JPH10210358A - Digital camera - Google Patents

Abstract

PROBLEM TO BE SOLVED: To suitably correct uneven illuminance, even when an oblique image is photographed after correcting it to a pseudo-front image. SOLUTION: An oblique image of the character information on a white board that is photographed by a CCD 20 is corrected to a pseudo-front image at an oblique image correction arithmetic part 322 and stored in an image memory 324. This corrected front photographed image is divided into plural blocks at a γ-characteristic setting part 325 for correction of its uneven illuminance, and the γ-characteristic is set to the pixel data included in every block, based on this pixel data. Then the corrected front image undergoes a γ-correction at a 2nd γ-correction part 328 for every block, based on the characteristic corresponding to the block. Thus, it is possible to set the proper γ-characteristic for every block by correcting the uneven illuminance, after an oblique photographed image has been corrected.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a digital camera which photoelectrically converts a subject light image into an electric image by an image pickup device such as a CCD, performs predetermined image processing, and records the image on a recording medium such as a hard disk card. It is about.

[0002]

2. Description of the Related Art Conventionally, in an image forming apparatus such as a digital copying machine, in order to enhance the clarity of information such as characters and figures copied on a recording sheet, an image captured by photoelectric conversion into an electric signal is taken. Image processing (gamma correction processing) is performed using a γ characteristic having a relatively large γ value (γ characteristic having a characteristic close to the binarization processing). In addition, in this gamma correction processing, generally, in order to reduce the influence of illuminance unevenness, as shown in FIG. 51, the captured image G is divided into a plurality of rectangular blocks B (1), B (2) in the sub-scanning direction. ,... B (n), and based on a histogram indicating the level distribution of the pixel data included in each block B (r), γ characteristics γ (1), γ (2),. n) and set each block B (r)
The pixel data in (r = 1, 2,... N) is subjected to gamma correction using the γ characteristic γ (r) for the block.

[0003] By this gamma correction, a white background portion above a predetermined level is uniformly converted to a constant white color, and a character portion (black character portion) below a predetermined level is uniformly converted to a constant black color. An image quality similar to the processing can be obtained.

Japanese Patent Laid-Open Publication No. Hei 6-113139 discloses that a captured image is divided into a plurality of partial image blocks, and a selected partial image block (partial image block of interest) and a plurality of adjacent partial image blocks. For each partial image block, a histogram of the level distribution of the pixel data included in the block is created for each block, and a threshold value of the partial image block of interest is set by a neural network using the data of the histogram.
An image binarization device that performs a binarization process on pixel data in a partial image block of interest using this threshold value is shown.

[0005]

By the way, a digital camera can freely control the image quality of an image photographed by image processing, so that the image quality of a photographed image can be appropriately adjusted according to the purpose of photographing and the type of subject. By doing so, there is an advantage that an image with more favorable image quality can be captured as compared with a camera that shoots on a normal silver halide film. For this reason, it is used not only for ordinary photography but also as a device for copying information such as characters and figures written on a whiteboard in a conference hall.

When photographing a whiteboard on which characters and figures are written by a digital camera, the main purpose of the photographing is to record information such as characters and figures on the whiteboard. On the other hand, similarly to the digital copying machine, it is desirable to perform gamma correction so that a white background portion (whiteboard portion) is skipped white to enhance the clarity of an information portion (character or graphic portion).

In this case, since the variation in character density and uneven illuminance on the whiteboard are large, the captured image is two-dimensionally divided into a plurality of blocks, and gamma correction is performed for each block to correct illuminance unevenness (shading). Correction).

That is, assuming that the whiteboard is illuminated by the ceiling light in the room and the sunlight outside the window, the unevenness of the illumination light causes uneven illuminance, and the entrance pupil of the photographing lens has an angle ω. The image of the off-axis object point incident at is cos ⁴
Due to the synergistic effect of the incident light amount distribution according to the so-called cosine fourth law, which becomes darker in proportion to ω, and the illuminance non-uniformity, the output distribution of an imaging device such as a CCD fluctuates greatly in the two-dimensional direction within the imaging surface.

For this reason, it is desirable to divide a captured image two-dimensionally into a plurality of blocks, and to perform gamma correction for each block in accordance with the illuminance in the block to perform illuminance unevenness correction. More preferably, when the γ characteristic set for each block greatly changes between adjacent blocks, a block size is set to avoid generation of a false line at a block boundary due to a sudden change in the γ characteristic. Is set as small as possible to set an appropriate γ characteristic for each block.

On the other hand, for example, in a conference hall, when information such as characters and figures on the whiteboard is photographed obliquely with respect to the whiteboard due to the positional relationship of the photographer's seat, the entire information such as characters is focused. Because of this, a perspective geometric distortion occurs in the captured image, and the readability of information is reduced. As a method of solving such a problem, the applicant corrects the perspective image distortion and shoots the image (ie,
Japanese Patent Application No. 8-10 proposes a digital camera capable of correcting an obliquely photographed image into a pseudo frontal photographed image (an image that looks like photographed from the front) and photographing it).
Nos. 1268 to 101271).

This digital camera calculates a photographing magnification at each pixel position in a photographing screen using an inclination angle of a subject with respect to the camera, a focal length of a photographing lens, and a subject distance. The geometrical image distortion is corrected by enlarging or reducing the partial image. For example, when shooting from the left diagonal direction with respect to the whiteboard,
Based on the center of the shooting screen, the left image is a close image, and the right image is a distant image. Therefore, the left image is scaled in the reduction direction, and the right image is scaled in the enlargement direction. As a result, a pseudo front image is corrected.

In general, when photographing a whiteboard in a conference hall, the whiteboard is hardly illuminated with uniform illuminance, and very few cases can be photographed from the front of the whiteboard. Normally, image processing based on both the illuminance unevenness correction function and the oblique image correction function described above is applied to a captured image of such a shooting scene.

In this case, since the oblique photographed image has a different photographing magnification and a different character size in the screen, if the illuminance unevenness correction is performed on the oblique photographed image, the number of characters included in the block (character density of the block) is reduced. It changes depending on the position of the block, and the variation in the shape of the histogram created for each block among the blocks increases. For this reason, a discontinuity in the white background level occurs due to the difference in the γ characteristic between adjacent blocks, making it difficult to perform appropriate illumination unevenness correction, and in some cases, a false line is generated at a block boundary. The problem described above also occurs.

In order to avoid this problem, it is conceivable to change the size of each block based on the photographing magnification in the photographing screen so that the character density of each block becomes substantially the same. Changes have to be made, which complicates the process.

The image binarizing apparatus described in Japanese Patent Application Laid-Open No. 6-113139 is mainly concerned with the binarizing process in a copying machine or a facsimile machine, like the conventional digital copying machine. Due to the structure, the document image is not taken in from an oblique direction, so that it does not have the oblique image correction function. Therefore, since there is no problem in performing both the illuminance unevenness correction and the oblique image correction as described above, there is no description or suggestion about such a problem.

SUMMARY OF THE INVENTION The present invention has been made in view of the above-mentioned problems, and is a digital camera capable of correcting illuminance non-uniformity which is suitable even when an obliquely shot image is corrected into a pseudo frontal shot image. Is provided.

[0017]

According to the present invention, there is provided an image pickup means having a pixel comprising a plurality of photoelectric conversion elements, photoelectrically converting a subject light image into a pixel signal, and taking in the image taken by the image pickup means. A digital camera comprising: a distortion correction unit configured to partially expand or reduce to correct a geometric distortion of the image, wherein the image correction unit divides the image corrected by the distortion correction unit into a plurality of small images. Means, γ characteristic setting means for setting a γ characteristic for the small image using pixel signals constituting the small image, and gamma correction using the γ characteristic set for each small image by the γ characteristic setting means. And gamma correction means for performing the following.

According to the above arrangement, the image captured by the imaging means is partially enlarged or reduced, and the geometric distortion occurring in the image is corrected. The image after the image distortion correction is divided into a plurality of small images, and a γ characteristic for the small image is set using an image signal constituting each small image.
Then, each pixel signal constituting the image after the image distortion correction is subjected to gamma correction using a γ characteristic corresponding to each small image,
Thereby, the illuminance unevenness correction of the captured image is performed.

According to the present invention, in a digital camera, a photographing lens for forming the light image of the subject on the image pickup surface of the image pickup means, and an angle setting for setting an angle between the image pickup surface of the image pickup means and the object surface. Means, distance measuring means for measuring a distance to a subject, and photographing in a photographing screen using the angle set by the angle setting means, the focal length of the photographing lens, and the subject distance measured by the distance measuring means. Calculating means for calculating a distribution of magnification, wherein the distortion correcting means corrects perspective distortion caused by oblique photographing by partially enlarging or reducing a captured image based on the distribution of photographing magnification. (Claim 2).

According to the above arrangement, an oblique image photographed with the object plane inclined at an angle θ with respect to the image pickup plane of the image pickup means is obtained by using the focal length of the photographing lens, the inclination angle θ, and the object distance. The distribution of the shooting magnification in the shooting screen is calculated,
By partially enlarging or reducing the captured image based on the calculation result, perspective distortion caused by oblique shooting is corrected. That is, the oblique image is corrected to an image as if photographed from the front.

The corrected photographed image is divided into a plurality of small images, and the γ characteristics for the small images are set using image signals constituting each small image. Then, gamma correction is performed on each pixel signal forming the image after the image distortion correction using the γ characteristic corresponding to each small image, and thereby illuminance unevenness correction of the captured image is performed.

[0022]

FIG. 1 is a perspective view showing the appearance of a digital camera according to the present invention. FIG. 2 is a rear view of the digital camera.

The camera 1 shown in FIG.
A D area sensor is provided, and image data captured by the CCD area sensor is recorded on a PCMCIA-compliant hard disk card (not shown). In addition,
In the present embodiment, a case where an area sensor is used as an image sensor will be described. However, a configuration may be adopted in which a line sensor scans a subject light image to capture image data.

In the camera 1, an image of a subject whose image pickup surface of the CCD area sensor is not parallel to the object surface (hereinafter, referred to as an oblique image) is an image of a subject whose image pickup surface and the object surface are parallel (hereinafter, a front image). ) (Hereinafter, this correction is referred to as oblique image correction).

That is, as shown in FIG. 3, for example, when a character or a figure drawn on the whiteboard 22 is photographed in a normal photographing mode from an obliquely left front position (a) with respect to the whiteboard 22, As shown in FIG. 4A, the image becomes an oblique image K in which the right end dimension is smaller than the left end dimension, as shown in FIG. When the image is captured in the image correction mode, the oblique image K is corrected to a front image K ′ almost as if shot from the front position (b) of the whiteboard 22, as shown in FIG.

Further, the camera 1 illuminates the whiteboard with illumination light (ceiling light) for a photographed image of information such as characters and figures written on the whiteboard (hereinafter, this type of binary information is referred to as character information). And sunlight outside the window) and CC
It has a function of correcting uneven illuminance caused by variations in sensitivity of the D area sensor.

That is, as shown in FIG. 5, for example, assuming that the whiteboard 22 is illuminated by the ceiling light in the room and the sunlight outside the window, the illuminance unevenness occurs due to uneven illumination light, and The image of the off-axis object point incident on the entrance pupil of the photographing lens at an angle ω darkens in proportion to cos ⁴ ω. As shown in FIGS. 6A and 6B, the output distribution of the CCD area sensor greatly varies two-dimensionally within the imaging plane. In the camera 1, when shooting in a character image mode described later, an appropriate γ characteristic is set for each pixel data or a plurality of pixel data units,
By performing gamma correction for each pixel data or each pixel data unit using the γ characteristic, the output fluctuation of the CCD area sensor which occurs two-dimensionally is reduced as much as possible.

In FIG. 6, the solid line indicates the case where nothing is written on the whiteboard 22, and the whiteboard 2
2 shows an output distribution of a white background portion, and a dotted line shows an output distribution of a character portion when a character is written on the whiteboard 22.

Here, the principle of the oblique image correction will be briefly described. Note that a one-dimensional image will be described for convenience of description.

FIG. 7 is a schematic structural view of the optical system of the camera 1. The optical system of the camera 1 includes a horizontally long rectangular CCD area sensor 20 (hereinafter referred to as CCD
Abbreviated as 0. ) Is arranged, and this imaging lens 2 and CC
A stop 21 is provided between the stop 21 and the lens D20. A light image such as a figure drawn on the white board 22 is transmitted through the photographing lens 2 and the aperture 21 and then formed on the image pickup surface of the CCD 20.

FIG. 8 is a view of the image pickup system in oblique photographing as viewed from directly above, in which the display surface of the whiteboard 22 (hereinafter, referred to as the object surface) and the image pickup surface of the CCD 20 have an angle θ (hereinafter, the inclination angle θ). ) Is shown only when the image pickup system is tilted.

In FIG. 8, a dashed line L is the optical axis of the taking lens 2. N0, N1 and N2 are CCDs 2 passing through points P, F and G on the whiteboard 22, respectively.
0 is a line segment parallel to the imaging plane. A point O is an intersection between the lens surface of the taking lens 2 and the optical axis L, a point Q is an intersection between the line segment N1 and the optical axis L, a point R is an intersection between the line segment N2 and the optical axis L, and D,
E is an extension of line segment N0 and line segment BF and line segment GC, respectively.
Is the intersection with

The light image between the FG on the white board 22 is C
An image is formed between the BCs on the imaging surface of the CD 20, but since the imaging surface and the object surface are inclined at an inclination angle θ, the light image BC formed on the imaging surface of the CCD 20 is equivalent to the image between the DEs. CC
The image is projected on the imaging surface of D20. CCD2
The imaging magnification at each of points A, B, and C on the imaging plane 0 is m _A , m _B , and m _C , and the object distance is D _A (= O
_{P), D B (= OQ} ), when the _{D C (= OR), m} B =
_{m A · OP / OQ = m} A · D A / D B, m C = m A · OP / O
Since R = m _A · D _A / D _C , m _B > m _A > m _C , and
The optical image formed on the imaging surface becomes an oblique image K as shown in FIG. 4A. Of the optical images BC, point A (the optical axis L and the imaging surface Intersection).

In the present embodiment, the photographing magnification mi (i = 1, 2,...) At each point between AC on the image pickup surface of the CCD 20.
n) and the imaging magnification mi ′ (i = 1,
2,... N ′), and enlarges the captured image of the optical image AC based on the imaging magnification mi, and reduces the captured image of the optical image BA based on the imaging magnification mi ′ to perform the oblique image correction. I have to.

Assuming that the subject distance at an arbitrary point between the BAs on the imaging surface of the CCD 20 is Di ', and the angle of view of the point (the angle between the line passing through the point and the point O and the optical axis L) is αi. , 'because it is = 1 + tan (αi) · tan (θ), the imaging magnification mi at any point' D _a / Di calculated from the inclination angle theta, photographing magnification m _a and shooting angle .alpha.i the following equation (1) Is done.

[0036]

(Equation 1)

[0037] In the equation (1), the photographing magnification m _A is m _A = a
· F / D _A (a; proportionality factor, f; focal length) is calculated from. Also, the expression of D _A / Di ′ = 1 + tan (αi) · tan (θ) can be expressed by the following equation 2 when the object distance D _B and the shooting angle of view α _B with respect to the point B in FIG. Desired.

[0038]

(Equation 2)

Further, when the object distance at any point between the AC in the imaging plane of the CCD 20 Di, the shooting angle of the point and _{βi, D A / Di = 1} -tan (βi) · tan (θ)
Since it is, the imaging magnification mi at any point between the AC inclination angle theta, following the imaging magnification m _A and shooting angle .beta.i (2)
It is calculated by the formula.

[0040]

(Equation 3)

D _A / Di = 1−tan (βi) · tan (θ)
Can be obtained by the same method as described above.

The illuminance unevenness correction will be briefly described. The image of the text information is mainly used for recording information, and it is required that the image quality is higher than the descriptiveness. It is desirable to clarify the information and reduce unevenness of the illuminance to make it easier to see as a whole.

In the present embodiment, as shown in FIG. 9, the captured image G is divided vertically and horizontally into a plurality of square blocks B (I) (I = 1, 2,..., 18 in FIG. With the white background level W calculated using the histogram of the level distribution of the pixel data included in B (I), the γ characteristic as shown in FIG. 10 is changed to the γ characteristic γ with respect to the center position O (I) of the block (I). (I), and further using the set γ characteristic γ (I), the pixel position P other than the center position O (I) of each block B (I)
Interpolates the γ characteristics γ (P) of these, and these γ characteristics γ (I), γ (P)
Is used to perform gamma correction of the pixel data at the corresponding pixel positions O (I) and P, thereby performing clarification of character information and correction of uneven illuminance.

FIG. 10 shows a case where the pixel data is A / D converted into 8-bit data, and the level value “255” of the input / output level indicates the maximum value. Further, in the γ characteristic shown in the figure, since all the pixel data at the input level W or higher are converted to pixel data saturated to the maximum level, the captured image has a uniform white background portion including the pixel data at the input level W or higher. The image quality is corrected to a lightness white image, whereby the contrast of the character information portion with respect to the white background portion is enhanced, and the character information is clarified.

If necessary, gamma correction is performed on the image after the illuminance non-uniformity correction using the γ characteristic as shown in FIG. That is, the image after the illuminance non-uniformity correction is corrected to the image quality in which the black background portion including the pixel data of the input level B or less is uniformly black with the minimum brightness. As a result, blackening of the character portion is performed in accordance with the density, the line thickness, and the line density of the characters and graphics written on the whiteboard 22, thereby optimizing the clarification of the character information. . The details of the gamma correction process in the illuminance unevenness correction will be described later.

Returning to FIG. 1, the camera 1 is provided with a photographing lens 2 composed of a zoom lens substantially at the center of the front surface, and a light projecting window 4 and a light receiving window for measuring a subject distance by an active distance measuring method above the photographing lens 2. 5 is provided, and a photometric window 3 for measuring the brightness of the subject is provided between the two windows. A finder objective window 6 is provided on the left side of the light projecting window 4, and a flash 7 is provided on the right side of the light receiving window 5.

The light projecting window 4 is a window for irradiating the subject with infrared light, and the light receiving window 5 is a window for receiving the reflected light of the infrared light from the subject. In the present embodiment, the active distance measuring method is adopted as the distance measuring method, but a passive distance measuring method may be used.

On the side of the camera 1, there is provided a card insertion slot 8 into which the hard disk card 13 is inserted and removed. Above the card insertion slot 8, a card ejection button 9 for ejecting the hard disk card 13 is provided. . When printing out the photographing result, the user can press the card ejection button 9 to remove the hard disk card 13 from the camera 1 and mount the hard disk card 13 in a printer to which the hard disk card can be mounted to print out.

The camera 1 is provided with a SCSI cable interface, and the camera 1 and the printer are
The image data may be directly transferred from the camera 1 to the printer by connecting with a cable, and the photographed image may be printed out.

In this embodiment, a hard disk card conforming to PCMCIA is used as a recording medium for image data. However, a memory card or a mini disk (MD) may be used as long as it can store photographed results as image data.
And other recording media.

On the upper surface of the camera 1, a shutter button 10 is provided at the left end, and a zoom switch 11 and a photographing / playback switch 12 are provided at the right end. The shutter button 10 is an operation button that turns on an S1 switch for instructing shooting preparation such as focal length adjustment and exposure control value setting when pressed halfway, and turns on an S2 switch for instructing release when fully pressed. The zoom switch 11 can slide left and right
It consists of a contact switch, and the zoom switch 11 is set to T (TEL
Slide to the E) side to the telephoto side and W (WID
By sliding to the E) side, the zoom ratio of the taking lens 2 can be continuously changed to the wide angle side.

The photographing / reproducing switch 12 is a switch for switching between a photographing mode and a reproducing mode. The photographing / playback switch 12 is composed of a two-contact changeover switch slidable left and right. When the photographing / playback switch 12 is set to the photographing (REC) side, photographing of a subject (recording of a photographed image on the hard disk card 13) And playback (PL
AY), the hard disk card 13
LCD display section 19 of the captured image recorded in the camera (see FIG. 2)
Monitor display on the computer.

On the back of the camera 1, as shown in FIG.
A main switch 14 for turning on the power and a finder eyepiece window 15 are provided at the upper left end and substantially at the center, respectively, and a mode setting switch 16 is provided on the right side of the finder eyepiece window 15.

The mode setting switch 16 has a function of switching between a normal photographing mode and an oblique image correction mode for correcting an oblique image into a front image, and has a function of setting an inclination angle θ (see FIG. 8).

The mode setting switch 16 comprises a horizontally long guide groove 16b provided with an angle scale 16a on the upper side and an operation button 16c movable along the guide groove 16b. The inclination angle θ can be set by setting the angle position to.

Note that the angle scale 16a is 15 °, 30 °, and 30 ° on the left and right sides of the central front position (angle 0 °).
An angle of 45 ° is provided, and three types of inclination angles θ can be set for each of the left and right sides. Here, the angle on the left side is an inclination angle when photographing from the left side toward the subject (hereinafter, this oblique photography is referred to as oblique photographing on the left side), and the angle on the right side is when photographing from the right side toward the subject (hereinafter, oblique photographing). This oblique photography is referred to as right oblique photography.). When the operation button 16c is set to the front position, since the inclination angle is 0 °, the oblique image correction is not performed on the photographed image, and the normal photographing mode is set.

Further, in this embodiment, the photographer can discretely set the inclination angle θ measured by the amount of eyes, but the inclination angle θ is continuously changed according to the slide amount of the operation button 16c. May be set.

Further, an illuminance unevenness correction switch 17 and a black density adjustment switch 18 are provided below the main switch 14. Further, an LCD display unit 19 is provided below the finder eyepiece window 15.

The illumination unevenness correction switch 17 is a switch for instructing the above-described illumination unevenness correction. Illumination unevenness correction switch 17 is ON / OF in which the operation button slides left and right
It is composed of an F switch. Illuminance unevenness correction switch 1
When the illuminance non-uniformity correction is instructed by 7, the captured image is divided into a plurality of blocks (small images), and a γ characteristic as shown in FIG. The gamma correction is performed using this. On the other hand, when the illuminance non-uniformity correction switch 17 is turned off, a γ characteristic suitable for normal photographing that has been set in advance (image quality with high descriptiveness capable of reproducing the gradation of the subject as faithfully as possible) is obtained. Gamma correction processing of the captured image is performed using the obtained γ characteristic).

The illuminance non-uniformity correction is intended to improve image quality deterioration due to illuminance non-uniformity when photographing character information written on a whiteboard, and is mainly applied when photographing such character information. Therefore, if an image obtained by capturing information such as characters and figures is referred to as a “character image” and an image obtained by capturing a landscape or a person is referred to as a “natural image” and the content of the captured image is divided into two types, the illuminance unevenness correction switch 17 is used for The switch switches image processing (especially gamma correction processing) of an image between a character image mode and a natural image mode.

When photographing a character image, the photographer sets the illuminance non-uniformity correction switch 17 to "ON" so that an image quality suitable for the character image (white information is appropriately emphasized by skipping a white portion white). Illuminance unevenness correction switch 1 when shooting a natural image.
By setting “7” to “OFF”, a captured image having an image quality suitable for a natural image (image quality with high descriptiveness) can be obtained.

The black density adjustment switch 18 is a switch for adjusting the black saturation level B (see FIG. 11) of the γ characteristic in the gamma correction of black emphasis on the image after the illuminance unevenness correction. The black density adjusting switch 18 is constituted by a three-contact switch whose operation button slides left and right. The black density adjustment switch 18 functions only when the illuminance unevenness correction switch 17 is set to “ON” (when the character image mode is set). In the character image mode, when the black density adjustment switch 18 is set to the OFF state, as shown in FIG. 12, the black saturation level of the γ characteristic is set to a predetermined level B0, and the black density adjustment switch When 18 is set to “dark”, the black saturation level of the γ characteristic is switched to a predetermined level B1 (> B0) larger than the predetermined level B0, and when set to “light”, black saturation of the γ characteristic is set. The level is switched and set to a predetermined level B2 (<B0) smaller than the predetermined level B0.

In the image processing in the character image mode, the white saturation level of the γ characteristic is automatically adjusted so as to remove the background portion white, and the density of the character portion is further changed by changing the black saturation level by the black density adjustment switch 18. Can be changed to adjust the contrast with the base (white background).

For example, comparing the characters written on the whiteboard with the characters written on the document, the characters on the whiteboard are generally thicker and larger than the characters on the document. When the same gamma correction is performed as in the case, the contrast of the character with respect to the background is reduced as compared with the case of the whiteboard. Therefore, when photographing an original, the black density adjustment switch 1
By setting “8” to “dark”, the black of the character portion can be enhanced, and the contrast of the character portion with respect to the background can be suitably adjusted.

In the present embodiment, the black saturation level is set to be switched in two stages. However, a multi-stage switching system may be used or the black saturation level may be switched continuously.

The LCD display section 19 displays a captured image on a monitor. When the photographing / reproduction switch 12 is set to the reproduction side, a captured image of a frame designated by a selection switch (not shown) is read from the hard disk card 13 and reproduced and displayed on the LCD display unit 19.

FIG. 13 is a block diagram of the camera 1 according to the present invention. In the figure, the same members as those described above are denoted by the same reference numerals. Also, the CCD driving unit 3
1 controls the imaging operation of the CCD 20 based on the shutter speed of the exposure control value input from the CPU 30. The CCD 20 includes a color area sensor, performs an imaging operation (charge accumulation operation) based on a control signal input from the CCD driving unit 31, converts pixel signals of each of R, G, and B color components into a time-series signal, and outputs an image. Processing unit 32
Output to

The image processing section 32 performs predetermined signal processing on the pixel signal output from the CCD 20 and outputs the processed signal to the hard disk card 13. The image processing unit 32 has A /
D converter 321, oblique image correction operation unit 322, data effective area operation unit 323, image memory 324, first γ characteristic setting unit 325, second γ characteristic setting unit 326, first γ correction unit 3
27, a second γ correction unit 328 and a switch circuit 329, and also performs diagonal image correction on diagonal images. When illuminance unevenness correction is instructed, a γ characteristic for illuminance unevenness correction is set for each block, and gamma correction is performed using the γ characteristic. At this time, the γ characteristics for the portion between the center positions of the blocks are interpolated, and the image signal of this portion is gamma-corrected using the interpolated γ characteristics, whereby discontinuity in image quality based on the difference in γ characteristics between the blocks is reduced. Be relaxed.

The A / D converter 321 converts each pixel signal contained in the image signal read from the CCD 20 into a digital signal (hereinafter, referred to as pixel data).

The oblique image correction operation section 322 performs an oblique image correction process when the oblique image correction mode is set. The oblique image correction calculation unit 322 enlarges and reduces the captured oblique image in the horizontal direction (FIG. 3, x-axis direction) and the vertical direction (FIG. 3, y-axis direction) based on the above-described oblique image correction principle. Is performed, and a pseudo front image is corrected.

FIG. 16 is a diagram showing an image processing method for correcting a photographed oblique image into a pseudo front image.
(A) is a figure which shows the image before correction, (b) is a figure which shows the image after correction.

In the figure, the xy coordinates are the orthogonal coordinates temporarily set on the image G (image composed of pixel data) captured by the CCD 20, and the origin is set at the center of the image G. An image K (corresponding to the whiteboard 22) that is photographed obliquely and included in the image G is an image to be subjected to oblique image correction (hereinafter, referred to as a correction target image).

In the present embodiment, since the focus adjustment position of the camera 1 is set at the center of the photographing screen (the origin O of the xy coordinates), the focus is accurately adjusted on the correction target image K in FIG. The image K1 on the y-axis corresponds to the portion on the y-axis. The image K1 on the left side of the y-axis is an image inclined so as to approach the camera 1 side, and the image K2 on the right side of the y-axis is separated from the camera side. It is an image inclined to.

Therefore, as shown in FIG.
Of the image G1 on the left side of the y-axis, the x-axis and y
By compressing the image in both directions of the axis and performing scaling processing so as to obtain an image G1 ', the trapezoidal oblique image K1 is corrected to a rectangular reduced image K1', and the image G2 on the right side of the y axis is x The trapezoidal oblique image K2 is corrected to a rectangular enlarged image K2 'by expanding the image in both directions of the axis and the y-axis and performing scaling processing so as to obtain an image G2'.

The processing method for enlarging and reducing a photographed image in oblique image correction is described in, for example, Japanese Patent Application Laid-Open No.
This is done in the manner described in No. 22. That is, since the object plane is not inclined with respect to the imaging surface of the CCD 20 in the y-axis direction, the captured image is scaled by k (>
In the process of correcting the image expanded in 1), each pixel data forming the captured image (image before correction) is converted into each pixel data forming the captured image obtained by capturing an image obtained by enlarging the original image at a magnification k. This is done by substitution.

The captured image in the y-axis direction is scaled by k (<
The process of correcting the image compressed in step 1) is also performed by replacing each pixel data forming the captured image with each pixel data forming the captured image obtained by capturing an image obtained by reducing the original image by the magnification k. .

The image expansion / compression scaling process in the x-axis direction can be performed in the same way as the image expansion / compression scaling process in the y-axis direction.
In the axial direction, the photographing magnifications mi and mi 'change, and the magnification k differs for each pixel. Therefore, the scaling processing of image expansion and compression is performed using the magnification ki for each pixel.

[0078] When the enlargement processing in FIG. 8, any of magnification ki at the pixel position reciprocal of the ratio of the photographing magnification mi at the pixel position relative to the photographing magnification m _A at point A between the AC and (m _A / mi) obliquely The product of a correction coefficient (1 / cos (θ)) for correcting an image into a front image, that is, ki =
(M _A / mi) / expressed as cos (theta). On the other hand, from the above equation (2), because it is _{mi / m A = 1-tan} (βi) · tan (θ) = cos (βi + θ) / (cos (βi) · cos (θ)), magnification ki is It is calculated by the following equations (3) and (4).

[0079]

(Equation 4)

In the case of compression processing, in FIG.
Correction coefficient reduction ratio ki at any pixel position 'is the photographing magnification mi at the pixel position relative to the photographing magnification m _A at point A' to correct the ratio of the reciprocal (m _A / mi ') and slant image of the front image between A product of (1 / cos (θ)) , i.e., expressed as _{ki '= (m a / mi} ') / cos (θ). On the other hand, from the above equation _{(1), mi '/ m A = 1} + tan (αi) · tan (θ) = cos (αi-θ) / from a (cos (.beta.i) · cos (theta)), reduction ratio ki 'Is calculated by the following equations (5) and (6).

[0081]

(Equation 5)

In this embodiment, the oblique photographed image is corrected by the scaling process as described above.
For example, the oblique photographed image is corrected by a so-called image padding method or a so-called pixel thinning method in which an image is enlarged by interpolating known pixel data into pixel data to be increased, or compressing an image by thinning out pixel data. You may do so. Further, in the present embodiment, the oblique photographed image is corrected by performing both the enlargement and compression processes. However, the oblique photographed image is corrected by only one of the enlargement and compression processes. You may.

When the scaling process is performed to compress the image G1, a portion where the pixel data is missing (see a region ΔG1 shown by oblique lines in FIG. 16B) occurs in the corrected image. When outputting to a monitor TV or printer,
The image becomes unnatural. In the example of FIG. 16B, the area ΔG
Since the pixel data of the noise level is output to the portion 1, the image quality of this portion becomes unstable at the time of reproduction, and the overall image becomes poor in appearance. For this reason,
In the correction process, for example, white dummy data may be set as pixel data in the area ΔG1 so that the missing portion becomes white.

The region ΔG1 has a wedge-shaped top and bottom, giving an unnatural feeling.
As shown in FIG. 7, a band-like blank area ΔG0 that can include the above-mentioned area ΔG1 is provided on the left side, upper side, and lower side (peripheral part in some cases) of the corrected image G ′. It is preferable to set dummy data of a specific color such as. By doing so, the reproduced image appears to be fringed, so that unnaturalness can be reduced.

When the normal photographing mode is set, the oblique image correction operation section 322 does not perform the correction processing, and the pixel data input from the A / D converter 321 is output to the image memory 324 as it is. You.

The data valid area calculation unit 323 calculates the area other than the dummy data (the area except for ΔG1 in FIG. 16 and the area except for ΔG0 in FIG. 17) in the image after the oblique image correction.
It is called the data valid area. ) Is calculated. When correcting the skew image to the front image and then setting the γ characteristic for illuminance unevenness correction for each block, in the block including the dummy data, an accurate γ characteristic cannot be obtained due to the dummy data. In such a block, as described later, in such a block, the γ characteristic is set using only valid pixel data excluding dummy data. The information of the data valid area calculated by the data valid area calculation 323 is used when extracting valid pixel data in each block.

The image memory 324 stores the oblique image correction operation unit 3
The pixel data output from the memory 22 is stored.
The image memory 324 has a capacity capable of storing pixel data of one captured image, so that image processing of the captured image can be performed collectively.

When the image is divided into blocks, the capacity of the image memory 324 of the memory is divided into a plurality of blocks at the maximum block size that can be set so that image processing can be performed in units of rows of the blocks. In this case, the capacity may be set to be able to store the pixel data included in the blocks arranged in at least one row, and the memory may be reduced. That is, for example, as shown in FIG.
Is the maximum block size of 3 × 3 blocks B (1) to B
If the image memory 324 is divided into (9), the capacity of the image memory 324 is changed to blocks B (1) to B (3),
(4) to B (6) and a capacity capable of storing the pixel data included in the blocks B (7) to B (9).

The first γ characteristic setting section 325 sets a γ characteristic for correcting illuminance unevenness of a captured image. The first γ characteristic setting unit 325 divides the captured image into a plurality of blocks, and sets γ characteristics for illuminance unevenness correction using pixel data included in each block for each block. The second γ characteristic setting unit 326 sets the γ characteristic for enhancing the black density of the image after the illuminance unevenness correction. The second γ characteristic setting unit 326 sets C
The gamma characteristic for black density enhancement is set based on the operation information of the black density adjustment switch 18 input from the PU 30.

The first γ correction section 327 is a circuit for performing gamma correction on a natural image, and the second γ correction section 328
Is a circuit for performing gamma correction on a character image. First
The γ correction unit 327 performs gamma correction on the pixel data read from the image memory 324 using a preset γ characteristic suitable for a natural image. The second γ correction unit 328 divides the captured image of the character image into a plurality of blocks, and sets γ for illuminance unevenness correction set by the first γ characteristic setting unit 325 for each block.
After performing gamma correction of the pixel data forming the captured image using the characteristics, gamma correction of the pixel data is performed using the γ characteristic for black density enhancement set by the second γ characteristic setting unit 326. Note that gamma correction for a character image will be described later.

The switch circuit 329 is connected to the image memory 3
24 and the connection between the first and second γ correction units 327 and 328. The switching control of the switch circuit 329 is performed by the CPU in accordance with the setting state of the illumination unevenness correction switch 17.
When the illuminance unevenness correction switch 17 is set to “OFF” (the natural image mode is set), the image memory 324 and the first γ correction unit 327 are connected. When the illuminance unevenness correction switch 17 is set to “ON” (character image mode is set), the image memory 324 and the second γ correction unit 32 are set.
8 is connected.

Here, a method of gamma correction (illuminance unevenness correction and black enhancement correction) for a character image will be described.
As described above, in the case of a character image, it is desirable to increase the clarity of the character portion relative to the white background portion. Therefore, in order to skip the white background portion white, as shown in FIG. The γ characteristic in which the output level is saturated at the level W is used.

The white saturation level W in the γ characteristic is
For example, a histogram of the level distribution of the pixel data of the green component forming the image of the character image is created, and the class having the maximum frequency within the range corresponding to the white background portion is set. That is, when a histogram of the level distribution of the green component pixel data is created for an image obtained by photographing the whiteboard 22 on which characters, figures, and the like are drawn, as shown in FIG. 19, the histogram generally corresponds to a white background portion (board portion). A level w corresponding to a peak of the mountain U corresponding to a mountain U and a mountain C corresponding to a black character portion (character portion) is set as a white saturation level W of the γ characteristic.

The white saturation level W is determined from the histogram of the green component pixel data constituting the entire captured image.
The gamma correction may be performed on the entire captured image by using the gamma characteristic.
The handwritten in No. 2 has a large variation in the character density (ratio of the character portion to the white portion), and in the case of photographing, the light source is not constant, unlike the case of a copying machine equipped with an illuminating device. Since the illuminance distribution greatly changes within the illuminance, the illuminance unevenness in the imaging screen becomes large. Therefore, it is preferable to divide the captured image into a plurality of blocks and perform gamma correction in block units using the γ characteristic set for each block. It is desirable to correct the illuminance unevenness by doing so.

In the camera 1 according to the present embodiment, FIG.
As shown in the figure, the captured image G is vertically and horizontally n (= K (vertical) × L
(Horizontal)) blocks B (I, J) (I = 1, 2,..., K, J
= 1, 2,... L), and for each block B (I, J), a γ characteristic for illuminance unevenness correction representing the block B (I, J) is set. In this case, block B (I,
The size (area) of J) is set to a size that can accommodate approximately 9 (= 3 × 3) characters. Thus, block B (I, J)
Is relatively set in relation to the number of characters because the peak U corresponding to the whiteboard 22 has a moderately steep peak when a histogram is created using the pixel data in the block. , The peak position w of the mountain U
Is to be surely detected.

That is, as shown in FIG. 21A, when the block size is set to be relatively small with respect to the character, the area occupied by the character portion in the block B (I, J) is large, Since the peak U of the portion corresponding to the whiteboard 22 is lowered, the peak position w of the peak U may be erroneously detected. As shown in FIG. When the size is set to a large size, the illuminance unevenness in the block B (I, J) is large, and the peak U of the portion corresponding to the whiteboard 22 of the histogram becomes gentle. This is because it may be detected.

In order to determine the block size, it is necessary to know the size of the character projected on the shooting screen. The size y 'of the character projected on the shooting screen is determined by the size of the character written on the whiteboard 22. Y ′ = y ·
m, and the size y of the character written on the whiteboard 22 is considered to be within a certain range even though there are individual differences.
The representative value y0 of the character size y written in 2 and the photographing magnification m
If 0 is experimentally determined, the size y ′ of the character projected on the photographing screen can be uniquely determined from the photographing magnification m.

In the present embodiment, the reference block size S0 is determined based on the size y0 of the character projected on the photographing screen at a certain photographing magnification m0, and the block size S0 at an arbitrary photographing magnification m is determined. Is calculated from the photographing magnification m0 and the block size S0 by an arithmetic expression of S = S0 · m / m0. Therefore, if it is assumed that a block having the block size S0 includes vertical and horizontal (i × j) pixel data, the number i ′ of the pixel data included in the block size S in the vertical direction is i · m / m0, Number in the horizontal direction
j ′ is j · m / m0.

In the present embodiment, the block size S is set so that the number of characters in the block is nine. However, this is merely an example, and the size y of the character written on the whiteboard 22 is set. Is changed, the number of characters in the block also changes. Therefore, the block size S is appropriately set to include an appropriate number of characters according to the setting of the representative value y0.

In the present embodiment, the block size S is changed in accordance with the photographing magnification m. However, the block size S0 is fixed, and in photographing a character image, the photographing magnification m is changed to the block size. Predetermined value m for S0
The adjustment may be made to be zero. That is,
As shown in FIG. 22, a block frame 43 corresponding to the block size S0 is displayed in the viewfinder frame 42,
The photographer may adjust the zoom ratio of the photographing lens 2 or the subject distance so that nine characters written on the whiteboard 22 in the block frame 43 are included. The block frame 43 may be displayed at all times, or may be displayed only when the character image mode is set.

Next, a method of determining a γ characteristic for correcting illuminance unevenness from a histogram of pixel data of a green component will be described.

(I × j) included in block B (I, J)
Pixel data g (1,1), g (1,2),... G (i-1, j), g (i,
In j), pixel data of X (%) set in advance is removed by integrating from the maximum level to the low level side, and a histogram of a level distribution is created using the remaining pixel data.
For example, if the total number of pixel data contained in a block is 1000
Assuming that there are no pixel data and X = 3%, 300 pixel data obtained by integrating sequentially from the pixel data at the maximum level q to the low level side is removed, and a histogram is formed using the remaining 9700 pixel data. Created. X% on high level
The reason for removing the pixel data is to avoid adverse effects such as noise.

This histogram generally has a two-peak distribution as shown in FIG. 23, and the peak U on the high level corresponds to the background portion of the whiteboard 22, and the peak C on the low level.
Is equivalent to the character part. The class p in FIG.
Is the maximum level of the pixel data included in the block B (I, J), and the class q (<p) is the maximum value of the class of the histogram.

When a histogram is created, the maximum class p
, The class w having the highest frequency in the distribution included in the range d set in advance to the low level is calculated, and this class w is set as the white saturation level W of the γ characteristic for illuminance unevenness correction. Since the block size is set to a predetermined size in relation to the number of characters in the range d, it is estimated that only the high-level mountain U is surely included in a block photographed with normal illuminance. Range. For example, when the pixel data is 8 bits and has a gradation of 0 to 255, the range d is set to about 48.

Therefore, for example, if the maximum class q is 200, the class w having the maximum frequency within the class range 152 to 200 is calculated. If the class w is, for example, w = 180, the white saturation level W = 180, and the γ characteristic as shown in FIG. 24 is determined.

In the oblique image correction mode, a region where dummy data exists is generated in a portion subjected to the compression processing. If the white saturation level W is set using the histogram of all the pixel data including the dummy data in the block including the dummy data, the correct white saturation level W will not be set. A histogram is created using only valid pixel data excluding data, and the white saturation level W is set using this histogram.

If the whiteboard 22 is not completely white and slightly colored, or if the white balance adjustment of the camera 1 is improper, the γ for correcting uneven illuminance is used.
Since the equivalent γ value of gamma correction performed using the characteristic and the γ characteristic for black enhancement is relatively large, the γ characteristic for illuminance unevenness correction set using the green component pixel data is used. It cannot be applied to gamma correction of pixel data of red component and pixel data of blue component.

That is, the photographing data in a certain area of the whiteboard 22 is not completely white, and the levels D _R , D _G , and D _B of the pixel data of each of the R, G, and B color components are, for example, (D _R , D _G , D _B ) = (130, 140, 125), and the equivalent γ characteristic of the γ characteristic for illuminance unevenness correction and the γ characteristic for black enhancement set using the pixel data of the green component ( The γ characteristic obtained by superimposing both γ characteristics) is shown in FIG.
When the gamma correction is performed on the pixel data of the red component and the pixel data of the blue component using the γ characteristic, the output of each color component becomes (D) as shown in FIG.
_R , D _G , D _B ) = (185, 255, 140)
The image after the gamma correction causes a large color shift in yellow-green.

When the γ value of the γ characteristic shown in FIG. 25 is small (when the inclination is gentle), the output difference between the respective color components after the gamma correction is small, so that the color shift hardly causes any problem.
Since the γ characteristic applied to the character image mode is for performing gamma correction close to the binarization process, the γ value is set relatively large, so that the γ characteristic set using the green component pixel data is used. It is difficult to apply gamma correction to pixel data of red component and pixel data of blue component.

As a method of avoiding the above-described coloring phenomenon of the white portion, the pixel data of each of the R, G, and B color components is converted into luminance data and color difference data, and gamma correction is performed using only the luminance data. Again, a method of reverse conversion to pixel data of R, G, B color components is conceivable. However, in this method, since color difference data is stored, for example, characters written on a white board are thinned with ink. If the character is a color character, the character remains faint even after gamma correction, and it is difficult to clearly reproduce the faint character.

In the present embodiment, the gamma characteristic for illuminance unevenness correction set by the pixel data of the green component is corrected to set a dedicated gamma characteristic for each color component, and a dedicated gamma characteristic is set for each color component.
By performing gamma correction on the characteristics, even light-colored characters can be clearly reproduced.

The gamma characteristic for each color component is, for example, a level margin value of "5" and an input level (D _R -5,
D _G -5, D _B -5) are formed so that white saturation level is set using the pixel data of each color component. For example, FIG.
In the example of the γ characteristic shown in FIG. 26, as shown in FIGS. 26A to 26C, the input level (125, 1) of each of the R, G, and B color components
35, 120) becomes the white saturation level 255,
The γ characteristics of each of the R, G, and B color components are set.

Note that the gamma correction is performed so that the colored white portion becomes white, so the colored portion is shifted from the original color. However, in a character image, information is more important than color reproducibility. Some color misregistration is considered acceptable.

In the above description, a histogram of the level distribution of the pixel data included in the block is created, and the white saturation level W is determined using this histogram (ie, the γ characteristic is set). However, the γ characteristic may be set by calculating pixel data instead of the histogram.

Now, each block B (I,
J) A gamma characteristic for illuminance non-uniformity correction is set for each block, and the gamma correction of the image is performed in block units using the γ characteristic. The image quality changes suddenly, which may cause a boundary line (false line). That is, the level of the white background suddenly changes at the boundary of the block, and this discontinuity in the level of the white background may occur as a boundary line.

Therefore, in this embodiment, each block B
The gamma characteristic for illuminance unevenness correction set for each (I, J) is defined as the γ characteristic for the pixel data at the center position of the block B (I, J), and the illuminance unevenness for the pixel data between the center positions of the adjacent blocks. The γ characteristic for correction is linearly interpolated using the γ characteristic for illuminance unevenness correction of both blocks, and the linearly interpolated γ characteristic
By performing gamma correction on pixel data other than the center position in characteristics, discontinuity in image quality based on differences in γ characteristics between blocks is reduced.

That is, as shown in FIG. 27, each of the blocks B (I, J), B (I, J + 1), B (I + 1, J), B (I + 1, J + 1) Assuming that the center positions are A, B, C, and D, γ for illuminance unevenness correction with respect to an arbitrary position P in an area AR1 surrounded by ABCD.
The characteristics are represented by blocks B (I, J), B (I, J + 1), B (I + 1, J), B
Linear interpolation is performed using the γ characteristic for illuminance unevenness set for each (I + 1, J + 1), and gamma correction of the pixel data at the position P is performed using the interpolated γ characteristic.

The interpolated γ characteristics for illuminance non-uniformity with respect to the position P are represented by blocks B (I, J), B (I, J + 1), B (I + 1,
J) and B (I + 1, J + 1), the calculated white saturation levels W _A , W _B , W _C , and W _D are treated as for positions A, B, C, and D. White saturation level W
_{A, W B, W C,} W D from the position A, B, C, is set by calculating the value W _P which internally divides the position P by the following equation (7) with respect to D.

[0119]

(Equation 6)

In the internal division method, blocks B (1,1) to B (1, L), B (2, L) to B (K, L),
In B (K, L-1) to B (K, 1) and B (K-1,1) to B (2,1), the gamma characteristic of the portion outside the center position of each block is not interpolated. For this portion, linear interpolation of the γ characteristic may be performed by the external division method.

The γ characteristic may be interpolated at all positions except the center position of each block B (I, J). However, a portion other than the center position of each block B (I, J) may be interpolated by a plurality of pixels. The block may be divided into blocks containing data (for example, 4 × 4 pixels to 6 × 6 pixels), and the γ characteristic may be linearly interpolated on a block-by-block basis to reduce the time for the interpolation operation.

In the above-described interpolation processing of the γ characteristic for correcting the uneven illuminance, since the γ characteristic is set for each pixel position, a block centering on each pixel position is set, and the pixels included in the block are set. A similar result can be obtained by setting the γ characteristic using the histogram of the data level distribution. However, this method requires a long time to calculate the γ characteristic because a large number of blocks are set in the captured image G. There is a disadvantage that requires. Further, since almost all of the pixel data overlaps between adjacent blocks, there is almost no difference in the created histogram, and there is no benefit in creating histograms for both blocks. It employs linear interpolation processing of γ characteristics that can perform calculations and can reduce the memory capacity.

FIG. 14 shows an oblique image correction calculation section 322 to first and second γ correction sections 3 for performing image processing of a color image.
FIG. 27 is a block diagram showing a configuration up to 27 and 328.

The oblique image correction calculation section 322, the image memory 3
24, first γ characteristic setting section 325, first and second γ correction sections 3
27, 328 and the switch circuit 329 are R,
There are three processing circuits having the same structure for the pixel data of each of the G and B color components.

For example, the pixel signal of the R color component is temporarily recorded in the image memory 324A after the image distortion is corrected by the oblique image correction calculation section 322A. In the natural image mode, the pixel data of the R color component stored in the image memory 322A is supplied to the first γ correction unit 32 via the switch circuit 329.
7A, and gamma correction is performed using a predetermined natural image gamma characteristic.

On the other hand, in the character image mode, the first γ characteristic setting section 325A sets the γ characteristic for illuminance unevenness correction from the histogram of the level distribution of the R pixel data included in each block for each block, and the second γ characteristic Setting part 3
26 sets the γ characteristic for black adjustment based on the adjustment value of the black density adjustment switch 17. Then, the pixel data of the R color component stored in the image memory 322A is read out to the second γ correction unit 328A via the switch circuit 329, and each block has a γ characteristic for illuminance unevenness correction and performs gamma correction. After that, gamma correction is performed using the gamma characteristic for adjusting black density.

The pixel signal of each of the G and B color components is processed in the same manner as the above-described pixel signal of the R color component.

FIG. 15 shows the first γ characteristic setting section 3.
FIG. 25 is a block diagram showing the internal configuration of a 25.

In the figure, a block size setting unit 32
5a sets a block size for dividing a captured image into small image blocks B (I, J). The block size setting unit 325a sets the block size S using the imaging magnification m input from the CPU 30 and the preset reference size S0 and reference imaging magnification m0.

The address generator 325b generates an address of pixel data included in each block B (I, J) based on the block size S0 set by the block size setting unit 325a. This address data is used for reading control of pixel data from the image memory 324 and for interpolation calculation in the white saturation level interpolation calculation unit 325e.

The histogram creating section 325c creates a histogram (see FIG. 23) of the level distribution of the pixel data contained in each block B (I, J). The white saturation level setting unit 234d sets the white saturation level W (see FIG. 24) of the γ characteristic with respect to the center position of each block B (I, j) using the histogram created by the histogram creation unit 325c. . The white saturation level interpolation calculation unit 325e uses the white saturation level W of the γ characteristic set for each block B (I, j) to calculate the white of the γ characteristic for a portion other than the center position of each block B (I, j). The saturation level W is set by interpolation.

The gamma characteristic setting section 325f includes a white saturation level setting section 325d and a white saturation level interpolation calculation section 325e.
The gamma characteristic for illuminance non-uniformity correction for each pixel data of the captured image is set using the white saturation level W set by the above.

Returning to FIG. 13, the card drive unit 33 controls the drive of the hard disk card 13 to record image data. The light emission control unit 34 controls light emission of the flash 7.

The LCD drive section 35 controls the monitor display of the captured image on the LCD display section 19 based on the control signal from the CPU 30. The memory 36 is a CPU
30 (data of the subject distance Di at each pixel position of the CCD 20)
And the photographing magnification mi).

The lens driving section 37 controls the focusing operation of the taking lens 2 based on the AF control value input from the CPU 30. Further, the zoom drive unit 38 is provided with
The zoom operation of the photographing lens 2 is controlled based on the drive signal input from the camera. The aperture driving unit 39 is a CPU 3
The aperture of the aperture 21 is controlled based on the aperture value Av of the exposure control value input from 0.

The photometric section 40 is composed of a light receiving element such as an SPC provided at a position behind the photometric window 3 and measures the luminance of the subject. The distance measuring unit 41 detects the subject distance, and is provided at a position behind the light projecting window 4, and is provided at a position behind the light projecting unit 411 that emits infrared light and the light receiving window 5, and reflected by the subject. A light receiving section 412 for receiving infrared light.

The CPU 30 is for centrally controlling the photographing operation of the camera. CPU30 has an imaging magnification calculator 301, in the oblique image correction mode, shooting at the object distance D _A and its distance measuring point in the detected distance measuring point in the distance measurement unit 41 (the center position A of the imaging surface of the CCD 20) magnification m _A
And the photographing magnifications mi and mi 'at the other pixel positions are calculated. Further, the CPU 30 has an exposure control value calculation unit 302, calculates an exposure control value (aperture value Av, shutter speed Tv) based on the luminance information of the subject detected by the photometry unit 40, and uses the calculated results as aperture drive values. Output to the section 39 and the CCD drive section 31. Also,
The CPU 30 has an AF control value calculation unit 303 and the distance measurement unit 4
_A lens drive amount for setting the taking lens 2 to the in-focus position is calculated based on the subject distance DA detected in step 1, and the calculation result is output to the lens drive unit 37 as an AF control value.

Next, the photographing control of the camera 1 will be described.
This will be described with reference to the flowcharts in FIGS. It is assumed that the shooting / playback switch 12 is set to the shooting side.

When the main switch 14 is turned on, the camera 1
Is activated, the camera is ready for shooting. When the zoom switch 11 is operated in this state (YES in # 2), the zoom lens in the photographing lens 2 is driven according to the operation direction and the operation amount, and the zoom ratio is changed (# 4). Thereafter, when the shutter button 10 is half-pressed and the S1 switch is turned on (YES in # 6), the process shifts to step # 8, and processing for photographing preparation is performed.

That is, first, the mode setting switch 16
It is determined from the setting position of the operation button 16c whether or not the mode is the oblique image shooting mode (# 8). If the mode is the oblique image shooting mode (YES in # 8), the inclination angle θ is determined from the setting position of the operation button 16c. The image is captured (# 10), and a correction calculation process is instructed to the oblique image shooting correction calculation unit 322 (# 1).
2). On the other hand, if the mode is not the oblique image shooting mode (N
O), the above steps # 10 and # 12 are skipped.

Subsequently, it is determined whether illumination unevenness correction is instructed (illuminance unevenness correction switch 17 is set to "ON") (# 14), and if illumination unevenness correction is instructed (# 14). If YES in # 14), the density information on the character portion is fetched from the set position of the black density adjustment switch 18 (# 16). Subsequently, the first γ characteristic setting unit 325 is set to be processable, and the density information is set to the second γ characteristic.
The γ characteristic for black enhancement is input to the characteristic setting unit 326, based on the density information, and the γ characteristic for black enhancement is input to the second γ correction unit 328 (# 18). Further, the γ correction unit is changed from the first γ correction unit 327 to the second γ correction unit 328.
(# 20). On the other hand, if the illumination unevenness correction has not been instructed (NO in # 14), the above-described step #
16 to # 20 are skipped.

Subsequently, infrared light for distance measurement is projected from the light projecting section 411 of the distance measuring section 41 toward the object (# 22), and the reflected light of the infrared light from the object is measured by the distance measuring section 41. 41 light receiving units 4
12 and the data for distance measurement is taken in (# 2).
4, # 26).

Subsequently, it is determined whether or not the mode is the oblique image correction mode (# 28 in FIG. 29). If the mode is the oblique image correction mode (YES in # 28), the image is taken from the captured distance measurement data. The distance D _A from the imaging surface to the subject at the center of the screen (point O in FIG. 16A) (the distance O in FIG. 8).
P) is calculated, and photographing angles of view αi and βi for each pixel position of the CCD 20 are calculated.

[0144] Then, the photographing angle .alpha.i, .beta.i, the object distance D _A and the object distance Di at each pixel position from the inclined angle theta (i.e., subject distance distribution in the imaging screen) is calculated, and further the focal length f and the subject photographing magnification m _a in the distance measuring point from the distance D _a is calculated, in the imaging magnification m _a, each pixel location by the calculation formula of the tilt angle θ and the photographing field angle .alpha.i, the from .beta.i (1) and (2) Shooting magnification mi, mi '
(That is, the distribution of the photographing magnification in the photographing screen) is calculated. Further, the magnification k is calculated by the above equations (4) and (6).
i and ki 'are calculated (# 30).

Subsequently, using the flowchart of the subroutine "data valid area calculation" shown in FIG. 31, the area of valid pixel data in the image after the oblique image correction (the area excluding the pixel position of the dummy data. Hereinafter, the data valid area) Is calculated (# 32).

In the calculation of the data valid area, first, the counter K for counting the pixel position is set to "1"(# 9).
0), the subject distance Di at the pixel position K = 1, the tilt angle θ, and the focal length f of the photographing lens 2 indicate the pixel position K =
The corrected pixel position (i ', j') of one image is calculated (# 92). As shown in FIG. 33, when the pixel data at the pixel position (i, j) is represented by g (i, j), the calculation of the data effective area is performed in the raster direction from the pixel position (1, 1). K = L · (I−1) + J
And the pixel data g (I, J) is the pixel data g (L ((I-1) + J)
It corresponds to.

Subsequently, it is determined whether the calculated corrected pixel position (i ', j') is within the effective image area (# 94). In this determination, the effective image area is (1 to
n, 1 to m), (i ′, j ′) becomes (1 to 1)
n, 1 to m). If the corrected pixel position (i ', j') is within the effective image area (YES in # 94), the data of the pixel position (i ', j') is stored in the memory 36 (# 9).
6) If the corrected pixel position (i ', j') is not within the effective image area (NO in # 94), step # 96 is skipped.

Subsequently, the count value of the counter K becomes "1".
After incrementing by # (# 98), it is determined whether or not this count value K is greater than the total number of pixels nm (# 98).
100), if K ≦ n · m (NO in # 100), the flow returns to step # 90 to determine whether or not the corrected pixel position of the next pixel data g (K) is in the effective image area. Done. Then, when K> nm (Y in # 100)
ES), it is determined that the above determination process has been completed for all pixel data, and the process returns.

Returning to the flowchart of FIG. 29, if it is determined in step # 28 that the photographing mode is the normal photographing mode (N in # 28)
O) Further, it is determined whether or not the illumination unevenness correction is instructed (# 34). If the illumination unevenness correction is not instructed (NO in # 34), the above steps # 30 and # 32 are skipped, and the illumination is performed. If unevenness correction is instructed (Y in # 34)
ES), the above steps # 30 and # 32 are skipped, and a light emission control signal is output to the light emission control unit 34 to stop light emission of the flash 7 (# 36). The fact that the emission of the flash 7 is prohibited in the normal shooting mode and when the illuminance unevenness correction is instructed is, for example, in a scene in which the white board 22 is shot from the front, the flash 7 automatically emits light. However, in this shooting scene, the flash light may be totally reflected by the whiteboard 22 and the characters of the shot image may not be legible, so that such shooting mistake is prevented.

[0150] Then, the object distance D lens driving amount for setting imaging lens 2 to an in-focus position based on _A is calculated (# 38), further the detected exposure control on the basis of the photometric data at the metering section 40 The value is calculated (# 40), whereby the photographing preparation processing ends and the camera enters a release standby state.

In the release standby state, when the shutter button 10 is fully pressed and the S2 switch is turned on (# 4)
(YES in 2), the process proceeds to step # 46, and the release operation is performed. On the other hand, if the shutter button 10 is half-pressed and the S1 switch is on, the release standby state is continued (loop of # 42 and # 44), the operation of the shutter button 10 is released, and the S1 switch is released. When it is turned off (NO in # 44), the process returns to step # 2.

When the operation shifts to the release operation, first, the data of the lens drive amount is output to the lens drive unit 37, and after the focus of the photographing lens 2 is adjusted (# 46), the data of the aperture value Av of the exposure control value is obtained. Is output to the aperture drive unit 39,
The aperture of the diaphragm 21 is adjusted (# 48).

Subsequently, it is determined whether or not illuminance unevenness correction has been instructed (# 50). If illuminance unevenness correction has been instructed (YES in # 50), the captured image is divided into a plurality of blocks. Block size is set (# 5)
2). The setting of the block size consists of a photographing magnification m _A calculated for the center of the screen and a preset reference photographing magnification m.
Set the block size _{S (= S0 · m A /} m0) in the center of the screen by using the 0 and block size S0, carried out by applying the block size as the block size in another position. That is, the captured image is divided into blocks of the size set at the center of the screen without changing the block size depending on the position of the screen.

Further, based on the distribution information of the photographing magnification mi and the set block size, a block in which pixel data is missing (a block including dummy data after correction) in a region where reduction processing is performed in the oblique image correction mode is extracted. Is performed (# 54). That is, as shown in FIG. 34, the oblique image G in FIG. 34A is corrected to the front image G ′ in FIG.
9) When divided into blocks, 11 blocks B (1,1) to 11 in the left half area of the reduced image G '
B (1,4), B (5,1) to B (5,4), and B (5,1) to B (5,4) are extracted as missing blocks of pixel data. On the other hand, if the illuminance non-uniformity correction is not instructed (NO in # 50), image processing is not performed for each block.
2, # 54 are skipped.

Subsequently, the data of the shutter speed calculated in step # 40 is output to the CCD driving section 31, and
The imaging operation (integration operation) by the CCD 20 is started (# 56). After resetting the charge of the photosensitive unit based on the drive control signal from the CCD drive unit 31, the CCD 20 captures an image of the subject by accumulating the charge in the photosensitive unit for a predetermined time (charge integration).

When the image pickup operation by the CCD 20 is completed,
Reading of the charges (pixel data) accumulated in each pixel of the photosensitive unit to the image processing unit 32 is started (# 58). CCD
As shown in FIG. 33, the 20 pixel data are sequentially read in the direction of the arrow for each vertical line and input to the image processing unit 32.

Subsequently, it is determined whether or not the mode is the oblique image correction mode (# 60 in FIG. 30). If the mode is the oblique image correction mode (YES in # 60), the pixel data read out from the CCD 20 is removed. Then, the oblique image correction operation unit 322 performs an oblique image correction process (# 62). The oblique image correction is performed for each vertical line. In an area where compression processing is performed, an area where pixel data is missing occurs. In this area, predetermined dummy data (for example, white data) is supplemented to perform oblique image correction. As shown in FIG. 35, for example, pixel data g1, g1 ', g2, and g at both ends of the line with respect to a pixel position where pixel data is lost in the vertical direction (a position indicated by oblique lines in FIG. 35A).
2 ′ is interpolated as dummy data g3, g3 ′, and, for example, the entire pixel data g4 of a known line is replaced with the pixel position where pixel data is lost in the horizontal direction (the position indicated by oblique lines in FIG. The interpolation may be performed as the dummy data g5. Then, the pixel data after the oblique image correction is stored in the image memory 324.

Subsequently, if a block from which pixel data is missing has been extracted in step # 54, a data valid area is calculated for each block (# 64).

Subsequently, it is determined whether or not illuminance unevenness correction has been instructed (# 66). If illuminance unevenness correction has not been instructed (YES in # 66), a subroutine "γ characteristic setting" shown in FIG. According to the flowchart of (1), the γ characteristic for illuminance unevenness correction for each block is set by the first γ characteristic setting unit 325 (# 68).

For setting the γ characteristic for correcting the illuminance unevenness of each block, first, a counter M for counting the number of blocks is set to “1” (# 110). The order of the blocks in the γ characteristic setting process is set in the raster direction in the block division shown in FIG.
M = L · (I−1) + J, and block B (I, J) corresponds to block B (L · (I−1) + J).

Subsequently, all pixel data in the data valid area included in the block B (M) is read (# 11).
2) A histogram as shown in FIG. 23 is created using the pixel data excluding the high-level X% of the pixel data (# 114). Subsequently, the class w corresponding to the peak value of the mountain U corresponding to the white portion of the histogram
Is calculated (# 116), and this class w is stored as the white saturation level W (M) of the γ characteristic for the block (M) (# 118).

Subsequently, the count value of the counter M becomes "1".
Then, it is determined whether or not the count value M is larger than the total number of blocks n (= K · L) (# 122), and if M ≦ n (# 122)
NO), the process returns to step # 110, and the next block B
The white saturation level W (I) is set for (M) (# 112 to # 120). When M> n, (#
YES at 122), it is determined that the setting of the white saturation level W (M) of the γ characteristic has been completed for all blocks B (M), and the routine returns.

Returning to the flowchart of FIG.
Interpolation is performed on the white saturation level W (I) of the γ characteristic for illuminance non-uniformity set for each block B (I), to correct illuminance non-uniformity at pixel positions other than the center position of each block B (I). Are set (# 70). Subsequently, while the set γ characteristic is input to the second γ correction unit 328,
The pixel data is read out from the image memory 324 to the second γ correction unit 328. The pixel data is subjected to gamma correction using the γ characteristic for illuminance unevenness correction corresponding to the pixel position, and then further to black color enhancement. Gamma correction is performed using the γ characteristic (# 72).

On the other hand, if illuminance unevenness correction has been instructed in step # 66 (NO in # 66), pixel data is read from the image memory 324 to the first γ correction unit 327, and this pixel data is set in advance. Gamma correction is performed using the gamma characteristic for a natural image (# 74).

Then, pixel data on which image processing such as gamma correction has been performed is sequentially written to the hard disk card 13 (loop of # 66 to # 76), and writing of all pixel data to the hard disk card 13 is completed. (YES in # 76), a control signal for ending the reading of the pixel data is output to the CCD driving section 31, and a control signal for ending the writing of the pixel data is output to the card driving section 33 (# 78). The photographing operation for one sheet is completed, and the process returns to step # 2 to perform the next photographing process.

In the above embodiment, histograms are created for all the set blocks B (I, J), and
The white saturation level W of the γ characteristic for illuminance non-uniformity correction is set from this histogram. FIG. 36 shows a case where the illuminance non-uniformity is relatively small in the vertical direction and large in the horizontal direction only. B (3,1), B (3,2),... B (3,9) in the horizontal direction passing through the center of the captured image G
, A white saturation level W of the γ characteristic is set from the histogram, and other blocks B (I, J) (I = 1, 2, 4, 5, J = 1, 2,... 9) For, the γ characteristic set in the block B (3, r) set in the column including the block may be applied. For example, block B (1,1) included in the first column,
Block B (3,1) for B (2,1), B (4,1), B (5,1)
Apply the γ characteristic set in.

On the other hand, when the illuminance unevenness is relatively small in the vertical direction and is large only in the horizontal direction, the vertical block B (1) passing through the center of the captured image G as shown in FIG. , 5), B (2,5),..., B (5,5), a white saturation level W of the γ characteristic is set from the histogram, and the other blocks B (I, J) (I =
For 1, 2,... 5, J = 1 to 4, 6 to 9), the γ characteristic set in the block B (r, 5) set in the row including the block may be applied. . For example, blocks B (1,1), B (1,2), B (1,1) included in the first row
3), B (1,4), B (1,6), B (1,7), B (1,8), B (1,9) are set in block B (1,5) Apply the γ characteristic. By doing so, it is possible to reduce the calculation time of the γ characteristic and the capacity of the memory for storing the set γ characteristic.

In the above-described embodiment, the entire captured image G is equally divided into a matrix and divided into blocks B (I, J).
Are set continuously, but a plurality of blocks B (I, J) may be set discretely in the imaging screen G as shown in FIG. In this manner, the number of blocks is reduced, so that the calculation time for setting the γ characteristic can be reduced as in the above example, and the capacity of the memory for storing the calculated γ characteristic can be reduced. Can be.

In the above embodiment, in the illuminance non-uniformity correction for the image after the oblique correction processing, a histogram of a level distribution is created using only valid pixel data in a block including dummy data, and the histogram is determined from the histogram. The gamma characteristic for illuminance non-uniformity correction for the block is set using the white saturation level W to be used. However, for such a block, the illuminance non-uniformity correction for the block is performed based on the level distribution of the pixel data included in the block. May not be set, and the γ characteristic for illuminance nonuniformity correction set in the neighboring block that does not include dummy data may be applied.

That is, in FIG. 34 (b), for example, blocks B (1,1), B (1,2), B (2,1) contain dummy data, so that these blocks B (1,1) ), B (1,2), B
For (2,1), the γ characteristic for illuminance unevenness correction set in the nearest block B (2,2) that does not include dummy data is applied. Similarly, for block B (1,3), the nearest block B (2,3) that does not include dummy data
The gamma characteristic for illuminance unevenness correction set in the above is applied, and for the block B (3,1), the γ for illuminance unevenness correction set in the nearest block B (3,2) that does not include dummy data. Apply properties.

In this case, the processing of the pixel data is performed as follows in FIG. That is, in the address generation unit 325b, the captured image is divided into a plurality of blocks B (I, J) by the block size input from the block size setting unit 325a, and the data effective area input from the data effective area calculation unit 323 is divided. While extracting the block B (I, J) containing the dummy data based on the information,
For each extracted block B (I, J), the nearest block B (I, J) that does not include dummy data is calculated. The calculation result is input from the address generation unit 325b to the white saturation level setting unit 325d as shown by the dotted line in FIG.

Then, pixel data is read from the image memory 324 in block units only for the block B (I, J) which does not include dummy data, and the histogram creation unit 32
In step 5c, a histogram of the level distribution of all pixel data included in the block is created, and a white saturation level W is set. The white saturation level setting unit 325d is the address generation unit 3
Block B containing dummy data input from 25b
Based on the information of the nearest block B (I, J) that does not include the dummy data for (I, J), for each block B (I, J) that includes the dummy data, the corresponding nearest block B (I, J) , J)
Is set for the white saturation level W.

If the area of the white background portion in the data effective area in the block is small, the peak U of the white background part may not be clearly shown in the histogram of the level distribution of the pixel data in the data effective area. There is a problem in the reliability of the white saturation level W of the γ characteristic for correction of the illuminance unevenness, but the method of applying the γ characteristic set in the nearest block not including the dummy data has such a problem. There is an advantage that reliability problems can be solved. In addition, since the γ characteristic of the closest block is applied, a false boundary line due to discontinuity in image quality due to a rapid change in γ characteristic between blocks does not occur.

As described above, in the method of dividing the captured image into blocks and detecting the level of the background (white) in the block from the histogram of the level distribution of the pixel data included in each block, the white If a white background portion other than the board 22 (for example, a background portion such as a white wall behind the white board) is included, the histogram does not have a typical two-peak distribution as shown in FIG. It is difficult to accurately detect the white background level.

For this reason, for example, when illuminance non-uniformity correction is performed in the case of monochrome photographing, a black color may be applied to the white board part especially in a block including the white board and the background part. The reason why such a false coloring phenomenon occurs is that the white saturation level set by the histogram is inappropriate.

That is, as shown in FIG. 39, in the block including the whiteboard and the background portion, the histogram of the level distribution of the pixel data corresponds to the whiteboard and the background portion as shown in FIG. It has a shape having two peaks U1 and U2 in a gray to white region.
Since the histogram of FIG. 40 is a block of the peripheral portion where the characters and the like of the whiteboard are not written, the histogram of FIG.
The mountain C corresponding to the character portion indicated by 9 does not appear.

For example, when the luminance of the background portion is higher than that of the whiteboard portion, such as when the whiteboard is arranged in front of a white wall having a high reflectance, the peak U1 corresponds to the background portion, and the peak U2 corresponds to the peak. Since gamma correction is performed using the gamma characteristic set as the white saturation level W for the class w corresponding to the peak value of the mountain U1, the white saturation level W is equal to or higher than the white saturation level W of the background portion. Is uniformly converted to a predetermined white, but a whiteboard portion having a white saturation level W or less is converted to a predetermined gray in a gray area.

Therefore, when gamma correction is further performed by the γ characteristic for black emphasis, the portion of the whiteboard portion lower than the black saturation level B is converted to black, which becomes the black coloration of the whiteboard portion. Will appear.

FIGS. 41A and 41B are diagrams for explaining black coloring by the above-described gamma correction for illuminance unevenness correction and gamma correction for black emphasis. FIG. 41A is a diagram showing a histogram.
FIG. 3B is a diagram illustrating a γ characteristic for correcting illuminance unevenness set based on a histogram, and FIG. 3C is a diagram illustrating a γ characteristic for black enhancement.

The white saturation level W of the γ characteristic shown in FIG. 17B is the value of the class w (= 200) corresponding to the peak value of the mountain U1 corresponding to the background portion of the histogram shown in FIG. Is set. Therefore, the class w ′ (= 10) corresponding to the peak value of the mountain U2 corresponding to the whiteboard portion
Assuming that 0) is a white level representing the whiteboard portion, the level value “100” is obtained by the γ characteristic shown in FIG.
Is gamma-corrected to a level value of “127”. After this, the black saturation level becomes “170”
When the gamma correction is performed using the gamma characteristic for black emphasis set to, the whiteboard portion at a level lower than the black saturation level is converted to a level value "0" and converted to black.

On the other hand, as shown in FIG. 42A, the level of the whiteboard portion is lower than the level of the background portion, but the area of the whiteboard portion in the block is larger than the area of the background portion. As shown in FIG. 41B, when the class w ′ (= 100) corresponding to the peak value of the peak U2 corresponding to the whiteboard portion is set to the white saturation level W of the γ characteristic, FIG. Even when black enhancement is performed using the same γ characteristic for black enhancement as the γ characteristic, the level value of the whiteboard portion converted to the level value “255” by the γ characteristic for uneven illuminance is converted to “0”. Since the whiteboard portion is maintained at the level value “255” without causing the whiteboard portion to fly appropriately white.

In color photography, a γ characteristic is set for illuminance unevenness correction for each of the R, G, and B color components. Therefore, in setting the γ characteristic of one of the color components, the white saturation level is determined by the level of the background portion. In this case, a false coloring phenomenon occurs in the whiteboard portion after the illumination unevenness correction and the gamma correction of black character enhancement.

[0183] R, G, white saturation level of γ characteristics of the respective color components B and _{W R, W G, W B} , for example, white saturation level W _R
Is set as in the example of FIG. 42, and the white saturation level W _G ,
If W _B is that is set as in the example of FIG. 41, for example R,
G and B color components D _R , D _G , and D _B are D _R = 200, D _G =
When gamma correction is performed on the whiteboard portion where D _B = 100, the red component of this white board portion is converted to D _R = 255, as described in the example of FIG. 42, and the green and blue components are converted to FIG. As described in the example of, D _G = D _B
= 0, so that the whiteboard portion is colored red.

When a whiteboard including a background is photographed in color, if a coloring phenomenon appears in a whiteboard portion in a boundary region between the whiteboard and the background of the captured image in the image processing, the image quality is remarkably deteriorated and becomes hard to see. Therefore, it is desirable to prevent the coloring phenomenon as much as possible.

In order to reliably prevent the coloring phenomenon, the photographer may configure the angle of view so that the entire whiteboard is included in the photographing screen. However, it is not always possible to configure the desired angle of view. Since there is no limitation, it is preferable that the coloring phenomenon can be reduced in the image processing.

As a method of reducing the coloring phenomenon in the image processing, a method of creating a histogram excluding pixel data of a portion (background portion) other than the whiteboard in the block can be considered. In this method, it is necessary to recognize only the whiteboard portion in the captured image. For example, the LCD display unit 19 displays the captured image in the release standby state on a monitor and provides an area designation member for the monitor display. It is preferable that the photographer can specify and input only the whiteboard portion by using the region specifying member. Alternatively, in general, a whiteboard is generally brighter white than a background portion and has higher luminance than the background portion. Therefore, a whiteboard may be automatically extracted using a luminance difference. In this method, a whiteboard portion is usually arranged at the center of the screen. Therefore, for example, a brightness value of the whiteboard portion is set from high-brightness pixel data at the center of the screen, and this brightness value and the brightness value of each pixel position ( (The level of the pixel data), it is possible to extract an area of only the whiteboard portion.

As another method for reducing the coloring phenomenon in image processing, as a γ characteristic for a block including a whiteboard and a background, the block closest to the block and not including the whiteboard and the background is used. The γ characteristic may be applied to the block so that the unevenness in the illuminance is not affected by the pixel data of the background portion.

FIG. 43 is a diagram showing the second γ characteristic setting section 325
FIG. 13 is a diagram showing a block configuration of the embodiment, in which the latter method is applied among the methods for reducing the coloring phenomenon in the image processing described above.

FIG. 28 is different from the block diagram of FIG. 15 in that a white saturation level correction unit 325h is added between the white saturation level setting unit 32d and the white saturation level interpolation calculation unit 325e, and the white saturation level correction unit 325h is added. A boundary block extraction unit 325g is added between the histogram creation unit 325c and the histogram creation unit 325c.

The boundary block extracting unit 325g extracts a block including a whiteboard and a background from the form of a histogram created for each block. That is, as shown in FIG. 41 (a) or FIG. 42 (a), a block having a shape having two peaks in a white region is extracted, and a γ characteristic to be applied to the extracted block. Is specified.

The white saturation level correction section 325h changes the white saturation level of the block extracted by the boundary block extraction section 325g to the white saturation level of the block designated by the boundary block extraction section 325g. This change is to change the white saturation level so that the white saturation level set in the block that does not include the whiteboard and the background and is close to the extracted block is applied to the extracted block. is there.

The extraction processing of the block including the whiteboard and the background in the boundary block extraction unit 325g is described in FIG.
The operation is performed as follows according to the flowchart of FIG.

First, a counter M for counting the number of blocks is set to "1"(# 130). Since the order of the blocks in the block extraction processing is performed in the raster direction in the block division shown in FIG. 20, M = L · (I−1) + J, and the block B (I, J)
Corresponds to block B (L (I-1) + J).

Subsequently, the class w corresponding to the peak value of the first peak of the histogram created for the block B (1)
Is detected. Taking the histogram shown in FIG. 40 as an example, a class w corresponding to the peak value of the mountain U1 is detected (# 132). The class w is performed in the same manner as the calculation of the class w for setting the white saturation level W of the γ characteristic for correcting the illuminance unevenness. That is, the class w having the highest frequency in the distribution included in the range d on the low level side from the maximum class p is calculated. In FIG. 40, the class p and the range d correspond to the class p (the maximum level of the pixel data included in the block B (1) and the range for detecting the class w in the histogram shown in FIG. Things.

Subsequently, a valley V between the peaks U1 and U2 is detected (# 134). The detection of the valley V is performed at a threshold value tnum (for example, the peak U
The calculation is performed by calculating a class lower than the peak value of 1 (frequency of 20% of fpnum) and determining whether or not this calculation result is continuous for a predetermined number tt (for example, 3 to 5) or more. . When tt or more classes lower than the threshold value tnum are continuous, the class range is detected as the valley V.

Subsequently, it is determined whether or not the valley V has been detected (# 136). If the valley V has not been detected (NO in # 136), the flow shifts to step # 144 to set the flag FLGM.
(M) is reset to “0”. Note that the flag FLGM
(M) is a flag indicating that the block includes a whiteboard and a background.

On the other hand, when the valley V is detected (Y in # 136)
ES), and a class w 'corresponding to the peak value of the second peak U2 is detected (# 138). The class w ′ is detected by calculating a class whose frequency is lower than the detected valley V and whose frequency is higher than a predetermined threshold value ynum (for example, a frequency of 30% of the peak value fpnum). The determination is made by determining whether or not the number yt (for example, 3 to 5) or more is continuous. When the number of classes higher than the threshold value ynum is continuous for yt or more, the class range is detected as the second mountain U2.

Subsequently, it is determined whether or not the second peak U2 is detected (# 140). When the second peak U2 is detected (YES in # 140), the flag FLGM (M) is set to "1". (# 142), and if the second peak U2 is not detected (NO in # 140), the process proceeds to step # 144.
The flag FLGM (M) is reset to “0”.

Subsequently, the count value of the counter M becomes "1".
Then, it is determined whether or not this count value M is greater than the total block number N (= kL) (# 148). If M ≦ N (NO in # 148) Then, the process returns to the step # 132, and the next block B (2)
Are performed as described above (# 132 to # 132).
148). Hereinafter, similarly, the same processing as described above is performed for each block B (M), and when M> N (# 14)
8), the determination as to whether or not all blocks include the whiteboard and the background has been completed, and the flag FLGM (M) set for each block is subsequently checked to determine whether the whiteboard and the background are included. Are extracted (# 150 to # 160).

In this block extraction, first, a counter M for counting the number of blocks is set to "1"(# 15).
0), it is determined whether or not the flag FLGM (1) for the block B (1) is set to “1” (# 15)
2) If the flag FLGM (1) has been reset to "0" (NO in # 152), the flow proceeds to step # 158.

If the flag FLGM (1) is set to "1" (YES in # 152), the flag FLGM (I) of the block B (I) adjacent to the block B (1) is further checked and the A block that does not include the board and the background is searched (# 154). This search is performed as shown in FIG.
This is performed for a 5 × 5 block centered on block B (M). The search method is shown in FIGS.
As shown in (5), the block B (M) is divided into four block groups of 3 × 3 blocks in which the block B (M) has a lower right corner, a lower left corner, an upper right corner, and a lower left corner, and is performed for each block group. It is.

In each block group, FIGS. 46 (a) to 46 (d)
The flag FLGM (I) is confirmed in the order of the number assigned to the block B, and the block B (I) in which FLGM (I) = 0 is first confirmed changes the γ characteristic of the block B (I) to the block B (M )
Is extracted as a block to be applied to the γ characteristic of.

In the case of block B (1), since block B (1) is located at the upper left corner of the captured image, FLGM (I) is confirmed for the adjacent block in the case shown in FIG.
For example, when FLGM (I) = 0 first in the third block (the block at the lower left of block B (1)),
The block B (I) is extracted as a block to be applied with the γ characteristic of the block B (I) to the γ characteristic of the block B (M).

Then, when the block B (I) is extracted,
The position information of the block B (I) is stored in association with the block B (1) (# 156).

Subsequently, the count value of the counter M becomes "1".
Then, it is determined whether or not this count value M is greater than the total block number N (# 148). If M ≦ N (NO in # 148), the flow returns to step # 152. Then, the same processing as described above is performed for the next block B (2) (# 152 to # 160).

Thereafter, the same processing as described above is performed for each block B (M), and when M> N is satisfied (# 1)
(YES at 60), it is determined that the designation of the block to which the γ characteristic is to be applied has been completed for all the blocks including the whiteboard and the background, and the process ends.

The information of the block extracted by the boundary block extraction unit 325g and the block having the γ characteristic to be applied to this block is input to the white saturation level correction unit 325h, and set for each block by the white saturation level setting unit 325d. Of the white saturation levels set, the white saturation level set for the block including the whiteboard and the background is determined by the white saturation level correction unit 325h for the block not including the specified whiteboard and the background. Replaced by the set white saturation level.

As described above, in the first γ characteristic setting unit 325 according to the second embodiment, for a block including a whiteboard and a background, the first γ characteristic is set using the histogram of the level distribution of the pixel data included in the block. Since the white background level of the board is not detected and the gamma characteristic is set using the whiteboard level of the whiteboard detected in a block that does not include the whiteboard and the background, it is based on an inappropriate setting of the gamma characteristic. Coloring at the boundary with the whiteboard background can be reliably prevented.

The first γ according to the second embodiment is
The characteristic setting unit 325 detects whether or not the block includes the whiteboard and the background by using the histogram of each block, but uses the detection result to determine the boundary position between the whiteboard and the background. Can be detected.

That is, since the flag FLGM indicating whether or not there is a boundary between the whiteboard and the background is set for each block, the boundary between the whiteboard and the background is extracted by extracting the block of FLGM = 1. A block including the block (hereinafter, this block is referred to as a boundary block) can be extracted.

The outline of the shape and position of the boundary can be known by connecting the extracted boundary blocks. When the captured image in which the whiteboard is projected in the center of the screen is divided into, for example, rectangular blocks as shown in FIG. 47, when a block of FLGM = 1 is extracted, the boundary including the boundary S indicated by stippling in FIG. Blocks are extracted,
By linking the boundary blocks, the outline shape (substantially horizontally long rectangle) of the boundary S can be determined.

Therefore, if the boundary position in each boundary block can be estimated in pixel units, the boundary position in the photographing screen, that is, the area of the whiteboard can be accurately known, and the γ characteristic between the whiteboard portion and the background portion can be determined. If they are made different, it is possible to obtain a more favorable image quality for the character information in the whiteboard even when the background is included in the shooting screen. Further, when the captured image is binarized to make the contrast between the background part (white background part) and the character part clear, it is preferable to change the binarization threshold between the whiteboard part and the background part. Binarization processing can be performed.

Next, a method of estimating the boundary position in each boundary block in pixel units will be described.

The image in the block is almost whiteboard white and background white (lower brightness than whiteboard)
In addition, when there is a certain luminance difference between the whiteboard portion and the background portion, the distribution of the histogram includes two peaks having a peak U1 of the whiteboard portion and a peak U2 of the background portion as shown in FIG. It becomes distribution, and both mountains U
1, U2 has a relatively sharp shape.

In such a case, the pixel data of the whiteboard portion in the block gathers around the peak U1 and the pixel data of the background portion gathers around the peak U2, so that the valley t1 between the peaks U1 and U2. Is divided into an area A1 having the level t1 or more and an area A2 having the level t1 or less, and the frequencies included in the areas A1 and A2 are a1 and a2, respectively. The value is close to the ratio S1: S2 between the area S1 of the board portion and the background portion S2.

In FIG. 47, for example, if the histogram created for the boundary block B1 including the horizontal boundary S is as shown in FIG. 48, the boundary S is substantially horizontal in the boundary block B1, It can be estimated that there is a boundary S at a position that internally divides the block B1 in the vertical direction at a power ratio a1: a2. The boundary position in the boundary block B2 including the boundary S in the vertical direction can be estimated in the same manner.

Therefore, the boundary block B adjacent in the horizontal direction
The frequency ratios a1: a2 are compared between the two, and if they are substantially the same or similar, it is estimated that these blocks B1 are blocks including only the horizontal boundary S. Similarly, the frequency ratios a1 ': a2' are compared between boundary blocks B2 adjacent in the vertical direction, and when they are substantially the same or similar, these blocks B2 are blocks including only the vertical boundary S. presume.

On the other hand, in the block B3 including the corner of the whiteboard, the frequency ratio a of the horizontal boundary block B1 is set.
1: Both a2 and the frequency ratio a of the vertical boundary block B2
Since a frequency ratio a1 ": a2" completely different from 1 ': a2' appears, the block B3 having such a frequency ratio a1 ": a2" is estimated to be a corner block, and as shown in FIG. Frequency ratio a1: a of the horizontal block B1
2 and the frequency ratio a1 'of the block B2 in the vertical direction adjacent thereto: a
It is estimated that there is a boundary S at a position that internally divides the inside of the block B3 into a key shape using 2 ′.

Since the corner block B3 can be determined from the connection state of the boundary blocks, the frequency ratio a1: a2 of the adjacent horizontal boundary block B1 and the frequency ratio a1 ': a2 of the adjacent vertical block B2. ′ May be used to immediately divide the inside of the block B3 into a key shape to estimate the position of the boundary S.

In the method of detecting the white background level of the whiteboard from the histogram of the level distribution of the pixel data, the whiteboard level of the whiteboard can be obtained even when a halftone material such as a photograph or a chart is attached on the whiteboard. Is difficult to detect accurately.

The reason is that, when the image in the block is a halftone image such as a photograph, most of the pixel data is spread in a gray area, and the histogram set for the block is, as shown in FIG. The mountain U and the mountain C in the character portion are not clearly shown, and the class w corresponding to the peak value of the mountain U in the white portion used for the histogram shown in FIG. This is because the class w corresponding to the peak value of the mountain U in the white background portion cannot be detected accurately and reliably even if the detection method is applied.

On the other hand, in a block including a halftone image such as a photograph, if the γ characteristic close to the binarization processing is performed, the image quality of the photograph or the like is rather deteriorated and an unnatural image is obtained. It is desirable to perform the same gamma correction as for a natural image.

Therefore, first, a block including a halftone image is extracted from the form of the histogram created for each block, and the γ characteristic (for example, applied to the first γ correction unit) for the predetermined halftone is extracted from the block. Γ
It is preferable to perform gamma correction based on characteristics or the like.

The following two methods can be applied to determine whether or not a block is a block including a halftone image from the shape of the histogram.

The first method is applied when the maximum frequency fp of the histogram is smaller than a predetermined threshold thd and the peak of the histogram is low, as shown in FIG. 49, from the class w having the maximum frequency fp. The variance is calculated for a distribution within a range bd set in advance on the low level side, and the calculation result is compared with a predetermined threshold thb to determine. The reason why only the pixel data within the range bd is used in the variance calculation is to reduce the influence of the pixel data corresponding to the character information. The threshold thb is a variance threshold that is experimentally obtained in advance and can be estimated as a whiteboard portion.

Therefore, the calculated variance is equal to the predetermined threshold value thb.
If it is larger, it is determined that the image of the block is not an image of the whiteboard portion.

The second method is applied when the variation of the histogram is relatively large. In the low level side from the class w having the maximum frequency fp, the Z% of the maximum frequency fp is obtained.
(For example, 50%), the class ph closest to the class w is calculated, and the class difference Δw (= w-ph) between the class ph and the class w having the maximum frequency fp is calculated as a predetermined value. This is determined by comparing with the threshold thph. Note that the predetermined threshold value thph is a class difference threshold value that is experimentally obtained in advance and can be estimated as a whiteboard portion.

Therefore, when the calculated class difference Δw is larger than the predetermined threshold thph, it is determined that the image of the block is not an image of the whiteboard portion.

In the above embodiment, the gamma correction of the image after the illuminance unevenness correction is performed using the γ characteristic for black emphasis in which the black saturation level B is variably set by the black density adjustment switch 18. However, instead of the gamma characteristic for black enhancement, gamma correction is performed so that the background of the captured image is skipped white using the gamma characteristic for uneven illuminance correction, and the lowest level h 'of the pixel data is calculated. h '
The gamma correction may be performed by using a γ characteristic for converting the level between the gray level and the white saturation level w in 256 gradations.

That is, assuming that a histogram as shown in FIG. 50A is obtained for the image after the illuminance unevenness correction, a class h 'having the lowest luminance in the black region is calculated from the histogram, and the class h' is calculated. The gamma characteristic shown in FIG. 4B is set using the white saturation level w. Then, gamma correction of the image after the illuminance unevenness correction may be performed using the γ characteristic.

In this method, gamma correction is performed so as to expand the luminance range of the image after the illuminance unevenness correction to 256 gradations, so that the dynamic range is widened and the quality of the captured image can be improved.

In the above embodiment, the case where the image distortion correction function for correcting perspective geometric distortion in oblique photographing and the illuminance unevenness correction function are combined is described, but the present invention is not limited to this. In addition, the present invention can be applied to a case where an image distortion correction function for correcting geometric distortion based on optical characteristics of an optical imaging system or geometric distortion based on processing characteristics of a signal processing system and an illuminance unevenness correction function are combined. it can.

[0233]

As described above, according to the present invention,
Digital camera capable of correcting geometric distortion, after correcting the geometric distortion, the image is divided into a plurality of small images, γ characteristics for each small image is set for each small image, Illuminance unevenness is corrected by performing gamma correction on a small image basis, so that the character density included in the small images is substantially uniform, and appropriate for each small image using the pixel signal that constitutes each small image. Γ characteristics can be set stably. This makes it possible to perform suitable illuminance unevenness correction even when an image has geometric distortion.

Further, according to the present invention, after correcting an oblique image photographed in an oblique direction to a pseudo front photographed image and then correcting the illuminance unevenness, for example, a character written on a white board Even if the image is obtained by obliquely capturing information such as a figure, it is possible to obtain an image with little informational distortion due to oblique photography and high illuminance unevenness and high information value.

[Brief description of the drawings]

FIG. 1 is a perspective view showing the appearance of a digital camera according to the present invention.

FIG. 2 is a rear view of the digital camera according to the present invention.

FIG. 3 is a diagram illustrating oblique photographing of a subject.

FIGS. 4A and 4B are diagrams for explaining oblique image correction, wherein FIG. 4A is a diagram showing an oblique captured image, and FIG. 4B is a diagram showing an image after oblique image correction.

FIG. 5 is a diagram illustrating an example of an illumination direction of illumination light of a whiteboard.

FIGS. 6A and 6B show the output distribution of the image sensor, in which FIG. 6A shows the output distribution in the vertical direction, and FIG. 6B shows the output distribution in the horizontal direction.

FIG. 7 is a schematic configuration diagram of an optical system of the digital camera according to the present invention.

FIG. 8 is a diagram of the imaging system in oblique imaging as viewed from directly above.

FIG. 9 is a diagram illustrating a state in which a captured image is divided into a plurality of blocks.

FIG. 10 is a diagram illustrating an example of a γ characteristic that emphasizes a white background set for each block.

FIG. 11 is a diagram illustrating an example of a γ characteristic in which a black portion is emphasized.

FIG. 12 is a diagram illustrating a relationship between black adjustment by a black density adjustment switch and a γ characteristic that emphasizes a black portion.

FIG. 13 is a block diagram of a digital camera according to the present invention.

FIG. 14 is a block diagram showing a configuration from an oblique image correction operation unit to first and second γ correction units for performing image processing of a color image.

FIG. 15 is a block diagram illustrating an internal configuration of a first γ characteristic setting unit.

FIG. 16 is a diagram for explaining a method of correcting an oblique image.
(A) is a figure which shows a diagonal image, (b) is a figure which shows the image after diagonal image correction.

FIG. 17 is a diagram showing a corrected image in which a missing portion of pixel data in a compression process is corrected.

FIG. 18 is a diagram for explaining the capacity of an image memory.

FIG. 19 is a diagram showing a general form of a histogram of pixel data forming a character image.

FIG. 20 is a diagram showing a state in which a captured image is divided into small images of a plurality of blocks.

FIGS. 21A and 21B show a state in which a captured image is divided into blocks of inappropriate sizes, where FIG. 21A shows a case where the block size is smaller than an appropriate value, and FIG. FIG.

FIG. 22 is a diagram showing a state in which a block frame is displayed in a finder field frame.

FIG. 23 is a diagram illustrating an example of a histogram of pixel data constituting a small image divided into blocks.

FIG. 24 is a diagram illustrating a γ characteristic determined using a histogram of pixel data.

FIG. 25 shows γ set using green component pixel data.
It is a figure showing an example of a characteristic.

26A and 26B are diagrams showing γ characteristics set for each color component using the pixel data of each color component, where FIG. 26A is for pixel data of a red component, FIG. 26B is for pixel data of a green component, (C) is for the blue component pixel data.

FIG. 27 is a diagram for describing interpolation calculation of γ characteristics for pixel data in an area surrounded by the center positions of four adjacent blocks.

FIG. 28 is a flowchart illustrating shooting control of the camera according to the present invention.

FIG. 29 is a flowchart showing shooting control of the camera according to the present invention.

FIG. 30 is a flowchart illustrating shooting control of the camera according to the present invention.

FIG. 31 is a flowchart of a subroutine “data valid area calculation”;

FIG. 32 is a flowchart of a subroutine “γ characteristic setting”.

FIG. 33 is a diagram illustrating a reading direction of pixel data of a CCD.

FIG. 34 is a diagram for explaining a method of oblique image correction.
(A) is a figure which shows a diagonal image, (b) is a figure which shows the pseudo front image after diagonal image correction.

35A and 35B are diagrams for explaining pixel data interpolation processing in oblique image correction, where FIG. 35A is a diagram illustrating vertical interpolation processing, and FIG. 35B is a diagram illustrating horizontal interpolation processing.

FIG. 36 is a diagram for describing a method of setting γ characteristics for illuminance unevenness correction for another block using γ characteristics for illuminance unevenness set in blocks arranged in the horizontal direction.

FIG. 37 is a diagram for explaining a method of setting γ characteristics for illuminance unevenness correction for another block using γ characteristics for illuminance unevenness set in blocks arranged in the vertical direction.

FIG. 38 is a diagram illustrating another block division method of a captured image for performing illumination unevenness correction.

FIG. 39 is a diagram showing a block including a boundary between a whiteboard and a background.

FIG. 40 is a diagram illustrating an example of a histogram created for a block including a whiteboard and a boundary.

FIG. 41 is a diagram for explaining the reason why the coloring phenomenon occurs in the illuminance unevenness correction and the γ correction processing for black enhancement,
(A) is a diagram showing a histogram, (b) is a diagram showing a γ characteristic for correcting illuminance unevenness, and (c) is a diagram showing a γ characteristic for black enhancement.

FIG. 42 is a diagram for explaining the reason why coloring phenomenon does not occur in the γ correction processing for illuminance unevenness correction and black emphasis,
(A) is a diagram showing a histogram, (b) is a diagram showing a γ characteristic for correcting illuminance unevenness, and (c) is a diagram showing a γ characteristic for black enhancement.

FIG. 43 is a block diagram illustrating an internal configuration of a first γ characteristic setting unit that reduces a coloring phenomenon in illuminance unevenness correction and black emphasis correction.

FIG. 44 is a diagram showing a flowchart of “block extraction” showing control for extracting a block including a background.

FIG. 45 is a diagram illustrating a search range for detecting a block that does not include the background around a block that includes the background.

FIGS. 46A and 46B are diagrams for explaining a search order of blocks not including a background, wherein FIG. 46A is a diagram illustrating a search order of blocks in the upper left region with respect to a block including a background, and FIG. Is a diagram showing the search order of the blocks in the upper right area, (c) is a diagram showing the search order of the blocks in the lower left area with respect to the block including the background, and (d) is the lower right area with respect to the lock including the background. FIG. 5 is a diagram showing a search order of blocks.

FIG. 47 is a diagram illustrating a method of detecting a block including a boundary between a whiteboard and a background.

FIG. 48 is a diagram illustrating an example of a histogram created for a block including a whiteboard and a background.

FIG. 49 is a diagram illustrating an example of a histogram created for a block including only a halftone image such as a photograph.

FIG. 50 is a diagram for explaining a method of expanding a dynamic range in γ correction in illuminance unevenness correction.
(A) is a figure which shows a histogram, (b) is a figure which shows a gamma characteristic.

FIG. 51 is a diagram illustrating a method of dividing a captured image into blocks in a digital copying machine.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Camera (digital camera) 2 Shooting lens 3 Metering window 4 Distance measuring light emitting window 5 Distance measuring light receiving window 6 Viewfinder objective window 7 Flash 8 Card slot 9 Card eject button 10 Shutter button 11 Zoom switch 12 Shooting / playback switch Reference Signs List 13 hard disk card 14 main switch 15 finder eyepiece window 16 mode setting switch (angle setting means) 17 illuminance unevenness correction switch 18 black density adjustment switch 19 LCD display unit 20 CCD area sensor (imaging means) 21 aperture 22 white board 30 CPU 301 shooting Magnification calculation unit (calculation means) 31 CCD drive unit 32 Image processing unit 321 A / D converter 322 Oblique image correction calculation unit (distortion correction unit) 323 Data valid area calculation unit 324 Image memory 325 First γ characteristic setting unit 325a Block size setting unit (image division unit) 325b Address generation unit 325c Histogram creation unit 325d White saturation level setting unit 325e White saturation level interpolation calculation unit 325f γ characteristic setting unit (γ characteristic setting unit) 325g Boundary block extraction unit 325h White saturation level correction Unit 326 Second γ characteristic setting unit 327 First γ correction unit 328 Second γ correction unit (gamma correction unit) 329 Switch circuit 33 Card drive unit 34 Light emission control unit 35 LCD drive unit 36 Memory 37 Lens drive unit 38 Zoom drive unit 39 Aperture drive Unit 40 Photometry unit 41 Distance measurement unit (distance measurement means) 42 Viewfinder field frame 43 Block frame

Claims (2)

[Claims] An image pickup means having a pixel composed of a plurality of photoelectric conversion elements, and taking in an object light image by photoelectrically converting it into a pixel signal, and partially enlarging or reducing an image taken in by the image pickup means. A digital camera comprising: a distortion correction unit configured to correct a geometric distortion of the image; an image division unit configured to divide the image corrected by the distortion correction unit into a plurality of small images; Gamma characteristic setting means for setting a gamma characteristic for the small image using a pixel signal to be performed, and gamma correction means for performing gamma correction using the gamma characteristic set by the gamma characteristic setting means for each of the small images. A digital camera characterized in that: 2. The digital camera according to claim 1, wherein the imaging lens forms the light image of the subject on an imaging surface of the imaging unit, and an angle for setting an angle between the imaging surface of the imaging unit and the object surface. Setting means, a distance measuring means for measuring a distance to a subject, and an angle set by the angle setting means, a focal length of the photographing lens, and a subject distance measured by the distance measuring means. Calculating means for calculating the distribution of the photographing magnification, wherein the distortion correcting means corrects perspective distortion caused by oblique photographing by partially enlarging or reducing the captured image based on the distribution of the photographing magnification. A digital camera characterized by being a thing.