JPH0537940A - Video camera - Google Patents

Abstract

PURPOSE:To obtain a focus appropriate for a person by detecting the skin- colored part of the person and finding a range. CONSTITUTION:A chrominance signal inputted from an image is converted into digital signals through A/D converters 31 to 33, the digital signals are inputted to a memory 34, a read skin-colored area signal is detected by a human detecting microcomputer 35, and an exposure level signal is formed by a screen dividing circuit 41 and sent to a control microcomputer 21. On the other hand, the microcomputer 35 executes individual extracting processing and judges a human based upon an extracted outline in accordance with the data read out from the memory 34. The microcomputer 21 controls a lens driving part 22 so that the area of the skin-colored part in the judged human part is fixed. Thereby the human can be focused independently of his position and human's size can be constantly controlled.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a video camera.

[0002]

2. Description of the Related Art In a conventional video camera, focus adjustment, exposure control, color adjustment and the like are usually performed in a fixed distance measuring area, photometric area and color measuring area. However,
Especially when a person is mainly used for photographing, it is natural to want to adjust the focus, exposure, color reproduction and the like of the person of interest. For that purpose, the person should be tracked regardless of the position on the screen so as to obtain proper focus, exposure, color reproduction, etc., centering on the skin color portion of the person.

However, there is no conventional example in which such a consideration is taken into consideration. For example, Japanese Patent Laid-Open No. 1-12018
In Japanese Patent Laid-Open No. 1-120178 and Japanese Patent Laid-Open No. 1-120178, the position and size of a subject are determined by using the difference information of frequency components obtained from inside and outside the detection area, and the position of the photometric area is controlled based on the determination result. However, neither extracting the skin color part of a person nor tracking it.

[0004]

Therefore, in the conventional video camera, ideal focus, exposure, color reproduction, etc. could not be realized in the case of photographing mainly by a person.

The present invention has been made in view of the above points, and provides an appropriate focus centering on a person's skin color portion.
It is an object of the present invention to provide a video camera capable of exposure, color reproduction, etc.

[0006]

In order to achieve the above object, the video camera of the present invention has a skin color part extraction circuit for extracting a skin color part from a video signal and a skin color part extracted by the skin color part extraction circuit. And a distance measuring area forming means for forming a distance measuring area slightly wider than the flesh-colored portion with respect to the flesh-colored portion determined to be the human portion.

In this case, when the skin color portion is detected in the vicinity of the previously detected skin color portion position, the distance measuring area forming means forms the distance measuring area without performing the determination by the determining means.

The video camera of the present invention comprises a skin color part extraction circuit for extracting a skin color part from a video signal, a determination means for determining whether or not the skin color part extracted by the skin color part extraction circuit can be identified as a human part, and And a means for driving the zoom lens so that the area of the flesh-colored portion determined to be the human portion becomes constant.

[0009]

As described above, in the video camera of the present invention, since the skin-colored portion of the person is detected and distance measurement is performed, the person can be focused regardless of the position on the screen of the person.
Further, since the skin color portion of the person is controlled to have a constant area regardless of the position on the screen, it is easy to control the size of the person to be constant.

[0010]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS FIG. 1 shows a known circuit of a video camera to which the embodiments of the present invention shown in FIG. 5 and subsequent figures are applied. Since the circuit itself of FIG. A general circuit of the video camera shown in FIG. 1 will be described. In the figure, 1 is a taking lens and 2 is a diaphragm. Here, the taking lens 1 is illustrated as a single lens,
Actually, it includes a focusing lens and a zoom lens. Reference numeral 3 denotes a color separation optical system that decomposes natural light into red (R), green (G), and blue (B) light and outputs the light.
For example, it is composed of a color transmission prism.

CCDs (charge coupled devices) 4, 5 and 6 which act on the R, G and B lights supplied from the color separation optical system 3, respectively, correspond to the R, G and B light signals. It outputs electrical signals R1, G1, B1. This CCD output signal R
1, G1, and B1 are added by the adding circuit 15 at an appropriate ratio, and then detected by the first detecting circuit 16. The detected output voltage is given to the diaphragm drive unit 17 to control the diaphragm 2. This control is, for example, when the subject is bright, the CCD output signal R
1, G1 and B1 increase, the detection output voltage rises, and the diaphragm 2 is narrowed down. This is performed by a feedback loop, and the diaphragm is controlled so that the output voltage of the detection circuit 16 is always constant.

Reference numerals 7, 8 and 9 denote AGC circuits, which are controlled by the level-detected outputs of the luminance signal Y output from the matrix circuit 12 in the subsequent stage. 10,
Reference numeral 11 is a voltage control amplifier for white balance inserted in the R channel and B channel. These amplifiers 10 and 11 process the outputs R ′, G ′ and B ′ of the white balance sensor 19 by the processing circuit 20. The gains are respectively controlled by the R ′ / G ′ and B ′ / G ′ signals thus obtained. For example, when the color temperature is low, the output of the white balance sensor 19 is R ′> when capturing a white subject.
Since G '>B', the two outputs of the processing circuit 20 are R '/ G'> 1 and B '/ G'<1, respectively, and depending on this ratio R '/ G', B '/ G'. The R-channel amplifier 10 lowers the gain, while the B-channel amplifier 11 increases the gain so that R2 = G2 = B2.

Reference numeral 12 is a matrix circuit, which is a primary color signal R.
2, G2, B2 to a luminance signal Y and two color difference signals R-
Y and BY are formed. The video signal processing circuit 13 in the next stage adds a synchronizing signal to the luminance signal Y, orthogonally modulates the R-Y and B-Y signals, adds a color burst and outputs as a chroma signal C, each of which is a deck section. Supply to 14. Reference numeral 21 denotes a control microcomputer (hereinafter referred to as “control microcomputer”), which controls various controls. One of them is the lens 1 via the lens drive unit 22.
Has a function of driving for focusing or the like. In addition, information such as the focal length of the lens and the lens extension amount is input from the lens driving unit 22.

By the way, in a video camera, an infrared AF method or a phase difference detection AF method has been conventionally used as AF (autofocus), but recently, a video signal using a video signal which is advantageous in terms of downsizing. The AF method has become mainstream, and this method is also used in FIG. In FIG. 1, the high frequency component of the luminance signal Y after matrix is separated by the bandpass filter 23, and the signal level is A
The A / D converter 24 performs A / D conversion. The screen division circuit 25 is controlled by the AF area signal from the control microcomputer 21, the central portion of the screen 28 shown in FIG. 4 is set as the AF area 29, and the data in this area 29 is integrated by the digital integration circuit 26.

Therefore, as the output data of the digital integrator circuit 26, a large value is obtained at the time of focusing to obtain a high AF evaluation value, and a small value is obtained at the time of non-focusing to obtain a low evaluation value. This data is integrated and read out for each field under the control of the control microcomputer 21.

FIG. 2 shows a flow chart (AF routine) of the AF control operation by the control microcomputer 21.
First, when this routine starts, in step # 5, the lens 1 (focusing lens in this case) is driven after waiting until the AF evaluation value based on the output data of the digital integration circuit 26 stabilizes (step # 10). Then, it is determined whether or not the AF evaluation value is increased by driving the lens (step # 15). If the AF evaluation value has increased, the process proceeds to step # 25 until the AF evaluation value decreases.
Move the lens as it is. Then, when the AF evaluation value starts to decrease, the focus is returned to the in-focus point in step # 30, and step #
At 35, the lens drive is stopped.

This situation will be described with reference to FIG. 3. For example, when the lens is driven in the infinity direction when the lens position before driving is P1, the AF evaluation value gradually increases,
It decreases after passing the focal point P2. Therefore, at P3, the lens is moved in the opposite direction (in this case, the near side direction), and the focal point P
Return to 2.

When the AF evaluation value does not increase in the determination of step # 15 in FIG. 2 above, it is considered that there is a focal point in the opposite direction, and the process proceeds to step # 20 to move the lens in the opposite direction (near side direction). Driving is performed, and this is performed until the AF evaluation value decreases in step # 25. After that, the lens drive is stopped after returning to the in-focus point (steps # 30, # 35). This situation will be described with reference to FIG. 3. When the lens before driving is at P4, for example, if this lens is driven in the infinity direction, the AF evaluation value does not increase. Therefore, driving is performed in the opposite direction at P5 on the way, and this is continued until the evaluation value decreases. In this case, the evaluation value decreases when the focal point P2 is exceeded and the evaluation value is further moved to the near side. Therefore, the lens is moved again in the opposite direction (infinity direction) and the focus P2
Then, the lens drive is stopped.

In the above description, the case where the lens position before driving is first moved to the infinity direction (that is, in step # 10) has been described, but the same applies when the driving in step # 10 is performed in the near side direction. And the lens is brought to the in-focus position P2. Now, in step # 40, it is determined whether or not there is a change in the subject from the state of being in focus, and if there is a change, the process returns to step # 5, and from the beginning A
Repeat F action. If there is no change, the flow ends.

Returning to FIG. 1, 27 is a trigger switch, and when this switch is turned on, the control microcomputer 21 detects this and changes the state of the deck section 14. That is, if the deck unit 14 is in the recording state before the switch 27 is turned on, the recording is stopped, and if not, the recording is started.

Above, the explanation of FIG. 1 is finished, and then the first embodiment of the present invention shown in FIG. 5 will be explained. In the circuit of FIG. 5 described below, the description of the parts denoted by the same reference numerals as those of FIG. 1 will be omitted in principle. In FIG. 5, reference numeral 30 is a microcomputer 2 for controlling the primary color signals R2, G2 and B2 which are AGC-controlled and whose white balance has been adjusted.
The color reproduction correction circuit performs color reproduction correction based on the color reproduction correction data from 1. The color reproduction correction circuit 30 is a circuit provided for adjusting a color to a user's preference by operating a color correction switch 50 described later, and its insertion position acts on R2, G2, B2 shown in the drawing. Instead of the position to be set, RY, B-
It may be a position that acts on Y.

Reference numerals 31, 32 and 33 denote the primary color signals R2,
It is an A / D converter that converts G2 and B2 into digital values, and the converted outputs are taken into the memory 34 as r, g, and b signals. The r, g, and b signals read out from the memory 34 after being taken into the memory 34 are processed by a person detecting microcomputer (hereinafter referred to as “person detecting microcomputer”) 35. The memory 34 has an image memory for each color for two fields, and reading and writing are repeated for each field. The person detection microcomputer 35 applies the first AF area signal to the first screen division circuit 25, the second AF area signal to the second screen division circuit 38, and the skin color area signal to the third screen division circuit 41.

Here, the second screen division circuit 38 passes the bandpass filter 36 and the A / D converter 37 to the digitized high frequency component of the luminance signal to detect the person 35.
It is gated on the basis of the second AF area signal from and is given to the digital integration circuit 39. The output of the digital integration circuit 39 indicates the AF evaluation value of the second AF area and is sent to the control microcomputer 21 as an AF signal. Bandpass filter 3
6, the A / D converter 37, the screen division circuit 38, and the digital integration circuit 39 are the bandpass filter 2 described in FIG.
3, the A / D converter 24, the screen division circuit 25, and the digital integration circuit 26 are related to the signals in the first AF area, but are different in that they are related to the second AF area. Is 23, 24, 25, 2
Same as 6.

The third screen division circuit 41 controlled by the skin color area signal from the person detection microcomputer 35 gates the low frequency component of the luminance signal Y extracted by the low pass filter 40. The gate output is averaged by the averaging circuit 42, digitized by the A / D converter 43, and sent to the control microcomputer 21 as an exposure level signal. From the diaphragm 2, diaphragm opening information is sent to the control microcomputer 21. On the contrary, from the control microcomputer 21, an aperture correction signal is given to the aperture drive circuit 17 via the D / A conversion circuit 47 and the addition circuit 48. For the AGC circuits 7, 8 and 9, the control microcomputer 21 passes the D / A converter 45 and the adder circuit 46 to the AG.
A C correction signal is provided.

Further, FIG. 5 shows the auto trigger switch 4
4, an APZ switch 49, and a color correction switch 50 are provided. Among them, the auto trigger switch 44 controls ON / OFF of the auto trigger, and when it is ON, the deck is started by the start / stop signal from the control microcomputer 21. ON / OFF of is automatically controlled. Also, A
The PZ switch 49 controls ON / OFF of the auto zoom. When the PZ switch 49 is ON, a zoom lens drive signal is output from the control microcomputer 21 to drive the lens 1 (zoom lens in this case). With this, the angle of view is automatically adjusted. The color correction switch 50 turns the color correction ON / OFF, and when it is ON, the color reproduction correction data is given to the color reproduction correction circuit 30 by the control microcomputer 21.

As compared with FIG. 1, one of the features of FIG. 5 is a person detection function. Image data for human detection is obtained by A / D conversion of the primary color signals R2, G2, B2 after white balance adjustment.
The digital data A / D converted by the converters 31, 32 and 33 is used. This digital data is written and read by the memory 34 for each field. Based on the data read from the memory 34, the person detection microcomputer 35 executes the individual extraction process shown in FIGS. 6 and 7.

When entering the routine for extracting an individual, the person detecting microcomputer 35 determines the presence or absence of the previous individual in step # 100. Since there is no previous individual in the first execution, the process proceeds to step # 105 to calculate the total number of pixels of the skin color part in the screen. This calculation is performed by the pixel data rij, gi of the memory 34.
j and bij are sequentially loaded into the person detection microcomputer 35, and r
The ratio of (i, j), g (i, j), b (i, j) is calculated, and as a result, the pixel determined to be the skin color part is f (i, j).
= 1 and f (i, j) = 0 except for the skin color, and counting the total number of pixels which are the skin color part.

Next, in step # 110, it is determined whether or not the total number of pixels of the flesh color portion is a certain value or more. If it is more than the certain value, the contour extraction of the flesh color portion is performed in step # 115, and step # 120. Calculate the straightness of the contour with. The contour extraction of the skin color part is performed by calculating the contour of the skin color part as an example with the following algorithm.

[0029]   f (i, j) -f (i-1, j) = δ1 When δ1 = 1, f (i, j) = 1                                         When δ1 ≠ 1, f (i, j) = 0   f (i, j) -f (i, j-1) = δ2 When δ2 = 1, f (i, j) = 1                                         When δ2 ≠ 1, f (i, j) = 0

Further, in calculating the straightness of the contour portion, if the contour portion has a large number of straight line portions, it can be considered to be a portion having a color close to skin color such as a wall or a floor. Therefore, it is calculated whether the contour portion has a large number of straight line portions. FIG. 8 is a diagram in which the pixels of the extracted contour portion are represented by black circles 51. Each grid represents a coordinate. Here, the direction of the next contour pixel is stored counterclockwise from the pixel marked with the circle mark 52 at the upper left of the screen according to the direction code of FIG. In this example, the code is stored as shown in FIG. The total number NB of changed pixels is calculated by setting the pixel where the code has changed to 1 and the pixel which has not changed to 0. The part where the code changes is the curved part, so the larger this number is, the closer the curve is.

In step # 125, it is determined whether or not there is an object close to a circle. This is based on the logic that an object close to a circle may be considered as the outline of a person. In this determination, the ratio NB / NA between the total number of pixels NA in the contour portion and the total number NB of changed pixels is calculated, and if the ratio is equal to or greater than a certain value, it is determined that there is something close to a circle (human outline). .

If it is determined in step # 125 that there is something close to a circle, the next step # 130
Calculate the skin color area in the contour near the circle with. This can be easily calculated by counting the number of pixels in the contour portion. next,
In step # 135, it is determined whether or not there is a skin-colored part having a certain size or more. Here, in the case of NO (that is, when it is not above a certain level), it means that the person is at a distant position or noise. If YES (that is, above a certain level),
In the next step # 140, it is determined whether or not there are a plurality of skin-colored portions having a certain area or more. If the number is one, it means that the individual has been determined, so the person detection microcomputer 35 sets the first AF area signal including the skin color portion in step # 145.
The first AF area signal is given to the first screen division circuit 25. Along with this, the above-described AF routine of FIG. 2 is executed in step # 150, and the AF evaluation value in the area is controlled to be maximum. The control of the AF routine is performed by the control microcomputer 21.

If there are a plurality of flesh-colored areas having a predetermined value or more in step # 140, an AF area signal is set for each and the AF routine shown in FIG. 11 is executed (however, in FIG. 11, the flesh-colored areas are If there are two, namely 1A
The AF routine when the F area signal and the second AF area signal are set is shown). FIG. 12 shows the relationship between the lens position and the AF evaluation value when two people are at different distances. In this case, in the processing of the AF routine, the closer one of the two persons is focused.

Now, the AF routine of FIG. 11 will be described. When this routine starts, first, in step # 300, the AF evaluation value is allowed to stabilize, and the lens 1 is not moved until then. This is to prevent the subject from moving and erroneous distance measurement. When the evaluation value is stable, the flag FLG is set in step # 305 (F
LG = 1). This flag FLG is a flag indicating that the lens is moved in the near direction. In accordance with the setting of this flag, the focus lens is driven in the near direction in step # 310, and in the next step # 315, A for the first AF area is set.
It is determined whether the F evaluation value (1) has increased.

If the AF evaluation value (1) has increased, then in step # 320 the AF evaluation value (2) for the second AF area.
It is determined whether or not is increasing. Here, if the AF evaluation value (2) also increases, the focus lens is continuously driven in the near direction. Then, in step # 325, A
It is determined whether or not the F evaluation values (1) and (2) both decrease,
If these decrease at the same time, it is judged that the peak of focusing has been exceeded at the same time, and the subject is set to the same distance in both the first and second AF areas (step # 330). Then, the focus lens is returned to the focal point (step # 355), and step #
At 360, the lens drive is stopped. Then step # 3
At 65, it is monitored whether or not the subject has changed (specifically, the AF evaluation values (1) and (2) are monitored), and if there is no change, this routine ends. If there is a change, it is determined that the subject has moved, the process returns to step # 300, and the flow from step # 300 is executed.

In step # 325, the AF evaluation value (1)
If both (2) do not decrease, the process returns to the step # 315 of determining whether or not the AF evaluation value (1) increases while the focus lens is driven in the near direction.

Next, the case where the AF evaluation value (1) is increasing and the AF evaluation value (2) is not increasing (NO in the determination in step # 320) will be described. In this case, the subject in the first AF area is closer to the current focus position, and the subject in the second AF area is farther from the current focus position. This algorithm is the first
Since the subject closer to whichever of the second AF areas is focused, the focus lens is continuously driven in the near direction.

In step # 335, the value of the AF evaluation value (1) is monitored, and if it starts to decrease, the flag F is detected in step # 350.
Determine the state of LG. If FLG = 1, the lens is driving in the near direction. Therefore, the point at which the AF evaluation value (1) starts to decrease following the AF evaluation value (2), that is, the peak of the AF evaluation value (1) of the near-side subject can be detected. Therefore, in step # 355, the focal point (peak point) is returned to and the focus lens drive is stopped (step # 3).
60).

If the flag FLG is not 1, the focus lens is being driven to the far side, and the point where the AF evaluation value (1) starts decreasing after the AF evaluation value (2) is the AF of the far side object. It is a point beyond the peak of the evaluation value. Therefore, FL
When G is not 1, step # 350 to step # 3
In step 70, FLG = 1, the focus lens is driven in the near direction (step # 375), and the first step # 31
Return to 5.

Next, in step # 315, A
If the F evaluation value (1) does not increase, step # 340
By advancing to, it is determined whether or not the AF evaluation value (2) increases, thereby determining whether or not the second AF area has the in-focus point. If the AF evaluation value (2) has increased in step # 340, it is determined in the next step # 345 whether or not the AF evaluation value (2) will decrease. If the AF evaluation value (2) has decreased, step # 350
The flag FLG is discriminated by. Then, if FLG is 1, it is determined that the focus is on the near side, and the processes in and after step # 355 are performed. If FLG is not 1, step # 370, #
After 375, the process returns to step # 315.

If the AF evaluation value (2) has not increased in step # 340, the focus lens is driven in the far direction in step # 380, and the flag FLG is set to FLG = 0. Then, in step # 390, the AF evaluation value (1)
(2) If both decrease, the process jumps to step # 330, and the processes after step # 330 are performed. But step #
If both AF evaluation values (1) and (2) do not decrease at 390, the process returns to step # 315 and the subsequent processing is executed.

As described above, in the AF routine,
If the AF evaluation values (1) and (2) start decreasing at the same time, the first,
It is determined that the second AF areas are at the same distance. During the driving in the near direction, the AF evaluation value (1) or (2) which is later decreased is focused. When driving in the far direction, the focus is first on the one that begins to decrease. In the AF routine, the driving direction of the focus lens is determined by the FLG, and when FLG becomes 1, it is determined that the person is closer, and the focus lens is returned to the in-focus point and the focus lens is stopped.

Now, returning to the flow of FIG. 6, step #
When the processing of the AF routine described above is completed at 160, steps # 165 to # 180 are executed. This is to select the person to be tracked because it is possible that the persons in the first and second AF areas move in different directions after the processing of the AF routine, if they are at the same distance.

In this case, the person near the center of the screen is prioritized (step # 170). If the persons in the first and second AF areas are equidistant from the center of the screen, the area The larger person is selected (steps # 175, # 180). If the persons in the first and second areas are not at the same distance, it is not necessary to select the person to be tracked, so steps # 170 to # 180 are skipped.

After performing the AF routine of step # 150 and the processes of steps # 160 to # 180, the individual is determined in step # 185, and this flow ends.
The above is the case where the individual extraction is the first time, so step #
The case where the process proceeds to step # 105 and the subsequent steps assuming that there is no previous individual in 100 has been described. However, if it is not the first time, it is determined in step # 100 that the previous individual exists, and the process proceeds to step # 190 in FIG. This step # 19
It also comes to 0 when NO is determined in each of the steps # 110, # 125, and # 135.

In step # 190, it is checked whether or not there is a skin color portion near the previous AF area. Specifically,
As shown in FIG. 3 (b), the previous AF area 61 is set as an AF area 62 that is one size larger, and the area of the skin color portion included therein is checked. Here, when there are a plurality of skin color portions, the larger one is selected (steps # 195, # 200). The AF area having the same size as the previous AF area 61 is moved within the AF area 62 which is one size larger than the previous one, and is set to the position including the most skin color portion (step # 205). When the entire AF area is occupied by the skin color portion, the AF area is enlarged by one (step # 210). The AF area part is determined as an individual in step # 215, and then the step # 22.
Focusing is performed by the 0 AF routine.

In step # 190, when there is no flesh color near the position of the previous AF area, (a) the person is at the same position, but the flesh color part disappears, such as when the person is facing backwards.
(B) It is possible that the person disappeared from the screen.
Therefore, the same AF area as the previous time is set (step # 22
5), the focusing operation is performed using the AF routine (step # 230), and if the difference between the distance measured last time and the distance measured this time is small, it is considered to be the case of (a) and the individual is determined (step # 230). 235, # 240).

When the distance between the previous time and this time is different, the AF evaluation value is checked to determine whether or not there is a highly reliable in-focus point (step # 245). The process proceeds to step # 240 to determine the individual. If there is no in-focus point, there is no person (step # 250), and an in-focus operation is performed in the AF routine processing in the default AF area (steps # 255 and # 26).
0).

Next, the main routine of FIG. 14 will be described. First, when the power is turned on, the AF area is initialized in step # 400, and then the focusing operation is performed in the AF routine in step # 405. Then, the individual extraction and refocusing operations are performed in the individual extraction routine of step # 410 (described in FIG. 6). Then, the process proceeds to step # 415 to determine whether or not the APZ mode is set. APZ mode is APZ
It is set by operating the switch 49. Where APZ
If the mode is set, APZ in step # 420
Routinely, the zoom lens drive speed direction is set so that the area of the skin-colored portion is constant. The flow of this APZ routine is shown in FIG.

In FIG. 15, first, in step # 500, the area of the skin color portion is calculated, and the value is set to S1. continue,
In step # 505, the size relationship between the previous area S0 and the current area S1 is checked. If S0> S1, step # 510
Go to and move the zoom lens toward the telephoto end, and press S0 <
If it is S1, the process proceeds to step # 515 to move toward the wide-angle end, and zooming is performed until S0 = S1. Then, when S0 = S1, step # 520
Then, the routine ends. In addition, the previous area S0
Alternatively, the operator may freely set the area to be compared with the area S1 at this time, or may be predetermined by the camera side.

Returning to FIG. 14, when the APZ routine is completed at step # 420, the process proceeds to the next step # 425, where the skin color AE routine is executed. This skin color A
The E routine is shown in FIG. When this skin color AE routine is entered, first in step # 600, the flag FLG for driving the lens 1 in the near direction is set to 0, and in step # 605.
Then, it is determined whether or not the level of the skin color part is 55 IRE or higher.

Here, the level of the skin color portion is gated by the screen division circuit 41, averaged by the averaging circuit 42 and A / D converted by the A / D converter 43, and then the control microcomputer 2
It is the level of the luminance signal of the flesh color portion input to 1. If the level of this skin color portion is 55 IRE or higher, it is determined in step # 610 whether it is within 75 IRE. If it is within 75 IRE, the routine proceeds to step # 610 to end this routine. That is, 55 IRE to 75
If there is no problem within the range of IRE, no correction is performed.

However, if it is determined in step # 605 that it is less than 55 IRE, the exposure is underexposure, so the aperture 2 is opened by a fixed amount in step # 625. As a result, when the aperture 2 is opened, the aperture 2 cannot correct any more. Therefore, the AGC amount is increased in step # 635, the process returns to step # 605, and the flow from step # 605 is executed. Finally, the signal level of the skin color part is set to 55IRE
Bring to the range of ~ 75 IRE. Note that step # 635
The increase of the AGC amount is performed by the control microcomputer 21 giving a gain increase signal as an AGC correction signal to the AGC circuits 7, 8 and 9 through the D / A conversion circuit 45 and the addition circuit 46.

If it is determined in step # 610 that the value exceeds 75 IRE, then overexposure has occurred, so the aperture 2 is closed by a fixed amount in step # 640. Then, the process returns to step # 605, and the flow from step # 605 is executed to finally set the signal level of the skin color part to 55 IRE to 7
Brings to the range of 5 IRE.

Returning to FIG. 14, when the flesh color AE routine in step # 425 ends, the next step # 43
At 0, it is determined whether or not the color correction switch 50 is turned on, and if it is turned on, a color correction routine is executed at step # 435. This color correction routine corrects the color reproduction of the extracted skin color portion based on the operation content of the color correction switch 50. The color reproduction is corrected by receiving the signal of the human part detected by the human detection microcomputer 35, tracking the human part, and sending the color reproduction correction data from the control microcomputer 21 to the color reproduction correction circuit 30. At this time, the color reproduction correction circuit 30 adjusts the gain of the input R2, G2, and B2 so that the ratio becomes a ratio based on the color reproduction correction data.
3 and B3 are output.

After executing the color correction routine, the process proceeds to step # 460 to determine whether or not the auto trigger switch 44 is auto-triggered. Here, if the auto trigger is applied, step # 46.
After executing the auto-trigger routine shown in FIG. 17 in step 5, the process returns to step # 410, and if not, the process directly returns to step # 410.

Here, the auto trigger routine will be described. In FIG. 17, first, it is determined whether or not the tape cassette is loaded in the deck portion 14, and if it is not loaded, the process proceeds to step # 725 to end this routine, but if it is loaded, the next step # At 705, it is determined whether the recording mode is set. If the recording mode is not set, the routine is ended in step # 725. If the recording mode is set, the process proceeds to step # 710 to determine the presence / absence of an individual. If there is an individual, the recording state is set in step # 715 and the process proceeds to step # 725. If there is no individual, step # 715.
At 720, the recording is paused and the process proceeds to step # 725. When the tape cassette is installed and the recording mode is set in this way, recording is performed when an individual is detected within the angle of view, and when the individual disappears from the screen, recording is automatically stopped. As a result, it is possible to reduce the waste of tape without missing a recording opportunity.

Although the embodiment of FIG. 5 described above shows only the video system and omits the audio system, the embodiment of FIG. 18 has the same video system as that of FIG. 5 and further features of the audio system. Is shown. In FIG. 18, an auto pan / tilt switch 70 controls ON / OFF of the auto pan / tilt. When it is ON, an electric pan head drive signal from the control microcomputer 21 causes a pan head drive circuit 77 to move the cloud. Stand 76
To pan / tilt. Arrows 78 and 79 indicate the movement of the platform 76 in this case. On the other hand, a signal indicating the amount of movement of the platform 76 is transmitted to the control microcomputer 21 via the platform drive circuit 77. The control microcomputer 21 stops the platform drive based on this movement amount signal.

Next, the sound f special correction switch 71 controls ON / OFF of the frequency characteristic automatic control.
In the case of D / D by the f special control signal from the control microcomputer 21
The frequency characteristic of the audio preamplifier 73 is controlled through the A converter 75 to equalize the audio signal from the microphone unit 72. The equalized audio signal is given to the deck section 14 through the automatic level control circuit 74.

The various controls relating to the embodiment of FIG. 5 described above are performed in the same manner in the case of FIG. 18, but a description thereof will be omitted here, and control of a voice system not shown in FIG. Will be described only. In the main flow chart shown in FIG. 19, steps # 400 to # 435 have already been described in FIG. In FIG. 19, when the color correction routine of step # 435 is completed, the process by the control microcomputer 21 proceeds to step # 440 to determine whether or not the sound f special correction switch 71 is ON, and if it is ON, step # 445 is executed. The voice routine shown in FIG. 20 is executed to control the voice band by the individual information within the angle of view.
If it is F, this routine is skipped.

To explain the voice routine, first, in FIG. 20, in step # 800, the ratio of the total number of pixels of the skin color portion within the angle of view calculated by the individual extraction routine described above to all the pixels is calculated. . Then, in step # 805, it is determined whether the ratio is 5% or more,
If it is less than 5%, the process proceeds to step # 825 to end this routine. If 5% or more, next step # 81
It is judged whether it is 20% or less at 0, and if it exceeds 20%, the high range and the low range are cut at step # 820, and 3 to 4 KH.
After emphasizing z, the process proceeds to step # 825.

If it is 20% or less (hence 5% to 20%), it is determined in step # 815 whether or not the photographing distance of the flesh-colored portion is 10 m or more. If it is 10 m or more, the process directly proceeds to step # 825 and 10 m. If closer, step # 820
After correcting the frequency characteristic in step # 825, the process proceeds to step # 825. Thus, in the voice routine, the number of pixels in the skin color part is 2
When it is 0% or more, or 5% to 20% and the AF distance information is within 10 m, it is determined that the person is close to the person when the person is conscious of shooting or when shooting, and the voice of the person becomes clearer. For recording, the peak of the frequency characteristic of the preamplifier 73 is set to 3 to 4 KHz.

Returning to FIG. 19, after executing the voice routine in step # 445 or when the result of the determination in step # 440 is NO, it is determined in step # 450 whether the auto pan / tilt mode is ON. , ON, the auto pan / tilt routine shown in FIG. 21 is executed in step # 455.

Now, the auto pan / tilt routine shown in FIG. 21 will be described. In this auto pan / tilt routine, when a person deviates from the angle of view, the panning and tilting motors are driven to track the subject. When this routine is entered, it is first determined in step # 900 whether or not there is an individual this time. If so, it means that the individual is within the angle of view.
No need to do subject tracking, so do nothing and step #
Proceed to 960 to end this routine. The presence or absence of this individual is determined based on the individual extraction data described above.

If it is determined in step # 900 that the individual does not exist this time, the flow advances to step # 905 to determine whether or not the previous individual exists. Here, if there is no individual last time, it is not possible to perform subject tracking, so in step # 955, the camera platform 7
After resetting 6 to the initial position, this routine ends in step # 960. If there was a previous animal, step #
At 910, the previous individual position is determined. In this embodiment, FIG.
As shown in FIG. 2, four areas (area A1 to area A4) are prepared in advance, and it is determined which of these four areas the individual position is located.

Therefore, steps # 915, # 925, #
At 935 and # 945, the position of the previous individual is checked in order from area A1 to area A4. If it is in area A1, tilt the platform 76 upward (step # 920), and if it is in area A2, tilt downward. (Step # 93
0). If it is in the area A3, it is panned to the left (step # 940), and if it is in the area A4, it is panned to the right (step # 950).

Returning to FIG. 19 again, after executing the auto pan / tilt routine in step # 455, or when the auto pan / tilt mode is not in step # 450, the process proceeds to step # 460. Step # 460, Step #
Although 465 exists also in FIG. 14, in the auto-trigger routine in the case of FIG. 19, when it is determined that there is no individual in step # 710 as shown in FIG.
In step 16, it is determined whether the mode is auto pan / tilt mode. If the mode is not auto pan / tilt mode, the recording is temporarily stopped (step # 720). If the mode is auto pan / tilt mode, it is determined. It is determined whether or not it has been completed (step # 71
8) If it is completed, the recording is temporarily stopped, and if not completed, the recording is performed (step # 715).

[0068]

As described above, in the video camera of the present invention, since the skin-colored portion of a person is detected and distance measurement is performed, it is possible to focus on the person regardless of the position on the screen of the person. Therefore, there is an effect that it is possible to realize good shooting mainly for a person. Further, since the skin color part of the person is controlled to have a constant area regardless of the position on the screen, APZ (Adbanced Pr
Compared with ogram zoom, it has the effect of making it easier to control the size of a person.

[Brief description of drawings]

FIG. 1 is a block circuit diagram of a video camera that forms part of an embodiment of the present invention.

FIG. 2 is a diagram showing a routine of AF control using a video signal adopted in FIG.

FIG. 3 is an explanatory diagram thereof.

FIG. 4 is an explanatory view of a part of FIG.

FIG. 5 is a block circuit diagram of a video camera embodying the present invention.

FIG. 6 is a flowchart showing a control routine for the individual extraction.

FIG. 7 is a flowchart following FIG. 6;

FIG. 8 is an explanatory view of the individual extraction routine.

FIG. 9 is an explanatory diagram of an individual extraction routine.

FIG. 10 is an explanatory view of an individual extraction routine.

FIG. 11 is a flowchart showing an AF control routine when there are a plurality of individuals.

FIG. 12 is an explanatory diagram thereof.

FIG. 13 is an explanatory view of the same.

14 is a flow chart showing the main routine of the embodiment of FIG.
Chart.

FIG. 15: Flow chart of APZ routine.

FIG. 16: Flow chart of skin color AE routine.

FIG. 17: Flow chart of auto trigger routine.

FIG. 18 is a block circuit diagram of a video camera showing another embodiment of the present invention.

FIG. 19 is a flowchart showing the main routine.

FIG. 20 is a flowchart of an audio routine.

FIG. 21: Autopan / tilt routine flow chart
To.

FIG. 22 is an explanatory diagram thereof.

FIG. 23 is a flowchart of an auto trigger routine.

[Explanation of symbols]

1 lens 2 aperture 7, 8, 9 AGC circuit 10, 11 White balance amplifier 14 deck section 21 Control microcomputer 25, 38, 41 screen division circuit 30 color reproduction correction circuit 34 memory 35 Human control microcomputer 44 Auto trigger switch 49 APZ switch 50 color correction switch 70 Auto pan / tilt switch 71 Voice f special correction switch 73 preamplifier 76 pan head

   ─────────────────────────────────────────────────── ─── Continued front page    (72) Inventor Hajime Sasaki             2-3-3 Azuchi-cho, Chuo-ku, Osaka             Kokusai Building Minolta Camera Co., Ltd. (72) Inventor Kenji Mizumoto             2-3-3 Azuchi-cho, Chuo-ku, Osaka             Kokusai Building Minolta Camera Co., Ltd. (72) Inventor Hiroaki Kubo             2-3-3 Azuchi-cho, Chuo-ku, Osaka             Kokusai Building Minolta Camera Co., Ltd. (72) Inventor Yoshihiko Higashi             2-3-3 Azuchi-cho, Chuo-ku, Osaka             Kokusai Building Minolta Camera Co., Ltd. (72) Inventor Takehiro Kato             2-3-3 Azuchi-cho, Chuo-ku, Osaka             Kokusai Building Minolta Camera Co., Ltd. (72) Inventor Hiroshi Otsuka             2-3-3 Azuchi-cho, Chuo-ku, Osaka             Kokusai Building Minolta Camera Co., Ltd.

Claims (3)

[Claims] 1. A skin color part extraction circuit for extracting a skin color part from a video signal, a judging means for judging whether or not the skin color part extracted by the skin color part extraction circuit can be specified as a human part, and it is judged as a human part. And a distance measurement area forming means for forming a distance measurement area in a slightly wider range than the skin color portion. 2. A distance measuring area forming means forms a distance measuring area without the judgment by the judging means when a skin color portion is detected in the vicinity of the position of the previously detected skin color portion. The video camera according to claim 1. 3. A skin color part extraction circuit for extracting a skin color part from a video signal, a determination means for determining whether or not the skin color part extracted by the skin color part extraction circuit can be identified as a human part, and a zoom lens. A video camera, comprising: a taking lens; and means for driving the zoom lens so that the area of a skin-colored portion determined to be a human portion is constant.