system 18 having b is connected to a display 20 for displaying a simulation image for verification, various operation instructions, and the like.

The scanner 12 is a device that photoelectrically reads an image photographed on the film F or the like one frame at a time. The scanner 12 has a white light source 22, a variable diaphragm 24, a color filter plate 26, and a reading light incident on the film F. Box 28, an imaging lens unit 32,
An image sensor 34 which is an area CCD sensor, an amplifier (amplifier) 36, and an A / D (analog / digital) converter 38 are provided.

In the photo printer 10, a scanner is used in accordance with the type, size, form (slide and film, etc.) of a film such as a new photo system (Advanced Photo System) or a 135-size negative (or reversal) film. A dedicated carrier that can be freely attached to the main body 12 is prepared, and by changing the carrier, it is possible to cope with various films and processes. An image (frame) photographed on a film and provided for print production is conveyed and held at a predetermined reading position by this carrier.

Here, at both ends in the width direction of the film F,
A bar code, such as a DX code, an extended DX code, or an FNS code, having various information such as a film type is optically recorded. Further, a magnetic recording medium is formed on the film of the new photographic system, and film information such as an ID number of the film, whether or not a flash is used at the time of photographing, the photographing date and time, and the like are recorded. The DX code and various types of information included in the magnetic recording medium are read by the carrier when the image of the film F is read, and various types of information are sent to the image processing device 14.

When reading an image photographed on the film F in such a scanner 12, the reading light emitted from the light source 22 and the light amount of which is adjusted by the variable aperture 24 enters the color filter plate 26 and is adjusted. After being illuminated and diffused by the diffusion box 28, the light is incident on the film F held at a predetermined reading position by the carrier and transmitted therethrough, thereby obtaining projection light carrying an image photographed on the film F. The projection light forms an image on the light receiving surface of the image sensor 34 by the imaging lens unit 32, and the image captured on the film F is photoelectrically read. The output signal of the image sensor 34 is amplified by the amplifier 36, converted into a digital signal by the A / D converter 38, and sent to the image processing device 14.

The color filter plate 26 is composed of R (red), G (green)
And a turret having B (blue) color filters,
Each color filter is inserted into the optical path of the reading light by being rotated by rotating means (not shown). In the illustrated example of the scanner 12, the color filters of the color filter plate 26 are sequentially inserted, and reading is performed three times, whereby the image captured on the film F is separated into three primary colors of R, G, and B. Read.

The scanner 12 reads an image photographed on the film F in two scans: a pre-scan for reading at a low resolution and a main scan for obtaining image data corresponding to print output. Do. The pre-scan is performed under pre-scan reading conditions set in advance so that the image of all the films targeted by the scanner 12 can be read without saturation of the image sensor 34. 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 image sensor 34 saturates at a density slightly lower than the minimum density of the image (frame). Therefore, the output signals of the prescan and the main scan are basically the same data except that the resolution and the output level are different.

The image data supply source for supplying image data to the image processing apparatus of the present invention is not limited to such a scanner 12. For example, the illustrated scanner 12 separates an image into three primary colors using a white light source and a color filter plate, but other than this, a light source that emits reading light of the three primary colors using an LED or the like individually. A scanner that reads an image by separating the image into three primary colors by using a scanner may be used. Alternatively, area C
Instead of a CD sensor, a scanner that reads a film image by slit scanning exposure using a three-color line CCD sensor may be used. Further, in addition to a scanner for photoelectrically reading a film, various kinds of images such as an image reading device for a reflection original, an imaging device for a digital camera, communication means such as a computer communication network, and a recording medium (drive thereof) such as a floppy disk. Image processing may be performed by receiving image data from a data source.

As described above, the output signal (image data) from the scanner 12 is output to the image processing device 14.
FIG. 2 shows a block diagram of the image processing device 14. As shown in FIG. 2, the image processing apparatus 14 (hereinafter, processing apparatus 1)
4) is a data processing unit 46, a Log converter 48,
A pre-scan (frame) memory 50, a main scan (frame) memory 52, a condition setting unit 54, a pre-scan processing unit 56, and a main scan processing unit 58 are provided.

Although FIG. 2 mainly shows parts related to image processing, the processing device 14 also controls and manages the entire photo printer 10, and other parts than those shown in FIG. , A CPU for controlling the whole, a memory for storing information necessary for the operation of the photo printer 10, and the like.
Further, the operation system 18 and the display 20 are connected to each part via the CPU or the like (CPU bus).

The R, G, and B output data output from the scanner 12 are converted by the data processing unit 46 into DC data.
Predetermined processing such as offset correction, dark time correction, and shading correction is performed. Next, the output data processed by the data processing unit 46 is Log-converted by a Log converter 48 using, for example, an LUT (look-up table), and is converted into digital image (density) data. Is stored in the pre-scan memory 50,
The main scan (image) data is stored in the main scan memory 52,
Each is stored.

The condition setting unit 54 includes a pre-scan processing unit 5
6 (the image processing unit 68) and the main scan processing unit 58
The image processing section 72 sets image processing conditions for each image (frame).
It has a setup unit 62, a key adjustment unit 64, and a parameter integration unit 66.

The scene discriminating section 60 is a section for discriminating the scene of the image using the prescan data stored in the prescan memory 50. In the present invention, the scene discriminating unit 60 basically determines a scene that does not require the minimum density (Dmin) reproducible by the printer 16 (hereinafter referred to as a person high-con scene), and other scenes that require this minimum density. A scene (hereinafter, referred to as a normal scene) is determined for each image. A person high-con scene means that an important part, such as a person's skin (especially a face), such as a portrait of a flash, which has a high image contrast and flashes, flies white on an image reproduced in a print. This is a possible scene. In other words, this is a scene in which proper reproduction of the print Dmin may be sacrificed in order to properly reproduce an important part such as a person's face.

As will be described in detail later, in the image processing apparatus and method according to the present invention, a person high-con scene has a bright part (low density) by dodging processing in image processing units 68 and 72 to be described later, as compared with a normal scene. ) Increase the area compression.

There is no particular limitation on the method of scene discrimination in the scene discriminating section 60, and various methods using photographing conditions, image analysis, image feature amounts, and the like as parameters can be used.

The parameters include, for example, whether or not flash shooting is performed and the result of face candidate extraction (center, shape, etc.).
And the color / density of the face candidate area, the exposure state (for example, undercuts and overs, as an example, a human high-con scene often has underscores), the density histogram shape (density distribution),
And image feature amounts such as the maximum density and the minimum density (image contrast) of the image. Whether or not to use flash photography may be heard from the client who made the print. In the case of a new photo system, information on whether or not a flash was fired at the time of photography recorded on the magnetic recording medium may be used. The face extraction may be performed by a known method using center detection, skin color detection, shape detection, and the like. The creation of the density histogram and the calculation of the image feature amount such as the image contrast are usually performed when various image processing conditions to be described later are set (in the illustrated example, set by the setup unit 62). May be. Further, if a density histogram is created, the exposure state can be determined.

Examples of the scene discrimination method using these parameters as parameters include, for example, an image (flash photographing, where the face is at the center, or the face area is equal to or more than a predetermined% of the whole), which satisfies predetermined conditions that are appropriately set. ) Is determined as a person high-con scene. It is also preferable to use a method of quantifying each parameter, creating a discriminant function, setting a threshold, and the like, and discriminating whether the image is a person high-con scene or a normal scene. Alternatively, instead of a binary determination as to whether or not the image is a person high-con scene, the probability of being a person high-con scene, such as “the probability of being a person high-con scene is 80%” or “10% of the same” is used. , Scene determination. The parameters used for the scene determination, the determination method, the determination criterion, and the like may be appropriately set experimentally for each model of the photo printer 10, for example.

The setup section 62 is a section for setting image processing conditions for each frame by image analysis using prescan data. Specifically, using the pre-scan data, creation of a density histogram of an image, and
Calculation of image feature amounts such as LATD (Large Area Transmission Density), a predetermined frequency% point of a density histogram such as a minimum density and a maximum density, and an average density is performed. Using the scene determination result (used at the time of dodging processing), the reading conditions of the main scan, various LUTs and the like in the image processing unit 68 and the image processing unit 72 are determined by a known method such as a matrix operation or an image processing algorithm. An image processing condition such as a matrix operation expression is calculated. This will be described in detail later.

The key adjustment unit 64 includes a density adjustment key, an adjustment key for each color of C (cyan), M (magenta), and Y (yellow), a gradation (γ) adjustment key, and a bright part set on the keyboard 18a. Various adjustment keys such as an adjustment key, a dark area adjustment key, a sharpness adjustment key, a saturation adjustment key, and a mouse 1
An image adjustment amount is calculated in accordance with various adjustment instructions and the like input by 8b and supplied to the parameter integration unit 66. The parameter integration unit 66 receives the image processing conditions set by the setup unit 74, and sets the image processing conditions in the image processing unit 68 of the pre-scan processing unit 56 and the image processing unit 72 of the main scan processing unit 58. The parameter integration unit 66 further adjusts (corrects) image processing conditions set for each part according to the adjustment amount of the image calculated by the key adjustment unit 64, creates correction conditions for performing the adjustment, and performs both adjustments. Make settings for the processing unit.

In the illustrated processing apparatus, basically,
The pre-scan data stored in the pre-scan memory 50 is processed in the pre-scan processing unit 56, and the main scan data stored in the main scan memory 52 is processed in the main scan processing unit 58. The pre-scan processing unit 56 includes an image processing unit 68 and a data conversion unit 70. On the other hand, the main scan processing section 58 includes an image processing section 72 and a data conversion section 74.

Image processing section 68 of prescan processing section 56
And the image processing unit 72 of the main scan processing unit 58 basically differs from the image processing unit 72 in that the pixel density of the image data to be processed is different.
It has a similar configuration and performs similar processing. Accordingly, in the following description, the image processing unit 72 of the main scan processing unit 56 will be exemplified and described.

The image processing section 72 (68) includes a first LUT 7
6. First matrix computing unit (hereinafter referred to as MTX) 78
And a dodging processing unit 80. As described above, the image processing conditions in each processing unit are set by the condition setting unit 54.

The first LUT 76 is the main scan memory 52
The image data stored in the (pre-scan memory 50) is read out, and gray balance (color balance) correction, density correction, and gradation correction are performed. LUTs for performing the respective corrections are connected in cascade. .
The first MTX 78 performs color correction on an image (image data) processed by the first LUT 76. That is,
Matrix calculations set according to the spectral characteristics of the film F, the spectral characteristics of the photosensitive material (photographic paper), the characteristics of the development processing, and the like are performed so that the image output to the print has an appropriate color, and color correction is performed. Do.

The dodging processing unit 80 is a part for performing dodging processing (giving a dodging effect in direct exposure printing by processing image data).
A low-pass filter (LPF) 84 and a second LUT 86
And an adder 88. When performing the dodging process, the image data processed by the first MTX 78 is sent to the second MTX 78 and the adder 88. On the other hand,
When the dodging process is not performed, the first MTX 78 and the data conversion unit 74 (70) are connected by bypass, and the image data is sent from the first MTX 78 to the data conversion unit 74.

The second MTX 82 generates light / dark image data (light / dark image data) of this image from the R, G and B image data sent from the first MTX 78. There is no particular limitation on the method of generating the light and dark image data.
And taking one third of the average value of the image data of B and
A method of converting color image data into light and dark image data using the YIQ rule is exemplified. For example, the expression “Y = 0.
With 3R + 0.59G + 0.11B, only the Y component defined by the YIQ may be calculated from the R, G, and B image data to obtain bright and dark image data.

The LPF 84 processes the light and dark image data generated by the second MTX 74 and extracts low-frequency components, thereby two-dimensionally blurring the light and dark image to obtain blurred image data of the read image. is there. LPF84
Is an FI that is normally used for generating blurred image data.
An R (Finite Impulse Respones) type LPF may be used, but an IIR (Infinite Impulse Respones) is used because a small circuit can generate large blurred image data.
It is preferred to use an LPF of the es) type.

Here, as described above, since the resolution of the image is different between the pre-scan data and the main scan data, if the processing is performed using the same LPF, the display 2
0 and the printer 16 reproduce images differently. Therefore, it is necessary to change the frequency characteristic of the LPF 84 between the pre-scan data and the main scan data according to the resolution. Specifically, the blur amount of blurred image data used for display on the display 20 may be reduced by the resolution ratio. That is, if the resolution ratio is m, the cutoff frequency of the pre-scan data is fc (p), and the main scan data is fc (f), "fc (p) ≒ mfc
(f) The LPF may be designed so as to satisfy "".

The blurred image data generated by the LPF 58 is stored in a second LUT 86 by a gradation compression LUT (hereinafter, referred to as a LUT).
(LUT).

As described above, the image density area that can be photographed on the film F is generally wider than the density reproduction area in printing. For example, in a backlight scene or flash photography, the image density area greatly exceeds the reproduction area of printing. An image in a density range may be taken.

FIG. 4 shows an example of a density histogram (film density) created by the setup section 62 from the read image data of the negative film. When the reproduction range in the image reproduced in the print is the density region indicated by the dotted line, all the pixels cannot be reproduced in the print in the images indicated by a to c. More specifically, the high-density portion exceeding the reproduction range (low reading signal intensity = large image data in the illustrated example), that is, the bright portion of the print (shooting scene) is skipped, and the low-density portion exceeding the reproduction range, The dark part of the print is crushed. Therefore, in order to obtain an image in which all of the image data is reproduced, the area from the highest density to the lowest density of the image (dynamic range of image density)
Must correspond to the reproduction range of the print. In other words, it is necessary to process the image data so as to compress the dynamic range by compressing the gradation of the image so as to give the same effect as the conventional dodging by direct exposure, and to correspond to the reproduction range of the print. There is.

In the illustrated example, the gradation of the main image data is compressed by adding the blurred image data processed by the compression LUT in the second LUT 86 to the main image data processed by the first MTX 78. And corresponds to a density reproduction area by printing. As a result, the gradation, density, and dynamic range of the bright and dark portions of the image data to be output are set to be appropriate, and an appropriate print in which a high-quality image is reproduced can be obtained even in a flash shooting scene or a backlight scene. It is possible to output stably. In other words, the compression LUT of the second LUT 86 is an LUT that performs processing on the blurred image data in order to obtain processing image data that makes the gradation and the like of main image data appropriate.

The compressed LUT is created by the setup unit 62 from the result of the image analysis and the result of the scene determination by the scene determination unit 60.

The setup section 62 has a basic LUT g shown in FIG.
_light and g _dark , which is a basic LUT for compressing a dark part shown in FIG.
The compression LUT of the bright part is changed to B × g _dark by A × g _light
The compression LUT of the dark part is set respectively. In FIG. 5, Tc is the optimum reproduction density of the face, and is usually about 0.4 to 0.7 in density D.

That is, the compression ratio f of the dodging process automatically set (auto-setup) by the setup unit 44
_auto (compression LUT) is determined by the following equation: f _auto = A × g _light + B × g _dark Here, the coefficients A and B are 0 ≦ A ≦
1, 0 ≦ B ≦ 1, and is appropriately determined according to the state of the image, specifically, the frequency of the light and dark parts, and the image feature amounts such as the maximum density, the minimum density, and the average density of the density histogram. , A compression LUT for a bright part and a dark part are set.

That is, for example, as shown in FIG.
In the case of the image of histogram b,
(Low density) is frequently used, and images such as nighttime flash photography
Can be determined. In the case of such an image, the light side pressure
Large shrinkage, ie g _lightLarge coefficient A
Set it properly. In night flash photography, etc.
The main subject such as an object is on the bright side of the histogram (on the film
High-density side of the image)
However, by performing such processing, the main
The density (brightness) of the subject of interest can be made appropriate.
Conversely, an image of a histogram c indicated by a two-dot chain line in FIG.
In the case of, scenes in snow or backlight
Can be determined. In the case of such an image,
Increase the compression ratio on the dark side, that is, g_darkPerson in charge of
Set the number B large. In backlit scenes, etc.
The subject is on the dark side of the histogram, resulting in a dark image.
However, by performing such processing,
Brightens the subject to create a high-quality image
You. Also, the maximum density and minimum density of the density histogram
Are significantly out of print gamut,
In order to reproduce the image of
It is necessary to increase the compression ratio.

As a method for determining the coefficients A and B, for example, in the density histogram shown in FIG. 4, the width of the density area deviating from the print reproduction area to the bright part side is represented by a, and the width of the density area deviating from the same dark part side is represented by a. FIG. 6A shows an LUT indicating the relationship between the density area a and the coefficient A which deviate from the bright side, as shown in FIG. 6 (B), an LUT indicating the relationship between the density region b deviating on the dark side and the coefficient B is prepared, and a method of determining A and B using this is exemplified.

Alternatively, as shown in FIG.
An LUT showing the relationship between the frequency on the dark side (cumulative% = X%) and the coefficient A, and the frequency on the light side (cumulative% = Y%) and the coefficient B as shown in FIG. LUT showing the relationship
8 is prepared from the histogram, and from the print reproduction limit P on the dark side and the print reproduction limit Q on the light side, the cumulative histogram is used. A method of calculating the cumulative percentage of the dark part and the cumulative percentage of the bright part and determining the coefficients A and B using the LUT shown in FIG. 7 is also exemplified.

Alternatively, the LU shown in FIGS.
A method of calculating coefficients A and B from T, taking the average of both, and determining coefficients A and B,
And a method of determining as B are also suitably exemplified. Furthermore, the coefficients A and B may be determined by selecting which LUT to use from the histogram.

Here, the image processing apparatus 14 according to the present invention
As described above, the scene determination unit 60 of the condition setting unit 54 determines whether the image is a person high-con scene or a normal scene, as described above. The setup unit 62 creates a compression LUT according to the result of the determination so that if the image is a person high-con scene, the bright part is compressed more than the normal scene. Also, in the case of a person high-con scene, the compression of the bright portion reduces the minimum density (white) of the image to the minimum density (Dm
in). In other words, in the case of a person high-con scene, white may not be appropriate in the image reproduced in the print.

As described above, according to the dodging process, the gradation of the image is compressed so as to correspond to the density reproduction range of the printer. The print can be properly reproduced. However, in a person high-con scene such as a scene in which a portrait is shot by flash, even if the dodging process is performed, the skin of the person may rarely fly white, resulting in an inappropriate image. Here, if the bright portion is strongly compressed by the dodging process, it is possible to stably and appropriately reproduce the skin and the like of a person even in a high-conscene scene of a person. However, if compression of bright parts is strongly applied, in a normal scene, for example, in a scene where it is necessary to appropriately reproduce white (the brightest part) such as snow in a snow scene or a cloud in a scene against the sky, it is difficult to print. The reproduced image becomes cloudy, which is not preferable as an image.

On the other hand, the image processing apparatus (and method) according to the present invention determines whether an image is a person high-con scene or a normal scene. In the case of a person high-con scene, the bright part is compressed more strongly. Therefore, even in the case of a person high-con scene, a main subject such as a face does not fly in the reproduced image. In the case of a normal scene, white such as snow and clouds can be properly reproduced. That is, according to the present invention, it is possible to stably output a high-quality print in which a high-quality image is reproduced regardless of a shooting scene.

Various methods can be used as a method of increasing the compression of a bright part in a person high-con scene compared to a normal scene. For example, in the case of a person high-con scene, the above expression “f _auto = A ×
In “g _light + B × g _dark ”, the coefficient A may be further multiplied or added by the coefficient a. FIG. 6A
As a method for determining the coefficient A corresponding to a normal scene, an LUT having a steeper slope may be used as the LUT for determining the coefficient A in the case of a person high-con scene. Alternatively, the LUT shown in FIG. 6B is used as a method for determining a coefficient A corresponding to a normal scene. In the case of a person high-con scene, an LUT in which a coefficient rises from a lower frequency region is used as an LUT for determining the coefficient A. May be used.

Further, a person high-con scene and a normal scene
With reference LUTs having different compression characteristics,
Bright compression of high-con scenes is stronger than normal scenes
May go. For example, in the case of a normal scene, FIG.
The reference LUT shown in FIG. _light)
In the case of a person high-con scene, it is shown in FIG.
The compression rise of the bright part is fast and the compression ratio is large.
Reference LUT (g_light-2) Is used.

Further, as described above, instead of a binary decision as to whether the image is a person high-con scene or a normal scene, the image is not displayed.
%, 10%, etc., the probability of a person being a high-con scene, etc., is used for scene discrimination, depending on the probability (reliability of discrimination). Conversely, when the compression strength is low, the compression strength of the bright portion may be changed continuously or discontinuously so that the compression approaches the normal scene. For example, a coefficient a is added to the coefficient A described above.
Is multiplied or added, the relationship between the probability of being a person high-con scene and the coefficient a may be converted into an LUT, and the coefficient A may be multiplied or added.

In the image processing apparatus of the present invention, the compression ratio by the dodging process is not limited to being automatically set by the set-up unit 62, but may be operated by a keyboard operation of the operator (for example, the tone adjustment key, the bright portion adjustment). Key, dark area adjustment key), the compression ratio f is adjusted by the operator.
_The compression LUT may be created by adding to _auto . Also,
In the image processing apparatus of the present invention, not only the gradation compression but also the gradation expansion may be performed in the case of an underexposed / overexposed image by the dodging process. The gradation may be expanded using, for example, a basic LUT having characteristics opposite to those of the basic LUT shown in FIG. In addition, regarding the adjustment of the compression ratio and the expansion of the gradation by the operator,
It is described in detail in JP-A-10-13680 by the present applicant.

The blurred image data processed by the second LUT 86 is sent to the adder 88. Adder 8
In step 8, the main image data processed by the first MTX 78 and sent directly to the adder 88 is added to the blurred image data. Thereby, the gradation of the main image data is compressed, and the same effect as that obtained by performing dodging by surface exposure is given to the reproduced image.

More specifically, in the blurred image data processed by the compression LUT in the second LUT 86, the bright portion (large image data) is minus and the dark portion is plus. Therefore, by adding this blurred image data to the main image data processed by the first MTX 78, the bright part of the main image data is small and the dark part is raised, that is, the gradation of the image data is compressed, and the gradation of the image data is compressed. The dynamic range is image data corresponding to the output density area in the printer 16. Here, in the present invention, the bright portion of the person high-con scene is compressed more as described above.

In this way, a dodging process is performed,
The pre-scan data of the compressed image data is sent to the data conversion unit 70, and the main scan data is sent to the data conversion unit 74. Note that various image processing units such as a sharpness processing unit may be arranged between the adder 88 and each conversion unit. The data conversion unit 70 of the prescan processing unit 56 converts the prescan data processed by the image processing unit 68 into image data corresponding to display on the display 20 using a 3D (three-dimensional) -LUT or the like. It is. On the other hand, the data conversion unit 74 of the main scan processing unit 58 similarly converts the main scan data processed by the image processing unit 72 using a 3D-LUT or the like, and converts the main scan data into image data corresponding to image recording by the printer 16. The part to be converted.

In the processing unit 14, the image data processed by the data conversion unit 66 of the prescan processing unit 54 is displayed on the display 20, while the image data processed by the data conversion unit 72 of the main scan processing unit 56 is processed by the printer 16.
, Respectively. The display 20 is not particularly limited, and various known display means such as a CRT (Cathode Ray Tube) and a liquid crystal display can be used.

On the other hand, the printer 16 exposes a photosensitive material (photographic paper) in accordance with the image data output from the main scan processing section 56 to record a latent image, and performs a developing process according to the photosensitive material ( Finish) Output as print. For example, after a photosensitive material is cut into a predetermined length corresponding to a print, three types of light beams, R exposure, G exposure, and B exposure, according to the back print recording and the spectral sensitivity characteristics of the photosensitive material (photographic paper) are applied. The latent image was recorded by modulating according to the image data (recorded image), deflecting in the main scanning direction, and recording the latent image by transporting the photosensitive material in the sub-scanning direction orthogonal to the main scanning direction. The photosensitive material is subjected to predetermined wet development processing such as color development, bleach-fixing, and washing with water, dried, printed, and then sorted and accumulated.

Hereinafter, the present invention will be described in more detail by describing the operation of the photo printer 10.

The operator requested to make a print of the film F loads the carrier corresponding to the film F into the scanner 12, sets the film F at a predetermined position on the carrier, and inputs necessary instructions such as a print size to be created. After that, an instruction to start printing is issued. Thereby, the aperture value of the variable aperture 24 of the scanner 12 and the image sensor 3
The storage time of No. 4 is set according to the prescan reading conditions, and then the carrier transports the film F, and the frame for performing print creation is transported to the reading position. When the film F is transported, magnetic information and a bar code such as a DX code recorded on a magnetic recording medium are read, and necessary information is sent to a predetermined portion. Next, pre-scanning of the frame is started, and as described above, the color filter plate 26
Are sequentially inserted, and the projected light in each is read by the image sensor 34, whereby the image of the frame is separated into three primary colors of R, G, and B and read photoelectrically.

The pre-scan and the main scan may be performed one frame at a time.
The pre-scan and the main scan may be performed continuously. In the following example, in order to simplify the description and clarify the operation, a case where the pre-scan and the main scan are performed frame by frame will be described as an example.

The output signal of the image sensor 34 by the pre-scan is amplified by the amplifier 36, and the A / D converter 3
8 and converted into a digital signal. The digital signal is
The data is sent to the processing unit 14, subjected to predetermined data processing by the data processing unit 46, converted into digital image data by the Log converter 48, and stored in the prescan memory 50.

When the pre-scan data is stored in the pre-scan memory 50, the setup section 62 of the condition setting section 54 reads out the pre-scan data and creates a density histogram of the image and calculates image feature amounts such as highlights and shadows. Then, the reading conditions for the main scan of the frame are set and supplied to the scanner 12. The pre-scan data is also read out by the scene determination unit 60, and determines whether the image is a person high-con scene or a normal scene as described above, and sends the determination result to the setup unit 62.

The setup section 62 further includes a density histogram, the calculated image feature amount, and a scene determination section 6.
0, or in accordance with an operator's instruction performed as necessary, the first LU.
LUT in T76, compressed L in second LUT 86
Image processing conditions for the image (frame) such as the UT in the pre-scan processing section 56 and the main scan processing section 58 are set and supplied to the parameter integration section 66. When the scene is a person high-con scene, the setup unit 62 creates the condition of the dodging process, that is, the compressed LUT so as to compress the bright part more strongly than the normal scene, as described above. is there. The parameter integration unit 66
The received image processing conditions are set in predetermined portions of the pre-scan processing section 56 and the main scan processing section 58.

When the image processing conditions are set, the display 20 becomes a verification screen, the pre-scan processing unit 54 reads out the pre-scan data from the pre-scan memory 50, and the pre-scan processing unit 56 executes the image processing corresponding to the image data. An image (pre-scan image) that is processed under the conditions and reproduces the pre-scan data is displayed on the display 20 as a simulation image (expected finish image).

Next, the operator performs verification by looking at the simulation image, and, if necessary, the keyboard 18a.
The color, density, gradation, etc. are adjusted using the adjustment keys and the like. The input of this adjustment is sent to the key adjustment unit 64, which calculates the adjustment amount of the image processing condition according to the adjustment input, and sends it to the parameter integration unit 66. The parameter integration unit 66 performs processing such as correction of image processing conditions set in both image processing units, calculation of correction conditions for performing the adjustment, and setting of both image processing units in accordance with the sent adjustment amount. Therefore, the image displayed on the display 20 also changes according to the correction, that is, the adjustment input by the operator.

When the operator determines that the image is proper (test OK), the operator issues an output instruction. In response, the image processing conditions are determined, and the main scan is started. If the test is not performed, the image processing conditions are determined when the image processing conditions are set by the parameter setting unit 66, and the main scan is started. The main scan is performed in the same manner as the pre-scan except that the accumulation time of the image sensor 34 and the aperture value of the variable aperture 24 are different.
Is amplified by an amplifier 36, is converted into a digital signal by an A / D converter 38, is processed by a data processing unit 46 of the processing device 14, is converted into a main scan data by a Log converter 48, and is converted into a main scan data. The data is sent to the memory 52. Next, the main scan memory 5 is operated by the main scan processing unit 58.
2, the main scan data is read out, and the image processing unit 72
The image processing is performed under the image processing conditions determined in, and further converted by the image data conversion unit 74 into image data for output corresponding to image recording by the printer 16, output to the printer 16, and a print is created.

Although the image processing apparatus and the image processing method of the present invention have been described in detail above, the present invention is not limited to the above-described embodiment, and various improvements and modifications can be made without departing from the gist of the present invention. Of course you can do it.

[0079]

As described above in detail, according to the present invention, even in the case of a person high-con scene, for example, when a person is photographed using a flash, the skin portion of the person can be stably removed. For example, it is possible to more stably obtain a high-quality image with respect to an image photographed in various scenes such as being able to appropriately reproduce.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating an example of a digital photo printer that uses an image processing apparatus according to an embodiment of the present invention for implementing an image processing method according to an embodiment of the present invention.

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

FIG. 3 is a block diagram illustrating an example of an image processing unit of the image processing apparatus illustrated in FIG. 2;

FIG. 4 is an example of a density histogram.

FIGS. 5A and 5B are examples of a basic LUT for creating a gradation compression LUT.

6A illustrates an example of an LUT for determining a coefficient applied to the basic LUT illustrated in FIG. 5A, and FIG. 6B illustrates an example of a coefficient applied to the basic LUT illustrated in FIG. 5B. An example of an LUT for performing this is shown below.

FIG. 7A shows another example of an LUT for determining a coefficient applied to the basic LUT shown in FIG. 5A, and FIG.
Shows another example of the LUT for determining the coefficient applied to the basic LUT shown in FIG.

FIG. 8 is an example of a cumulative histogram of density.

FIG. 9 is another example of a basic LUT for compressing a light portion.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 10 (Digital) photo printer 12 Scanner 14 (Image) processing device 16 Printer 18 Operation system 18a Keyboard 18b Mouse 20 Display 22 Light source 24 Variable aperture 26 Color filter plate 28 Diffusion box 32 Imaging lens unit 34 Image sensor 36 Amplifier 38 A / D converter 46 Data processing unit 48 Log converter 50 Prescan (frame) memory 52 Main scan (frame) memory 54 Condition setting unit 56 Prescan processing unit 58 Main scan processing unit 60 Scene discrimination unit 62 Setup unit 64 Key adjustment unit 66 Parameter integration unit 68, 72 Image processing unit 70, 74 Data conversion unit 76 First LUT 78 First MTX 80 Dodging processing unit 82 Second MTX 84 LPF 86 Second LUT 88 Adder

Claims (5)

[Claims] 1. A scene discriminator for discriminating, from image information, a scene of an image carried by the image information, analyzing the image information, and according to the analysis result and the scene discrimination result by the scene discriminator, A setting unit that sets compression processing conditions for compressing the gradation of an image with different intensities or characteristics; and a processing unit that processes image information according to the compression processing conditions set by the setting unit. Image processing device. 2. The scene discriminating section discriminates a scene of an image into a scene where a minimum reproducible density is unnecessary and a required scene in a reproduced image. The image processing apparatus according to claim 1, wherein the compression processing condition is set such that a bright portion of the image is compressed more strongly than a scene requiring the minimum density. 3. The image according to claim 1, wherein the gradation of the image is compressed separately for a light portion and a dark portion of the image depending on a frequency component of the characteristic.
Or the image processing device according to 2. 4. A method according to claim 1, wherein a scene of an image carried by the image information is determined from the image information, and different intensities or characteristics are determined according to a result of the analysis of the image information and a result of the determination of the scene.
An image processing method comprising compressing the gradation of an image. 5. A scene of an image which is distinguished from a scene which does not require a minimum density and a scene which requires a minimum density which can be reproduced in a reproduced image. The image processing method according to claim 4, wherein the gradation is strongly compressed.