JP2004274367A - Digital camera - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a digital camera capable of properly adjusting the white balance when taking pictures of a scene with a green background such as in a dark forest or woods. <P>SOLUTION: The white balance gains Ra, Ga and Ba which are the outputs of a white balance gain calculating section 40 for obtaining the white balance gains on the basis of the luminance data and hue data of a photographed image are corrected by a white balance gain correction section 60 by using a green correction amount Grni from a green data operating section 50, and white balance correction is performed by the corrected white balance gains Rg, Gg and Bg. The section 50 obtains the green correction amount Grni being the data indicating the characteristics of the scene with green background of an object on the basis of a distribution in R/G-B/G coordinates of the hue data. To obtain the green correction amount Grni, a regression line based on the hue data is obtained, and then the green correction amount is obtained by using the inclination of the regression line and an error between the regression line and the coordinate data of the hue data in a green detecting frame. <P>COPYRIGHT: (C)2004,JPO&NCIPI

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a digital camera having a white balance adjustment function corresponding to a shooting scene.
[0002]
[Prior art]
Conventionally, in a digital camera having a function of determining whether a shooting scene is an outdoor scene or an indoor scene, the shooting scene is determined based on the hue (RGB information) of the subject, and according to the determination result The white balance (hereinafter sometimes referred to as “WB”) is automatically adjusted. However, if the shooting scene is determined based only on the hue of the subject, erroneous determination of the shooting scene may occur if the subject's color temperature distribution is uneven or if the subject is a single color or similar color object. This is likely to occur and the white balance is often misadjusted. As a result, color failure has occurred due to white balance deviation.
[0003]
Therefore, in the invention described in Patent Document 1, based on the hue and brightness of the subject, it is determined whether the shooting scene is an outdoor scene or an indoor scene, thereby determining the shooting scene misclassification. The white balance is prevented from shifting.
[0004]
[Patent Document 1]
Japanese Patent Laid-Open No. 2000-224608
[Problems to be solved by the invention]
However, in a green scene, especially in a dark forest or forest where there is not much external light, the hue data distribution is under the illumination of white fluorescent lamps, metal halide light sources, mercury lamps, etc. Since it is very similar to a shooting scene, even in the invention described in Patent Document 1, the shooting scene may not be accurately determined.
[0006]
In this case, the white balance adjustment suitable for a white fluorescent lamp or the like for indoor photography is mistakenly performed despite outdoor photography. For example, if you take a picture of a person with a green plant in the background and make an incorrect white balance adjustment, the color of the green plant becomes grayish and the skin color becomes magenta. Becomes very bad.
[0007]
The present invention has been made in view of the above circumstances, and an object of the present invention is to provide a digital camera capable of appropriately adjusting the white balance at the time of shooting a green scene such as in a dark forest or a forest.
[0008]
[Means for Solving the Problems]
The digital camera of the present invention is a digital camera having a white balance adjustment function according to a shooting scene, and a luminance detection means for detecting the luminance of the subject, and the photographed image of the subject is divided into a plurality of areas. A hue detection means for detecting the hue of the image, a white balance adjustment data calculation means for obtaining white balance adjustment data based on the luminance data and the hue data, and distribution of the hue data in R / GB / G coordinates Based on the data, green data calculating means for obtaining data indicating the green scene likeness of the subject, white balance correcting means for correcting the white balance adjustment data based on the data obtained by the green data calculating means, and the white Based on the white balance adjustment data corrected by the balance correcting means, White balance adjustment means for adjusting the color balance, and the green data calculation means obtains a regression line based on the hue data existing in the green detection frame in the R / GB / G coordinates, and the regression line The data indicating the green scene-likeness is obtained based on the inclination of the image and the error between the regression line and the coordinate data of the hue data in the green detection frame.
[0009]
Here, R / G indicates the ratio of the signal strength of the R (red) component and the signal strength of the G (green) component, and B / G is the signal strength of the B (blue) component and the G (green) component. The ratio to the signal strength is shown. The R / GB-G / G coordinates are coordinates with R / G on the x-axis (horizontal axis) and B / G on the y-axis (vertical axis). The white balance adjustment data is, for example, a white balance adjustment gain. Further, the data indicating the likelihood of a green scene is calculated as correction amount data for correcting white balance adjustment data obtained based on luminance data and hue data, for example.
[0010]
The digital camera according to the present invention includes a digital camera in which the green data calculation means obtains data indicating the green scene likeness based on the number of previous hue data whose error is within a predetermined range. In this way, the green scene-likeness can be obtained more systematically.
[0011]
In the digital camera of the present invention, the green data calculation means is a data indicating the green scene-likeness depending on whether a detection frame of a predetermined light source in the R / GB / G coordinates exists on an extension line of the regression line. Including those that modify In this way, the influence of other light sources and the influence of outdoor daylight can be accurately distinguished, and white balance adjustment can be performed appropriately.
[0012]
The digital camera of the present invention includes one in which the white balance correction means performs correction using white balance adjustment data in the case of fine weather according to the data indicating the green scene likeness.
[0013]
In the digital camera of the present invention, the white balance adjustment data calculation unit determines whether the shooting scene is an outdoor scene or an indoor scene based on the luminance data and the hue data, and determines the color temperature of the subject. , Including those for obtaining white balance adjustment data based on the determined data.
[0014]
DETAILED DESCRIPTION OF THE INVENTION
Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 is a diagram showing a schematic configuration of a digital camera showing an embodiment of the present invention. Note that FIG. 1 mainly shows the functions of the digital camera, and does not accurately correspond to the hardware configuration.
[0015]
In the digital camera shown in FIG. 1, the subject image formed on the light receiving surface of a solid-state imaging device (hereinafter referred to as “CCD”) 13 through the photographing lens 11 and the diaphragm 12 is converted into each photoelectric conversion unit of the CCD 13. Thus, the signal charge is converted into an amount corresponding to the amount of incident light and stored. The signal charge accumulated in each photoelectric conversion unit is read to the shift register by a read gate pulse given from the CCD drive circuit 14, transferred by a register transfer pulse, and sequentially read as a voltage signal corresponding to the signal charge. The CCD 13 has a so-called electronic shutter function that can sweep out the accumulated signal charge with a shutter gate pulse and thereby control the charge accumulation time (shutter speed).
[0016]
The voltage signal sequentially read from the CCD 13 is sent to a correlated double sampling circuit (CDS circuit) 15 where the R, G, B signals for each pixel (photoelectric conversion unit) are sampled and held, and A / D It is sent to the converter 16. The A / D converter 16 converts the R, G, and B signals sequentially given from the CDS circuit 15 into 10-bit (0 to 1023) digital R, G, and B signals and outputs the signals. The CCD drive circuit 14, the CDS circuit 15, and the A / D converter 16 are driven in synchronism with a timing signal supplied from the timing generation circuit 17.
[0017]
The R, G, and B signals output from the A / D converter 16 are sent to the CPU 19 through the integrating circuit 35, and once stored in the memory 18, are then sent to the digital signal processing circuit 20. Multipliers 36R, 36G, and 36B are provided between the integrating circuit 35 and the CPU 19, and an adjustment gain value for adjusting variations in the devices is input to the multipliers 36R, 36G, and 36B. It has become. The integration circuit 35 integrates R, G, and B signals for each area obtained by subdividing the subject image into a plurality of areas (here, 64 × 64 divisions). The CPU 19 performs overall control of each circuit based on input from the camera operation unit 31 including a shutter button and the like, and executes a series of controls such as auto focus control and automatic exposure control to control the entire photographing operation of the digital camera. To do. Also, white balance gain calculation processing described later is performed.
[0018]
The digital signal processing circuit 20 includes a synchronization circuit 21, a white balance adjustment circuit 22, a gamma correction circuit 23, a YC signal generation circuit 24, and a memory 25. The synchronization circuit 21 converts the dot-sequential R, G, B signals read from the memory 18 into simultaneous equations, and outputs the R, G, B signals to the white balance adjustment circuit 22 simultaneously.
[0019]
The white balance adjustment circuit 22 includes multipliers 22R, 22G, and 22B for adjusting digital values of the R, G, and B signals. The R, G, and B signals are respectively multiplied by the multipliers 22R, 22G, 22B. The other inputs of the multipliers 22R, 22G, and 22B are gain values Rg, Gg, and Bg, which are white balance adjustment data for each of the R, G, and B signals. A multiplication signal by the multipliers 22R, 22G, and 22B is white. The adjusted R ′, G ′, and B ′ signals are output to the gamma correction circuit 23.
[0020]
The gamma correction circuit 23 changes the input / output characteristics so that the white balance adjusted R ′, G ′, B ′ signals have the desired gamma characteristics, and the 10-bit signal becomes an 8-bit signal. And output to the YC signal generation circuit 24. The YC signal creation circuit 24 creates a luminance signal Y and chroma signals Cr and Cb from the gamma-corrected R, G, and B signals. These luminance signal Y and chroma signals Cr and Cb (YC signal) are stored in the memory 25. The YC signal stored in the memory 25 at the time of shooting is compressed into a predetermined format by a compression circuit (not shown) and then recorded on a recording medium such as a memory card.
[0021]
Next, gain values Rg, Gg, and Bg that are white balance adjustment data output from the CPU 19 will be described. When the white balance setting is set to the manual mode by the operation of the camera operation unit 31, a gain value corresponding to the set light source is output. On the other hand, when the white balance setting is set to the auto mode, the shooting scene is determined based on the integrated value from the integrating circuit 35, and a gain value corresponding to the shooting scene is output. When the strobe 34 is used, a gain value corresponding to the strobe light source is output regardless of the white balance setting mode.
[0022]
FIG. 2 is a schematic functional block diagram of a white balance gain calculation processing unit that calculates the white balance gain in the auto mode. The white balance gain calculation processing unit calculates the white balance gain using the integrated values of the R, G, and B signals that are output from the integrating circuit 35, and is realized by the CPU 17 executing a predetermined program. The
[0023]
2 includes a white balance gain calculation unit (WB gain calculation unit) 40, a green data calculation unit 50, and a white balance gain correction unit (WB gain correction unit) 60, and includes a WB gain. The calculation unit 40 includes a luminance / hue calculation unit 41, an outdoor / indoor determination unit 42, a color temperature determination unit 43, and a WB gain selection unit 44.
[0024]
The luminance / hue calculation unit 41 calculates luminance data (Y) and hue data (R, G) of the subject image from the integrated value of each area obtained by subdividing the subject image into a plurality of areas (here, 64 × 64 divisions). , B) are calculated and given to the outdoor / indoor determination unit 42 and the color temperature determination unit 43, respectively. The outdoor / indoor determination unit 42 determines whether the shooting scene is an outdoor scene or an indoor scene based on the luminance of the subject, and supplies the determination to the WB gain selection unit 44. Further, the color temperature determination unit 43 determines the color temperature of the subject based on the hue data of the subject and similarly supplies the color temperature to the WB gain selection unit 44.
[0025]
The WB gain selection unit 44 determines the light source of the subject based on the outputs of the outdoor / indoor determination unit 42 and the color temperature determination unit 43, and selects the WB gains Ra, Ga, and Ba corresponding to the determined light source. WB gains corresponding to various light sources are stored in advance in a storage unit (not shown).
[0026]
The green data calculation unit 50 calculates data indicating the likelihood of the subject's green scene based on the integration value of the integration circuit 35, and the calculated data indicating the likelihood of the green scene is the WB gain correction unit as the green correction amount Grni. 60. The green scene-likeness of the subject indicates a certain level in an outdoor green scene such as a leaf green. Note that a method of calculating the green correction amount Grni, which is data indicating the likelihood of a green scene, will be described later.
[0027]
The WB gain correction unit 60 corrects the WB gains Ra, Ga, and Ba corresponding to the light source according to the green correction amount Grni. If the subject looks like a green scene, the WB gains Rd, Gd, and Bd of the daylight are corrected. To a gain value that takes into account. Specifically, the corrected gain values Rg, Gg, and Bg are obtained by the following equation.
[0028]
Rg / Gg = (Ra / Ga) × Grni + (Rd / Gd) × (1-Grni)
Bg / Gg = (Ba / Ga) × Grni + (Bd / Gd) × (1-Grni)
[0029]
That is, when it is determined that the scene is completely green, the green correction amount Grni = 0, the WB gain is the WB gains Rd, Gd, and Bd of the daylight, and the image is an indoor light source. When the determination is made, the green correction amount Grni = 1 is set, and the WB gains Ra, Ga, Ba calculated by the WB gain calculation unit 40 are set. If it is not a completely green scene, but it seems to be a green scene, the WB gain corresponding to the light source is changed by changing the green correction amount Grni between 0 and 1 in accordance with the degree of the green scene. Values obtained by adding Ra, Ga, Ba and daylight WB gains Rd, Gd, Bd at a ratio corresponding to the green correction amount Grni are output as corrected gain values Rg, Gg, Bg.
[0030]
Next, the calculation of the green scene likeness by the green data calculation unit 50 will be described. The green scene-likeness is calculated using data obtained by plotting 64 × 64 hue data obtained by subdividing the subject image into R / GB / G coordinates as shown in FIG. FIG. 3 shows emission colors of various light sources and hue detection frames of various subjects on R / GB / G coordinates, and a green scene is mainly plotted in a green detection frame 81 of FIG. As a result of plotting, when a predetermined number or more of the hue data of 64 × 64 divided areas enter the green detection frame 81, it is determined that the hue of the subject is green. However, in the green detection frame 81, not only the photographed image of the green scene but also the photographed image under the illumination of a white fluorescent lamp, a metal halide light source, a mercury lamp, etc. are plotted as described above. Or the influence of the light source is determined by the following calculation.
[0031]
When a predetermined number or more of hue data is plotted in the green detection frame 81, a regression line is obtained by the least square method based on the distribution of 64 × 64 hue data on the R / GB / G coordinates. Based on the number of hue data present in the green detection frame 81, the slope of the regression line, and the error between the regression line and the coordinate data of the hue data in the green detection frame 81, the green scene likelihood is calculated.
[0032]
FIG. 4 shows a schematic flow of greenness determination processing. The digital camera shown in FIG. 1 can be set by the camera operation unit 31 to enable or disable the WB gain correction function based on the green scene detection. In step 401, the green scene correction function is turned on / off. Judging. If the correction function is off, the green correction amount Grni, which is data indicating the likelihood of a green scene, is set to 1 (step 409).
[0033]
If the correction function is on, it is determined in step 402 whether or not a predetermined number or more of the 64 × 64 divided area hue data falls within the green detection frame 81. Therefore, the green correction amount Grni is set to 1 (step 409).
[0034]
When a predetermined number or more is included in the green detection frames 81, a regression line is obtained by the least square method based on the hue data included in the green detection frames among the hue data of 64 × 64 divided areas (step 403). Since the equation for obtaining the regression line by the method of least squares is well known, the description thereof is omitted. Assuming that the slope of the regression line obtained here is a and the intercept is b, the regression line is represented by (B / G) = a × (R / G) + b. Then, it is determined whether or not the obtained inclination a is equal to or greater than the threshold value S (step 404). If not, the green correction amount Grni is set to 1 (step 409). This judgment is influenced by the frame of the daylight light source (outdoor sunny) in the case of a green scene, and the inclination becomes positive, whereas in the case of an indoor light source scene, the distribution varies throughout. Is to be used.
[0035]
If the slope a of the regression line is greater than or equal to the threshold value S, the distribution of hue data is calculated in step 405. Specifically, an error between the regression line and the coordinate data of the hue data in the green detection frame 81 is obtained, and the number of hue data whose error is within a predetermined value is obtained. The method for obtaining the error will be described with reference to FIG. It is calculated whether or not the coordinates (gri, gbi) of each hue data in the green detection frame 81 satisfy the following expressions (1) and (2). Here, Yid and Xid are error thresholds.
[0036]
| Gbi− (a × gri + b) | <Yid (1)
| Gri- (gbi-b) / a | <Xid (2)
[0037]
For each hue data, the judgments of the formulas (1) and (2) are made, and the number of hue data satisfying both formulas is set as the number of green scenes. That is, since the green scene is affected by the daylight light source, it is determined that the error from the regression line 82 is small.
[0038]
Then, the green correction amount Grni is obtained from the number of hue data satisfying the equations (1) and (2) using the membership function shown in FIG. 6 (step 406). The membership function in FIG. 6 determines that if the number of hue data satisfying equations (1) and (2) is greater than GRN_G1, the rate of green scenes is determined to be high, and the green correction amount Grni is set to 1 or less, and GRN_G2 If it is more, it is determined that the scene is completely green and the green correction amount Grni is performed.
[0039]
Subsequently, in Step 407, it is determined whether or not a green scene determination condition is satisfied depending on the presence or absence of a light source detection frame on an extension line of the regression line. This determination is for preventing erroneous correction when the slope of the regression line is not influenced by daylight. That is, it is estimated that the hue data of the divided areas is included in the green detection frame 81, the slope of the regression line is positive, and the error from the regression line is small, but this is due to the effect of ordinary daylight. This is because the influence of the light source is also considered.
[0040]
This will be described with reference to FIG. If the regression line obtained by the method of least squares is represented by the reference numeral 71 in FIG. 7, it can be determined that the scene is a green scene because there is a daylight detection frame on the extension line of the regression line. . Therefore, it is determined that the green scene determination condition is satisfied, and the green correction amount Grni obtained in step 406 is output (step 408). However, when the regression line obtained by the method of least squares is represented by reference numeral 72 in FIG. 7, there is a detection frame of a warm white fluorescent light source on the extension line, which is considered to be an influence of the light source. Therefore, the green correction amount Grni is set to 1 (step 409). What kind of light source detection frame is on the extension line of the regression line can be easily determined from the slope and intercept of the regression line obtained by the least square method.
[0041]
【The invention's effect】
As described above, according to the present invention, it is possible to provide a digital camera capable of appropriately adjusting the white balance at the time of photographing a green scene such as a dark forest or a forest.
[Brief description of the drawings]
FIG. 1 is a diagram showing a schematic configuration of a digital camera showing an embodiment of the present invention. FIG. 2 is a schematic functional block diagram of a white balance gain calculation processing unit. FIG. 3 is a diagram showing various types on R / GB / G coordinates. FIG. 4 is a diagram showing a schematic flow of greenness determination processing. FIG. 5 is a diagram showing an error between the regression line and the coordinate data of the hue data in the green detection frame. FIG. 6 is a diagram illustrating an example of a membership function of green scene-likeness. FIG. 7 is a diagram illustrating green scene determination based on the presence or absence of a light source detection frame on an extension line of a regression line.
DESCRIPTION OF SYMBOLS 11 ... Shooting lens 12 ... Aperture 13 ... Solid-state image sensor (CCD)
14 ... CCD drive circuit 15 ... correlated double sampling circuit (CDS circuit)
16 ... A / D converter 17 ... timing generation circuit 18 ... memory 19 ... CPU
DESCRIPTION OF SYMBOLS 20 ... Digital signal processing circuit 21 ... Synchronization circuit 22 ... White balance adjustment circuit 23 ... Gamma correction circuit 24 ... YC signal creation circuit 25 ... Memory 31 ... Camera operation part 32 ... Lens drive unit 33 ... Aperture drive unit 34 ... Strobe 35 ... Integration circuit 40 ... WB gain calculation unit 41 ... Luminance / hue calculation unit 42 ... Outdoor / indoor judgment Unit 43 ... color temperature determination unit 44 ... WB gain selection unit 50 ... green data calculation unit 60 ... WB gain correction unit 81 ... green detection frame

Claims (5)

A digital camera having a white balance adjustment function according to a shooting scene,
Luminance detection means for detecting the luminance of the subject;
Hue detection means for detecting the hue of each area by dividing the photographed image of the subject into a plurality of areas;
White balance adjustment data calculating means for obtaining white balance adjustment data based on the luminance data and the hue data;
Green data calculating means for obtaining data indicating the green scene likeness of the subject based on the distribution of the hue data in R / GB / G coordinates;
White balance correction means for correcting the white balance adjustment data based on the data obtained by the green data calculation means;
White balance adjustment means for performing white balance adjustment based on the white balance adjustment data corrected by the white balance correction means,
The green data calculation means obtains a regression line based on the hue data existing in the green detection frame in the R / GB / G coordinates, and the slope of the regression line, and the regression line and the green detection frame A digital camera for obtaining data indicating the green scene-likeness based on an error from the coordinate data of the hue data. The digital camera according to claim 1,
The green data calculation means is a digital camera for obtaining data indicating the green scene likeness based on the number of the hue data whose error is within a predetermined range. The digital camera according to claim 1 or 2,
The green data calculation means is a digital unit that corrects data indicating the green scene-likeness depending on whether or not a detection frame of a predetermined light source in the R / GB / G coordinates exists on an extension of the regression line. camera. The digital camera according to any one of claims 1 to 3,
The white balance correction unit is a digital camera that performs correction using white balance adjustment data in the case of fine weather according to data indicating the green scene-likeness. The digital camera according to any one of claims 1 to 4,
The white balance adjustment data calculating means determines whether the shooting scene is an outdoor scene or an indoor scene based on the luminance data and the hue data, determines a color temperature of the subject, and based on the determined data A digital camera that seeks white balance adjustment data.

Images