JP3343921B2 - Person adaptive exposure control device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an exposure control device for photographing equipment such as a video camera.

[0002]

2. Description of the Related Art In a video camera or the like, the characteristics of how well a person can be photographed are required. One of the factors for successfully photographing a person is to photograph a person with an appropriate exposure. In this regard, the conventional exposure control apparatus merely detects the level of the video signal and controls the AGC circuit and the like so that the level becomes constant. Therefore, whether the main subject is a person or a landscape, the exposure control mode is the same.

[0003]

However, according to the research results of the television system, in the case of photographing a person, the exposure level of the face is 60 to 70 IR as a video signal level.
It is stated that a natural and beautiful expression of a person is possible on a CRT of a television receiver if adjusted to about E.
The prior art does not guarantee that the person's face is always automatically adjusted to this video signal level. Also, the exposure level of the main subject cannot be selectively and automatically adjusted between a person and a person other than the person. The present invention has been made in view of such a point, and an object of the present invention is to provide an exposure control device that can shoot with different exposures in a case of general landscape shooting and a case of a person shooting. .

[0004]

Means for Solving the Problems] exposure control device of the present invention for achieving the above object, the main object is shooting whether a person
A person determining means for determining based on chromaticity information indicating a skin color in the shadow video signal, and different exposures when the main subject is determined to be a person by the person determining means and when the main subject is determined to be other than a person And exposure control means for setting. At this time, the determination means uses the shape and area on the imaging plane of the chromaticity information obtained from the chromaticity information indicating the skin color in the video signal captured by the video capturing device and the photographic magnification of the photographic system to determine the face of the person. It is determined whether or not the person is a person by comparing with a parameter indicating. Further, when the person determining means determines that the main subject is a person, the exposure controlling means controls the exposure so that the luminance signal level of the face of the person is set to a value near 65IRE.

[0005]

According to this configuration, when photographing a person,
Exposure control (for example, face exposure control) suitable for portrait photography can be performed. Also, when photographing a person other than a person, exposure control suitable for the photographing (for example, normal exposure control based on the luminance signal level of a photographed image) can be performed.

Further, the photographing device is a video photographing device, and the judging means determines the shape and area of the chromaticity information on the imaging plane, which is obtained from the chromaticity information indicating the flesh color in the photographed image, and the photographing of the photographing system. The apparatus according to claim 1, wherein a determination is made as to whether the person is a person by comparing the parameter with a parameter indicating a person's face using a magnification. Since the skin color is detected based on the chromaticity signal, the color detection accuracy is increased, so that more accurate exposure control can be realized. It is conceivable to use a color difference signal as an alternative to the chromaticity signal. However, even if the color information of the subject does not change, if the exposure level changes, the signal level changes, so that the color detection accuracy is poor. On the other hand, the chromaticity signal indicates color information that is invariant to the exposure level, so that the color detection accuracy is high. Furthermore, by setting a person parameter in advance and comparing and referring to a skin color portion imaged using a photographing magnification, a person can be detected easily and with high accuracy.

[0007]

Embodiments of the present invention will be described below. FIG. 1 is a block circuit diagram of a video camera embodying the present invention, wherein 1 is an imaging lens unit, and 2 is an aperture. The imaging lens unit 1 converts the zoom information and the distance measurement information into a photographing magnification calculation unit 1.
Output to 3. 3 is a CCD (charge couple)
ed device), and outputs an R (red), G (green), and B (blue) signal component of a captured image. Reference numeral 4 denotes an exposure control unit that detects the average or peak level of the video signal obtained by the imaging unit 3 and controls the aperture so that the level becomes a constant value.

Reference numerals 5, 6, and 7 denote AGC circuits whose gains are controlled by the control processing unit 12 so as to start operation at low illuminance when the aperture 2 is open. 8, 9 is R
A variable gain amplifier for white balance inserted into each of a channel and a B channel, and a control process based on color temperature information of subject illumination light obtained by an external light type white balance sensor (WB sensor) unit 14 The gain is controlled via the unit 12 to achieve a predetermined white balance. In this case, the control processing unit 12 prepares control data corresponding to the color temperature information in advance, and outputs control data corresponding to the color temperature information provided from the white balance sensor unit 14 to output the variable gain amplifiers 8 and 9. Give to. The control data is D / A
It is assumed that the signal is converted and then guided to variable gain amplifiers 8 and 9.

Next, reference numeral 10 denotes a signal processing unit which performs gamma correction and matrix processing on the input R, G, and B signals to form a luminance signal Y, and forms and outputs a color signal CS. In this case, it is assumed that horizontal and vertical synchronization signals are added to the luminance signal Y. Also, the color signal CS
Is the color subcarrier (3.58 MHz for NTSC standard)
Is formed by performing balanced modulation with RY and BY signals and a color burst signal is added. The luminance signal Y output from the signal processing unit 10 is separately provided to a low-pass filter 30, where the signal is band-limited, and is taken into the control processing unit 12.

When detecting the low illuminance state based on the level of the input luminance signal, the control processing unit 12 controls the AGC circuits 5, 6, and 7 so that the level becomes a desired level. Note that this control is a normal exposure control referred to in this specification, and is distinguished from exposure control when the main subject is a person (described later). In the person automatic exposure mode, AGC circuits 5, 6, and 7 are controlled by a control signal from the control processing unit 12.
Performs a predetermined operation.

[0011] 11 is a white-balanced R,
This is a skin color detection unit that receives the G and B signals and generates skin color detection signals B and C, receives a clock CLK, and receives an instruction signal A from the control processing unit 12. The skin color detection signal C is always output, but the skin color detection signal B is output only when the instruction signal is input. Both of these skin color detection signals B and C are provided to the control processing unit 12. Reference numeral 15 denotes a person data section storing the person data, and when the request signal D is received from the control processing section 12, the person data is given to the control processing section 12.

Reference numeral 13 denotes a photographing magnification calculating unit for calculating a photographing magnification β, and a focal length f of the photographing lens unit 1 based on zoom information of the imaging lens unit 1
In addition to obtaining the value, the distance Ls from the subject to the front principal point of the imaging lens unit 1 is obtained from the distance measurement information, and the photographing magnification β is calculated from β = f / (Ls−f) from the f and Ls values, respectively. , To the control processing unit 12. Reference numeral 24 denotes a manual switch for setting a person automatic exposure mode.
2 is notified of the setting of the person automatic exposure mode.

The skin color detecting section 11 is configured as shown in FIG. Reference numeral 16 denotes a chromaticity information extracting unit which outputs chromaticity information of the captured video signal from each of the R, G, and B signals as quantized data.
The R, G, and B signals are output to the matrix unit 17 and mixed by the following equation. Assuming that the mixed output signal of the matrix unit 17 is M, M = (R + 1.8G + B) /3.8 (0 ≦ M ≦ 1: normalized signal). The R, M, and B signals are output to A / D (analog-to-digital conversion) sections 18, 19, and 20, respectively, and a clock signal CLK having a higher frequency than φH synchronized with the horizontal and vertical synchronization signals φH, φV (for example, , NTSC
(The standard color subcarrier signal).

The quantized R, M, B signals are divided by a divider 2
1 and 22, a division result R / M signal using the R signal as the numerator and the M signal as the denominator, and a division result B / M signal using the B signal as the numerator and the M signal as the denominator are obtained. A / D section 1
8, 19 and 20 are based on the sampling theorem, respectively.
It is assumed that the filter has band limiting means for preventing aliasing noise and an appropriate number of quantization bits. Division units 21 and 22
May be, for example, a look-up table method using a memory or a divider having a logic configuration.

The mixing ratio of the M signals (R: G: B = 1: 1)
1.8: 1) is U + indicating the sum of stimuli in the UVW color system.
The V + W value is expressed by using the R, G, and B stimulus values of the three primary colors of the NTSC standard image receiving device. To be exact, the following formula is used. However, since the division units 21 and 22 have the same configuration, they are used approximately. M = 0.261R + 0.469G + 0.270B (0 ≦ M ≦ 1: normalized signal)

Therefore, each of the R / M and B / M signals is
When an imaging system conforming to the C standard is used, the chromaticity information of the captured video signal is a uniform perceptual chromaticity diagram.
IE1960 UCS chromaticity diagram ((u, v) chromaticity diagram) or C
The IE1976 UCS chromaticity diagram ((u ′, v ′) chromaticity diagram) can be quantized uniformly along the (G, R) axis and the (G, B) axis. FIG. 3 shows this state (R / M, B
/ M The quantization bit number of each signal is 5 bits, and (u ′,
v ') A chromaticity diagram was used).

According to this quantization method, quantization noise for human chromaticity perception can be processed approximately uniformly within the range of chromaticity information extraction. Therefore, considering the minimum number of quantization bits, the flesh color of a person There is an advantage that the difference in chromaticity and the chromaticity other than the skin color can be subjected to data processing corresponding to the chromaticity perception.

The R / M and B / M signals of the chromaticity information extracting section 16 are sent to a skin color comparing section 23. The skin color comparison unit 23 has a skin color chromaticity data group prepared in advance, and can detect the presence or absence of a skin color in the captured video signal by comparing the input chromaticity information R / M and B / M signals. . The detection results are output as skin color detection signals B and C. If a flesh color is detected, a flesh color detection signal C is sent to the control processing unit 12 immediately.

The control processing section 12 has a memory section,
For example, in a field memory configuration, after outputting the instruction signal A to the flesh color comparing unit 23, the flesh color detecting signal B is fetched from the flesh color comparing unit 23 at every cycle of the clock signal CLK, and the imaging screen is displayed by the horizontal and vertical synchronizing signals φH and φV. Pattern indicating the position of the skin color chromaticity corresponding to the image (skin color pattern)
For one field. For example, as shown in FIG. 4, if (a) is a captured human subject, (b) is a skin color pattern stored in the memory unit by the above-described operation. However, the pattern is represented as “hatched” pixels when there is a skin color, and as “white” when there is no skin color.

The operation of the memory unit for obtaining the flesh color detection signal B is as follows.
At a certain timing, the control processing unit 12 checks the skin color detection signal C, and outputs the instruction signal A to the skin color comparing unit 23 by the control processing unit 12 only when there is a skin color. Next, the flesh color comparing section 23 outputs a flesh color detection signal B corresponding to the next field period from the time when the instruction signal A is received to the control processing section 12, and the memory section of the control processing section 12 stores this for one field. I do.

The person data section 15 previously holds n sets of data (referred to as person data H) corresponding to the area S of the face and its circularity K as parameters indicating the characteristics of the face of the person. The control processing unit 12 includes a person data unit 15
To output an instruction signal D. The person data unit 15 outputs the person data H to the control processing unit 12 in order or in accordance with the signal D.

Here, the circularity K will be described. The degree of circularity is a scale factor representing the complexity of the shape of a figure, and is defined in the present invention as a coefficient K obtained by the following equation when the area of a certain figure is S and its peripheral length is L. K = L ² / S (K ≧ 4π) For example, the value is 4π (minimum value) in the case of a circle, and 16 in the case of a square. The value increases as the shape becomes more complicated. Since the face part (skin color part) of a person is circular and simple in shape, a relatively small circularity K can be obtained. Therefore, it is possible to discriminate from an object other than a person having skin color chromaticity by this value. .

The control processing unit 12 determines whether or not the skin color area in the skin color pattern is a person, based on the skin color pattern obtained from the skin color detection signal B for one field, the photographing magnification β, and the person data H. If it is determined that the person, the control processing unit 12
Controls the AGC circuits 5, 6, and 7 so that the average level of the luminance signal Y in the area becomes 65 IRE. If it is determined that the person is not a person (landscape), the control processing unit 12 continues controlling the AGC circuits 5, 6, and 7 as is, that is, does nothing newly.

Hereinafter, the control operation by the control processing unit 12 in the present embodiment will be described. First, the procedure for determining whether or not the skin color area portion in the skin color pattern is a person will be described with reference to the flowcharts of FIGS. It is assumed that a plurality of flesh-colored areas exist in the flesh-colored pattern, and it is checked whether or not there is a flesh-colored area corresponding to the face. When this flow is started, first, in step # 5, the switch 24
Is turned on or off to determine whether or not the person automatic exposure mode is set. If the switch 24 is OFF, the process proceeds to step # 55 to perform the exposure processing in the normal automatic exposure mode. If the switch 24 is ON, the exposure processing in the person automatic exposure mode after step # 10 is performed.

In step # 10, the presence or absence of a flesh color is determined by examining the flesh color detection signal C of the flesh color comparing section 23. If there is no flesh color, the process proceeds to step # 55 to perform the normal exposure mode. If there is a skin color, in step # 15 the photographing magnification calculating unit 1
3, the current photographing magnification β is taken. Then, step #
20 and the previously acquired photographing magnification β becomes a predetermined value β.
It is checked whether it is within the range of 1 and β2 (β1 <β2).
Here, β1 is a threshold constant for assuming that a landscape is being photographed even if the main subject is a person if the size of the face of the person is smaller than a certain size in the imaging size of the photographing element. Is a threshold constant at which the size of a person's face begins to grow unnaturally in the shooting size.

If the photographing magnification β is outside the range of β1 to β2, it is determined that the main subject is a landscape, and the flow proceeds to step # 55. If β is in the range of β1 to β2, step # 25
Then, the instruction signal A is supplied to the skin color comparing section 23 to obtain the skin color detection signal B. Then, step # 30
The skin color detection signal B from the skin color comparison unit 23 to the control processing unit 12
Into a memory (field memory) in the device to obtain a flesh color pattern.

Next, in step # 35, contraction / expansion image processing is performed on the flesh color pattern. Shrinking / dilating image processing is described in the document "Image Processing System and How to Use It", Yoji Marutani
A technique described in the Nikkan Kogyo Shimbun, which refers to a process of converting a projecting shape portion or an acute angle portion existing in an area showing a flesh color in a flesh color pattern into a relatively smooth contour portion. For example, if skin color is extracted in a state where hair is applied to the forehead of the face, a projection-shaped portion corresponding to the hair portion is generated in the skin-colored area. Convert the part to a relatively simple shape,
Improve face detection accuracy. This process is repeated as appropriate to obtain an optimal shape. In addition, since this process eliminates other skin color area portions that are very small with respect to the skin color area portion, there is also an effect that the number of times of comparison with the person data H performed in post-processing is reduced.

After the processing in step # 35 is completed, the process proceeds to step # 40, in which the skin color not yet subjected to the labeling image processing has been given priority from the center of the screen among the plurality of skin color areas in the skin color pattern. Select one area part. The reason is that when the person is the main subject, the probability that the face is located near the center of the screen is high. In the next step # 45, it is determined whether or not there is a skin color area selected in the processing in step # 35. If the determination result is present, the process proceeds to step # 50. If not, the main subject is determined to be other than a person (landscape), and the process proceeds to step # 55.

In step # 50, labeling image processing is performed on the skin color area selected in step # 40. Labeling image processing refers to distinguishing a plurality of connected pixel groups by attaching the same label symbol to connected pixels and attaching different label symbols to different connected pixel groups. At the same time, the number N of pixels constituting the skin color area is counted. The number N of pixels is a value corresponding to the area of the skin color area.

Next, the process proceeds to step # 55 of FIG. 6, and in the step # 55, a variable for selecting a specific one of the n sets of person data H is set to i, and the first is selected (denoted by Hi).
Perform processing. Then, in the next step # 60, an instruction signal D including the variable i = 1 is output to the person data section 15, and in step # 65, the person data section 15 outputs the person data Hi (Si, Ki) corresponding to the variable i = 1. Receive).

Thereafter, at step # 70, N / (Si
• β ² ) and N / (Si · β ² )
It is determined whether the ratio falls within a predetermined ratio γ1 to γ2. Here, N is the number of pixels, and Si is the person data. This determination in step # 70 is based on the image sensor (CCD) obtained by multiplying the number of pixels N (corresponding to the skin color area) of the skin color area and the area magnification β ² at the current photographing magnification β by the area Si of the human face. By comparing the face area Si · β ² in the above, it is determined whether the skin color area is a person or a person other than a person (landscape).

In this determination, N / (Si · β ² ) is γ1 to γ
If not, the person data variable i is incremented by 1 in step # 85, and the process proceeds to step # 90.
If the person data variable i is within the number n of person data sets, the process returns to step # 60. If the number of persons has exceeded n, the process returns to step # 40. If N / (Si · β ² ) is in the range of γ1 to γ2 in the determination of step # 70, the process proceeds to step # 75, where the perimeter L of the flesh-colored area labeled in the processing of step # 50 is calculated. Ask. The perimeter L is obtained by counting the pixels constituting the contour of the skin color area, and obtaining data corresponding thereto. However, since the meaning of the pixel differs depending on the manner in which the outline pixels are arranged, correction is performed in accordance with the manner.

Next, in step # 80, an operation of | L ² / N−Ki | ≦ ε1 is performed on the number of pixels N, the person data Ki, and the perimeter L, and whether or not it is within a predetermined error ε1 is determined. judge. In this determination, the circularity L ² / N of the area is compared with the circularity Ki of the person's face to determine whether the area is a person or other than a person (landscape). N
If O (does not fit), the process proceeds to step # 85, and if YES (fits), step # 10 in FIG.
Go to 0. In the case of YES here, it is determined that the main subject is a person.

The person detection is performed according to the above procedure. Next, an exposure control method after the main subject is determined to be a person will be described with reference to FIG. First, in step # 100, an instruction signal A for obtaining a skin color detection signal B is output to the skin color comparing section 23, and the process proceeds to the next step # 105.

In step # 105, a skin color detection signal B is obtained from the skin color comparing section 23, and a logical product is obtained from the skin color area determined as a human face in the previous processing, and the number of pixels Na as a result is counted. Since this process requires some time in the process of determining whether or not the previous skin color area portion is a person, a new latest skin color detection signal B is newly obtained, and A
This is because the ND signal is regarded as the current skin color area. When the process of step # 105 is completed, the process proceeds to step # 110, where the ratio Na / N of the number of pixels N of the skin color area portion in the process of step # 50 and the number of pixels Na in the process of step # 105 is determined. Is compared with a predetermined value γ3.

If Na / N is smaller than γ3, it is determined that there is a change in the main subject or the photographing situation, and the process returns to step # 5. If Na / N is γ3 or more, step # 11
5 and the luminance signal level in the skin color area is 65IR
Nb is obtained by counting pixels above E. This point will be described with reference to FIG. 8. The luminance signal Y output to the control processing unit 12 is output to the comparison unit 26 only during a period corresponding to the gate signal by the gate unit 24. The gate signal is supplied to the AND circuit 25.
And an AND signal obtained by performing a logical product of the signal of the skin color area determined as the person's face and the skin color detection signal B. That is, the signal indicates the current skin color area. The gate signal is ANDed with the clock CLK by the AND circuit 28 to become a pixel signal pulse PNa of the skin color area.

On the other hand, the comparison section 26 compares the signal level of the input luminance signal Y with the reference luminance signal level, and outputs a comparison output signal to the AND circuit 27 when the signal level of Y is equal to or higher than the reference luminance signal level. . The AND circuit 27 performs an AND operation on the output signal of the comparison unit 26 and the clock signal CLK.
, A pixel signal pulse PNb having a Y signal level equal to or higher than the reference luminance signal level is obtained. Regarding the reference luminance signal level, according to the research results of the television system, the reproduction luminance level of the human skin color is about 60 to 70 IRE.
Is considered to be optimal, and here, it is assumed to be 65IRE.

Reference numeral 29 denotes a counting unit for counting the signal pulses PNa and PNb from the AND circuits 28 and 27, and outputs the count values Na and Nb of the signal pulses for one field, respectively. Now, returning to FIG. 7, when the processing in step # 115 is completed, in step # 120, step # 10
5 and the number of pixels Nb having a luminance signal level of 65 IRE or more obtained in the processing of # 115 and # 115, the value of Nb−Na / 2 is determined to determine the magnitude relationship between 0 and Nb−Na / 2.

The meaning of this determination is the number N of pixels of the skin color area portion.
It is intended to set the average luminance signal level of the flesh-colored area to 65IRE by checking whether the luminance signal level indicated by a exists half by half with respect to 65IRE. Then, according to the comparison result between the above equation and 0, A
The gains of the GC circuits 5, 6, 7 are increased or decreased by a predetermined unit level. That is, if the above equation (Nb-Na / 2) is larger than 0, the process proceeds to step # 125, and the AGC circuit 5,
Lower the gains of units 6 and 7 by unit level, and then step # 1
It returns to the process of 00. If the above expression is smaller than 0, the process proceeds to step # 130 to increase the gain of the AGC circuits 5, 6, and 7 by a unit level, and then returns to the process of step # 100. If 0, the flow ends.

Next, another embodiment of the exposure control method is shown in FIG.
This will be described with reference to FIG. 9 and 10, when the main subject is determined to be a person, the degree of the normal exposure control and the degree of the person automatic exposure control are determined by the relative ratio of the size of the person to the shooting angle of view. Weighting is applied to this. As the ratio of the person increases, the normal automatic exposure control continuously shifts to the person automatic exposure control.

First, in step # 200, the comparison unit 26 sets the initial reference luminance signal level Yref as 65IR, for example.
Set E. Then, steps # 205 and # 21
0 and # 215 are steps # 100 and # 10 of the previous embodiment.
5. After performing the same processing as in # 110, in step # 220, Nb is obtained by counting the pixels in which the luminance signal level in the skin color area is equal to or higher than Yref.

Next, in step # 225, the value of (Nb−Na / 2) is obtained for the number of pixels Na of the skin color area portion and the number of pixels Nb of which the luminance signal level is equal to or more than Yref, and is compared with 0. If Nb-Na / 2> 0, step # 230
Then, the reference luminance signal level is raised by a predetermined unit level from the Yref value. If Nb-Na / 2 <0, the reference luminance signal level is lowered by a predetermined unit level in step # 235, and the process returns to step # 205. When Nb-Na / 2 = 0,
Proceed to step # 240 in FIG. By the loop processing up to this point, the average luminance signal level Yref of the current flesh color area by the normal automatic exposure control is determined.

In the following processing, a target set value of the luminance signal level of the skin color area is obtained. The set value is the Yref value and the reproduction luminance signal level 65IRE of the human skin color.
Is a composite value obtained by performing weighting with the photographing magnification β value. First, in steps # 240, # 245, and # 250, the same processing as in steps # 100, # 105, and # 110 is performed, and in step # 255, the current photographing magnification β is obtained from the photographing magnification calculator 13 to obtain a weight coefficient. Ask for WE. The weighting coefficient WE is determined by the following formula, for example, from the thresholds β1 and β2 of the photographing magnification β used in the process d.

WE = (β−β1) / (β2−β1) or WE = {β ² − (β 1) ² } / {(β 2) ² − (β 1) ² }. However, WE = 0 when β <β1, and WE = 1 when β> β2. Therefore, the value range of the weight coefficient WE is 0
≤WE≤1.

Subsequently, in the next step # 260, using the weighting coefficient WE obtained in the processing of the step # 255, the average luminance level Yref of the current skin color area and the reproduced luminance signal level 65IRE of the human skin color are obtained. A weighted composite value, that is, a target reference luminance signal level is obtained by the following equation. (65−Yref) · WE + Yref Then, the value of the above equation is set as a new Yref value.

Hereinafter, in steps # 265 to # 280, the above-described steps # 220, # 120, and # 12 are respectively described.
5, the same processing as in # 130 is performed, and steps # 275 and # 130 are performed.
When proceeding to 280, after performing those processes, the process returns to step # 240.

[0047]

As described above, according to the present invention, when photographing a person, exposure control suitable for photographing a person can be performed. Also, when photographing a person other than a person, exposure control suitable for the photographing can be performed. In particular, in the present invention
The person is determined based on the chromaticity information indicating the skin color in the captured video signal.
Is determined based on the color difference signal.
However, the accuracy of color detection (accordingly, the accuracy of determination) is high. In addition, the determination means compares the shape and area on the imaging plane of the chromaticity information obtained from the chromaticity information indicating the skin color in the captured image with a parameter indicating the face of a person using the imaging magnification of the imaging system. Is performed, a chromaticity signal indicating color information that is invariant with respect to the exposure level is created, and a skin color is detected based on the chromaticity signal. Since the detection accuracy is increased, more accurate exposure control can be realized. Furthermore, by setting a person parameter and comparing and referring to a skin color part imaged using the photographing magnification, a simple and highly accurate person detection can be performed.

[Brief description of the drawings]

FIG. 1 is a main part block circuit diagram of a video camera embodying the present invention.

FIG. 2 is a block circuit diagram showing a configuration of the skin color detection unit.

FIG. 3 is a CIE1976 UCS chromaticity diagram.

FIG. 4 is a view showing a human subject and a stored skin color pattern;

FIG. 5 is a flowchart illustrating an operation flow of person detection according to the embodiment;

FIG. 6 is a flowchart following FIG. 5;

FIG. 7 is a flowchart of exposure control according to the embodiment.

FIG. 8 is an explanatory diagram thereof.

FIG. 9 is a flowchart illustrating exposure control according to another embodiment of the present invention.

FIG. 10 is a flowchart following FIG. 9;

[Explanation of symbols]

 5, 6, 7 AGC circuit 10 Signal processing unit 11 Skin color detection unit 12 Control processing unit 13 Shooting magnification calculation unit 15 Person data unit 16 Chromaticity information extraction unit 23 Skin color comparison unit

────────────────────────────────────────────────── ─── Continuing on the front page (72) Keizo Ochi 2-3-13 Azuchicho, Chuo-ku, Osaka-shi Osaka International Building Minolta Camera Co., Ltd. (72) Yoshihiko Higashi 2-chome Azuchicho, Chuo-ku, Osaka-shi No. 3-13 Osaka International Building Minolta Camera Co., Ltd. (72) Inventor Takehiro Kato 2-3-1 Azuchicho, Chuo-ku, Osaka City Osaka International Building Minolta Camera Co., Ltd. (72) Inventor Hiroshi Otsuka Osaka 2-3-3 Azuchicho, Chuo-ku Osaka International Building Minolta Camera Co., Ltd. (56) References JP-A-3-148987 (JP, A) JP-A-2-287225 (JP, A) JP-A-1 −114188 (JP, A) (58) Field surveyed (Int. Cl. ⁷ , DB name) H04N 5/235

Claims (4)

(57) [Claims] 1. An exposure control device for a photographing device, comprising:
Color indicating the flesh color in the captured video signal to the main subject is whether a person
A person determination unit that determines based on degree information, and an exposure control unit that performs different exposure settings when the main subject is determined to be a person and when the main subject is determined to be other than a person by the person determination unit. Exposure control device. 2. An exposure control device for a photographing device, comprising:
A person determining means for determining whether or not the main subject is a person;
The main subject is determined to be a person by the person determination means
The exposure setting differs depending on when
Exposure control means for performing exposure control
The photographing device is a video photographing device, and the determination unit determines the shape and area on the imaging plane of the chromaticity information obtained from chromaticity information indicating skin color in the captured video signal and the photographic magnification of the photographic system. An exposure control device that determines whether a person is a person by comparing the parameter with a parameter indicating a person's face. 3. The exposure control means according to claim 1, wherein when the person determination means determines that the main subject is a person, the exposure control means sets the luminance signal level of the face of the person to a value near 65IRE. 3. The exposure control device according to claim 2, wherein the exposure control is performed. 4. The exposure control apparatus according to claim 3, wherein the size of the area of the face of the person occupying in the screen is weighted by the photographing magnification.