JPH05207361A - Automatic exposure controller - Google Patents

Abstract

PURPOSE:To attain a stable correction for the spatial movement of a luminance distribution by setting areas which are mutually overlapped, detecting the low- pass components of a video signal, and putting emphasis on a dark in a frame set at a center at the time of a back light. CONSTITUTION:The average luminance of areas on a screen which include at least more than one picture element in vertical and horizontal directions, and which are mutually overlapped, is detected by an LPF 50. The effective maximum value of a signal to be housed in a dynamic range is estimated by a correction by a effective maximum valve estimating part 54 by using the result, and exposure correction amounts are calculated by a correction amount calculating part 56 by using the effective maximum value and the luminance value of a subject found by a central photometric part 55. In the luminance value found by the central photometric part 55, the average luminance of the dark part in the central frame is detected at the time of a back light. A diaphragm control or the pedestal correction of the video signal are operated by using the calculated exposure correction amounts by a diaphragm control part 9 and a signal processing circuit 5.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to automatic exposure correction for video cameras.

[0002]

2. Description of the Related Art Automatic exposure control devices such as video cameras are
The aperture is controlled so that the output video signal level becomes constant. As the aperture control, an average value method for averaging the entire one screen, a peak value method for detecting the maximum value in the screen, and a method in which both are mixed are used. The details of the conventional automatic exposure control apparatus will be described below with reference to FIG.

A subject image is formed on an image pickup device 3 through a lens 1 and a diaphragm 2, converted into an electric signal, and output through a signal processing circuit 5 which performs γ processing and the like. At that time, the diaphragm control is performed as follows using the signal from the image sensor 3. The average value detection circuit 7 and the peak value detection circuit 8 detect the average value and the peak value of the brightness of the entire screen, respectively, and the diaphragm drive circuit 9 drives the diaphragm 2 based on the results of the two detection circuits.

However, in the method based on the average value of the image signal, when the brightness distribution of the image is wide, for example, when an image of a person against a bright sky is taken, an image pickup state called backlight occurs. This is because the luminance distribution of the person to be imaged is lower than the average luminance value of the entire image, so the signal distribution corresponding to the person in the video signal is unevenly distributed in the low video signal portion, and It is a phenomenon of blackening.

In order to avoid such a phenomenon, it is necessary to correct the distribution in the image signal of the image portion to be imaged to a value higher than that in the case where the image average luminance is controlled. For example, the image quality at the time of backlighting can be improved by correcting the aperture in the opening direction so as to keep the image signal level of the imaging target portion high.

Therefore, there has been considered a method in which an input switch is provided at the input of the diaphragm drive circuit 7 and a user manually gives a switching signal at the time of backlighting to perform diaphragm control with a preset fixed control amount.

However, with this configuration, there are cases where proper aperture control cannot be performed due to an erroneous operation by the user. Further, since the control amount of the diaphragm is fixed, it is not possible to perform control suitable for each image. In order to solve this problem, it is necessary to automatically determine the backlight state and automatically perform the correction according to the backlight intensity of each image.

Therefore, there has been considered a method in which a region where a subject is considered to be present, for example, a central portion is metered, the backlight state is determined from the level difference from the background, and correction is performed according to the determined backlight state. As an automatic diaphragm device using this method,
For example, it is shown in Japanese Patent Laid-Open No. 2-18578. Using FIG.
The diaphragm control method will be described. First and second detection means 2
In 3 and 24, respectively, as shown in FIG.
The output signal level of each image sensor in the area and the second area around the area is detected. The outputs of the first and second detecting means 23 and 24 respectively pass through amplifiers 31 and 32, and a comparator 33 that compares the levels of the first region and the second region determines the degree of back light. That is, the first and second detection means 23, 2
When the output signals of 4 are x and y, and the amplification degrees of the first and second amplification units 31 and 32 are m and n, respectively,
When (x · m) ≧ (y · n), the forward light state, (x · m) <
When it is (y · n), it is determined that the backlight is on.

On the other hand, the first, second and third gate portions 20,
Areas 21 and 22 are divided into three areas t as shown in FIG.
The respective levels of 1, t2 and t3 are detected, and the addition section 34 combines the three signals through the first, second and third gain control sections 25, 27 and 29, respectively. The gains of the first, second and third gain control units 25, 27 and 29 are determined by the comparator 33.
Receiving the output of the first, second and third control units 26, 28,
Controlled by 30. The aperture control unit 9 uses the addition unit 34.
Aperture control is performed so that the output signal of 1 is maintained at a predetermined level.

Since the level difference between the first region (center) and the second region (periphery) is small in the normal light state, the comparator 33 determines that the light is normal light. At this time, the first, second and third control units 26, 28 and 30 change the gains G1, G2 and G3 of the first, second and third gain control units 25, 27 and 29 to G1> G2.
> G3 is set. However, G1 is G2, G
It is slightly larger than 3. Therefore, the aperture control close to the conventional full-screen photometric method is performed to make the average brightness of the entire screen constant.

In the backlit state, the first region in the center is darker than the second region in the periphery, so that the comparator 33 determines the backlight. At this time, the first, second,
The third control units 26, 28, 30 are set so that G1 >>G2> G3. However, G1 is significantly larger than G2 and G3. In this way, by setting the gain in the central region to be larger than that in the peripheral region and performing the center-weighted photometry method, it is possible to control the aperture according to the backlight condition.

[0012]

However, the conventional method has the following problems.

First, in the conventional method, the screen is divided into a plurality of areas (see FIG. 4), and the degree of backlighting is determined from the average luminance of the divided areas. However, the average value of the divided areas is
The correction state sometimes changed because the camera moved a little or the subject changed a little.

This problem occurs because the number of representative points (sample points) with respect to the size of the area (averaging area) for detecting the average brightness is small, so that aliasing distortion occurs when the brightness distribution is sampled. However, this causes a measurement error of the luminance distribution. However, when the number of sampling points is increased, the number of conditional expressions for determining the degree of backlighting increases, which makes it difficult to design the system.

Secondly, in the center-weighted method, since the correction amount is calculated by focusing on the preset central area, the setting (size, position) of the central area becomes a problem. That is, if a fixed frame is set near the center as shown in FIG. 5, there will be objects other than the dark subject in the frame,
The correction that is suitable for the subject cannot be performed, or the correction amount becomes unstable with respect to the movement of the subject or other objects.

Thirdly, when it is determined that the backlight is strong, the weighting is applied to the central area. Therefore, when most of the peripheral area is a high brightness area, many pixels are saturated beyond the dynamic range. Become. Therefore, the entire screen may be difficult to see due to the correction.

In view of the above point, the present invention has an object to provide an automatic exposure control device which is stable against spatial movement of a luminance distribution and performs correction in consideration of pixels saturated after correction.

[0018]

An effective maximum value estimating unit for estimating a maximum value of a video signal to be included in a dynamic range of a video signal processing circuit by exposure correction, and an exposure based on an output of the effective maximum value estimating unit. An exposure correction amount calculation unit for calculating a correction amount and an exposure control unit for performing aperture control and correction of a pedestal of an image pickup signal using an output of the exposure correction amount calculation unit are provided.

[0019]

According to the present invention, by providing the above means, it is possible to provide an automatic exposure apparatus which is stable in the spatial movement of the luminance distribution and which can perform correction in consideration of pixels saturated after correction.

[0020]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of an automatic exposure control device according to the present invention will be described with reference to FIG.

Of the respective elements, the same components as those in FIGS. 2 and 3 are designated by the same reference numerals. The image of the subject is the lens 1,
An image is formed on the image sensor 3 through the diaphragm 2 and converted into an electric signal. The converted signal is amplified by the amplifier 4, passes through the signal processing circuit 5 that performs γ processing, etc., and is sent to the video signal output terminal 6. At this time, diaphragm control is performed as follows based on the signal obtained from the image sensor 3.

First, the low-frequency component of the input video signal is filtered by the LPF 50 from the signal of the image pickup device. In this embodiment, this LPF is, as shown in FIG. 6, 1/3 of the horizontal / vertical frame width.
This is performed by detecting the average luminance in each ¼ width region 601 arranged in the horizontal and vertical directions with two offsets. As a result, there are a total of 1024 regions of 32 vertical and 32 horizontal in the entire screen. However, when installed in this manner, there is a portion where the area exceeds the image area around the image area. At these points, it is necessary to perform a process corresponding to the loss of the image data in the portion beyond the image boundary when obtaining the average value. As a processing method for this purpose, there is a method in which the area of the region for which the average value is obtained is made variable, or the insufficient image data is replaced with a fixed value or a boundary value. Since this point does not form the subject of the present invention, details thereof will be omitted as one of these methods.

Here, the size of the averaging region is determined by the size of the detection target used for judging the backlit state. When the area is small, it becomes vulnerable to noise such as spot light, and when the area is large, the reaction to the backlight is large. Get dull.

The number of sampling points for the averaging area is 8 in each of the vertical and horizontal directions, but here it is assumed that one cycle detects a change in the luminance distribution about the width of the averaging area. Eight sample points were set. This will be described with reference to FIG. For simplicity, it is assumed that the input signal is one-dimensional and that the averaging uses a signal of width W. Now, assume that a signal whose averaged result is as shown in FIG. 7 is input (1 cycle =
W). As in the conventional case, the number of sampling points is 1 for the averaging area W.
In this case, the value varies in the range of -1 to 1 depending on the position of the sampling point. As described above, when eight sample points are set for the averaging area W, the error due to the shift of the sample points is about 0.4 at the maximum. In this way, if the number of sample points is large, the error due to the position of the sample points is small, and stable determination can be performed with respect to the movement of the camera or the subject. If the system scale can be increased, the number of sampling points can be increased to improve the accuracy.

Next, a method of performing backlight determination and correction amount calculation using the output of the LPF 50 will be described. Here, the correction amount means α and β when the original signal Y is linearly converted as Y ′ = α * Y + β, and corresponds to the gain correction by the diaphragm and the pedestal correction in the signal processing circuit, respectively.

The central photometer 55 detects the average luminance Yt of the central area. The method of detecting Yt will be described later.

The dark area calculation section 51 counts the number of areas in which the average brightness detected by the LPF 50 is below a certain threshold value X.

The maximum brightness detecting section 52 detects the maximum brightness among the average brightnesses in the area detected by the LPF 50.

The effective maximum value estimating unit 54 calculates the effective maximum value using the output Ymax of the maximum brightness detecting unit 52 and the output St of the dark area calculating unit 51. Here, the effective maximum value is the limit value of the brightness saturated by the correction, and is the maximum value of the brightness signal that is within the dynamic range after the correction, that is, the lower limit of the brightness signal that saturates the dynamic range after the correction. The details of the effective maximum value will be described below.

Consider a case where a backlight image is input and gain correction is performed so as to raise the average luminance Yt of the central region to a certain value. When the number St of dark areas is large, it is considered that there are few parts that are saturated by the correction and the image quality is less deteriorated. Therefore, the gain correction amount α can be set large. That is, there is no problem even if the luminance value saturated by the correction is considerably low, and the effective maximum value can be set low. On the other hand, when St is small, the area saturated by the correction is large and the screen may be difficult to see. In this case, it is easier to see the correction so that most of the signal is within the dynamic range of the camera. That is, the effective maximum value is
It takes a value near the maximum value Ysat of the dynamic range.

When the effective maximum value having the above properties is expressed as a function h of St, a monotonically decreasing function as shown in FIG. 8 is obtained. However, when h (St) is larger than the actual maximum value Ymax in the screen, image deterioration does not occur even if the effective maximum value is corrected to Ymax. Therefore, the effective maximum value Ytop is set to Ytop
= Min (Ymax, h (St)).

The output Ytop of the effective maximum value estimating unit 54, the output St of the dark region area detecting unit 51, and the output Yt of the central photometry unit 55 obtained as described above are sent to the correction amount calculating unit 56. Be done. In the correction amount calculation unit 56, the equation (1),
The gain α and the pedestal β of the exposure correction amount are calculated based on (2).

[0033]

[Equation 1]

[0034]

[Equation 2]

The target brightness calculating unit 57 sets a target value of the average brightness of the entire screen for each vertical scanning (frame or field) from each correction amount obtained by the correction amount calculating unit 56.
That is, the target brightness Yall 'is set to Yall' = Yinit *
α, Y init: Calculate as initial target brightness.

The aperture controller 9 controls the average luminance Yal of the entire screen.
The diaphragm is controlled so that l is kept at the above target brightness Yall ′.

Next, details of the operation of the system will be described. When an image having a large St is input due to strong backlight, the effective maximum value estimation unit 54 sets the effective maximum value Ytop low due to the nature of the function h (St). The correction amount calculation unit 56 calculates α and β satisfying (Equation 1), and the subject brightness at the center is changed from Yt to K, and the pixel having the brightness Ytop is Ys.
Corrected to at.

On the other hand, when an image is input in the backlit state, but the St is not so large, the effective maximum value estimation unit 54 sets the effective maximum value to a large value. When the correction amount calculation unit 56 calculates the correction amount based on (Equation 1), the brightness Yt of the central region is corrected to a constant value K, but the brightness value saturated by the correction is high, and the aperture correction amount α
Becomes smaller. Therefore, the brightness of the subject can be increased without increasing the number of pixels saturated by the correction, and the image is not deteriorated.

When a normal-light image is input, both the aperture correction and the correction of the signal processing circuit are fixed, and the correction is the same as the normal full-screen average value control.

When the above control is actually performed, the number St of regions having dark brightness varies depending on the correction amount at that time. Therefore, when the threshold value X is fixed, the calculated correction amount always varies. In order to continue measuring the same St when the image state is the same, it is necessary to change the threshold value X according to the correction amount at that time. Therefore, the threshold value setting unit 58
Then, the threshold value is updated every vertical scanning. That is, the initial threshold value is X and the previously calculated gain is α '.
Then, the threshold value X ′ at this time is set as X ′ = X * α ′.

Next, with reference to FIG. 9, a method of detecting the average brightness of the central area in the central photometry section 55 will be described.

As described above, when the central area is fixed, the correction amount does not adapt to the darkness of the object, or the object and other objects are not present because the object other than the dark object exists in the area. It becomes unstable when the object moves. In order to improve this problem, in the backlit state, the dark portion in the center is extracted and the control that emphasizes the average luminance is performed as follows.

The central frame average detection section 61 detects the average luminance Yc within the frame set at the center.

The central frame dark area average detection unit 62 and the central frame dark area area detection unit 63 detect portions in the central frame whose luminance is less than or equal to a threshold value, and detect the average luminance Yd and the area Sd of that portion, respectively. To do. However, the threshold is calculated by the dark region threshold setting unit 60, and here, the maximum and minimum of the average luminance in the region detected by the LPF 50 of FIG.
As in, threshold value = (Ymax + Ymin) / 2 is set.

The in-center-frame determining unit 64 determines which of the average brightness Yc in the central frame and the average brightness Yd in the dark portion in the central frame is adopted as the brightness in the central region. That is, using the output Sd of the central frame dark area area detection unit 63 and the output St of the dark area calculation unit 51, (1) St> K
When 1 and Sd> K2, the brightness Yt of the central area = the output Yc (2) St <= K1 or Sd <of the average detection unit in the central frame
= K2, the brightness Yt of the central area is set to the output Yd of the central frame dark area average detection unit.

When the brightness of the central area is detected in this way, in the backlight condition, a correction amount that emphasizes the dark portion in the central frame can be calculated, and stable correction that does not depend on the bright background image can be performed. However, if there is no dark area above a certain level in the center frame, it may be dangerous to attach importance to the small area as a subject. For example, if the subject in the center becomes brighter than a certain amount, such as over-exposure, if you extract and correct the dark area in the center frame, the exposure will match the dark background, and the subject will appear saturated. Disappear. Therefore, the brightness of the central area is set as the dark area within the frame only when the area Sd of the dark area is K2 or more as described above.

In this way, by obtaining the central luminance by extracting the dark portion in the center in the backlight state, the central luminance value in the correction amount calculation unit described above in the backlight state in which the correction amount is large is stably exposed. Is controlled.

[0048]

As described above, according to the present invention,
After the correction, it becomes possible to perform the correction in consideration of the saturated pixel, and the stable exposure to the spatial movement of the brightness distribution can be performed, and the automatic exposure that can appropriately maintain the state of the saturated pixel in the image according to the image. A control device can be provided.

[Brief description of drawings]

FIG. 1 is a configuration block diagram of an automatic exposure control apparatus according to an embodiment of the present invention.

FIG. 2 is a configuration block diagram of a conventional automatic exposure control device.

FIG. 3 is a configuration block diagram of a conventional automatic exposure control device.

FIG. 4 is an explanatory diagram of first and second regions extracted by the detection means of the device shown in FIG.

FIG. 5 is an explanatory diagram of each region extracted by the gate unit of the device shown in FIG.

FIG. 6 is an explanatory diagram of a region to be averaged by the LPF of the same device.

FIG. 7 is an explanatory diagram of the relationship between the number of sample points for the area averaged by the LPF of the same apparatus and the error due to the position of the sample points.

FIG. 8 is a diagram showing the relationship between the number St of dark areas and the effective maximum value.

FIG. 9 is a block diagram of the configuration of the central photometry unit in FIG.

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Tsutomu Mori 1006 Kadoma, Kadoma City, Osaka Prefecture Matsushita Electric Industrial Co., Ltd.

Claims (6)

[Claims] 1. An effective maximum value estimating unit for estimating a maximum value of a video signal to be within a dynamic range of a video signal processing circuit by exposure correction, and an exposure correction amount is calculated based on an output of the effective maximum value estimating unit. An exposure correction amount calculation unit for
An automatic exposure control apparatus comprising: an exposure control unit that controls an aperture and corrects a pedestal of an image pickup signal by using an output of the exposure correction amount calculation unit. 2. A dark area calculation unit for counting the number of pixels of a predetermined value or less in a screen, and a signal within a dynamic range of a video signal processing circuit when a diaphragm is opened to a limit where image quality deterioration does not occur before correction. An effective maximum value estimation unit that estimates the maximum value of the converted values from the output of the dark region calculation unit, a central photometry unit that detects the average luminance of the image pickup signal in the central region, the effective maximum value estimation unit, and the A correction amount calculation unit that calculates a gain correction amount and a pedestal correction amount of an image pickup signal from the output of the central photometry unit, and an aperture control of a product of the gain correction amount of the output of the correction amount calculation unit and a predetermined value. A target brightness calculation unit for setting a target value of, a full screen level detection unit for detecting the average brightness of the image pickup signal of the full screen, and an output so as to match the output of the full screen level detection unit with the output of the target brightness calculation unit. A diaphragm control unit for controlling, Automatic exposure control apparatus characterized by comprising a signal processing control unit adding the correction amount of the pedestal of the output of the correction amount calculation unit to a signal obtained from the image element. 3. A level detector for detecting a signal level of a low frequency component of a spatial frequency of a video signal, an exposure correction amount calculator for calculating an exposure correction amount based on the output of the level detector, and the exposure correction. An automatic exposure control apparatus comprising: an exposure control unit that controls an aperture and corrects a pedestal of an image pickup signal by using an output of a quantity calculation unit. 4. The automatic exposure according to claim 3, further comprising a level detection unit which includes at least one pixel in the vertical and horizontal directions and which detects an average value of areas on the screen which overlap each other. Control device. 5. A dark area calculation unit that counts outputs of a level detection unit that are equal to or less than a predetermined value, and an exposure correction amount that calculates an exposure correction amount by using the output of the dark region calculation unit as a judgment value for backlighting. The automatic exposure control apparatus according to claim 4, further comprising a calculation unit. 6. A central frame in-plane average detection section for detecting average brightness in an area set in the center of the screen, and a central frame for extracting pixels having a predetermined value or less in the area and detecting the average brightness. A dark area average detection unit, a central frame inner determination unit that sets the brightness of an object by switching the outputs of the central frame inner average detection unit and the central frame dark region average detection unit according to the backlight condition of the screen, and the central frame A correction amount calculation unit that calculates an exposure correction amount according to the brightness of the subject set by the internal determination unit;
An automatic exposure control apparatus comprising: an exposure control unit that controls an aperture and corrects a pedestal of a video signal by using an output of the correction amount calculation unit.