JPH0399586A - Automatic exposure adjustor - Google Patents

Abstract

PURPOSE:To facilitate the detection of the brightness distribution on a picked-up screen and to attain smooth exposure adjustment by adopting the fuzzy deduction employing a ratio of brightness evaluations of two optional areas among plural areas resulting from divisions of the screen as an input variable for the detection. CONSTITUTION:A picked-up screen is divided into areas A1-A6, and normalized brightness evaluation values V1-V6 are outputted from normalizing circuits 51-56. Then a brightness decision circuit 57 detects the brightness distribution of each area based on the evaluation values V1-V6 and the priority of each area is decided. Then the fuzzy deduction employing a ratio of brightness evaluations of two optional areas as an input variable is adopted from the decision processing. Since ambiguous border information is used as it is, the detection of the brightness distribution is facilitated. Moreover, the exposure adjustment in which the contrast change in the screen is smooth is implemented by using the priority as the result of the detection for the exposure control.

Description

DETAILED DESCRIPTION OF THE INVENTION (a) Field of Industrial Application The present invention relates to an imaging device such as a video camera that automatically adjusts exposure.

(B) Conventional Technology In video cameras, control of the brightness level of a captured video signal using aperture, gain, etc., or so-called exposure adjustment, is a very important issue along with focus control.

Conventionally, as this automatic exposure adjustment mechanism, a method of detecting the average brightness level, peak value, etc. of the brightness level of the image capturing screen, and controlling the aperture and the gain for the captured image signal based on these levels has been popular. With this method, if there are high-brightness areas such as light sources on the screen, or if the background is dark,
Appropriate exposure may not be obtained for the main subject due to surrounding effects.

Therefore, the applicant of the present invention previously proposed countermeasures for these problems in Japanese Patent Application No. 4344/1983.

This measure involves dividing the image capture screen into multiple areas in advance, extracting the captured video signal for each area, and integrating the low-frequency components over one field to evaluate the brightness that indicates the brightness level of each area. Each brightness evaluation value is compared with a standard value that is expected to be obtained when an abnormal brightness part such as a light source exists in the area, and if the area exceeds the standard value, it is determined that an abnormal brightness part exists. By determining that the abnormal brightness area exists, the average value of the brightness evaluation values of the area other than the area where the abnormal brightness area exists matches the target value, thereby eliminating the influence of the abnormal brightness area on the entire imaging screen and confirming that the abnormal brightness area exists. In other words, the exposure is adjusted so that the subject in the area where the subject is not photographed is in the optimum exposure state.

(c) Problems to be Solved by the Invention As mentioned above, the detection result of the so-called brightness distribution of the image capturing screen is divided into two parts, and the presence or absence of an abnormally bright area or the position of the abnormally bright area on the screen is detected by dividing the image capturing screen. Executing exposure adjustment accordingly is extremely effective when photographing in backlit or road lighting conditions.

However, when detecting the luminance distribution as in the conventional example,
In a configuration where cases are simply classified based on whether the brightness evaluation value exceeds the reference value, it is impossible to distinguish between a state in which the brightness evaluation value slightly exceeds the reference value and a state in which the brightness evaluation value extremely exceeds the reference value. Therefore, for example, if the brightness evaluation value of an area that includes an abnormal brightness area is approximately equal to the reference value, this brightness evaluation value will be greater than or equal to the reference value at a certain point, and will be less than the reference value at a certain point, and the influence of the abnormal brightness area will be Exposure control is performed each time, such as ignoring the abnormal brightness area and keeping the area where there is no abnormal brightness part in the optimal exposure state, or maintaining only the area where the abnormal brightness part exists under the influence of the abnormal brightness part in the optimal exposure state. As a result, the brightness of the entire screen changes intermittently, resulting in an unsightly screen.

In addition, if a light source that was initially low brightness gradually becomes high brightness, the brightness evaluation value of the area containing this light source will eventually exceed the standard value, and the exposure of the light source will be adjusted just before and after exceeding the standard value. Since the degree of influence on the image changes greatly, the brightness of the image capture screen (area that does not include the light source) abruptly switches from dark to bright at this boundary, resulting in an unsightly screen.

(d) Means for Solving the Problems The present invention detects the luminance distribution of the imaging screen by using fuzzy inference using the ratio of the luminance evaluation values of any two regions of a plurality of regions into which the imaging screen is divided as an input variable. It is characterized by the fact that it is carried out using

(E) Function Since the present invention is constructed as described above, it is extremely easy to detect the brightness distribution of the image capturing screen, and if exposure control is performed using this detection result, the brightness of the image capturing screen changes rapidly. This allows smooth exposure adjustment without any hassle.

(f) Example An embodiment of the present invention will be described below with reference to the drawings.

FIG. 1 is a circuit block diagram of the device of this embodiment.

The incident light passes through the lens (1) and the diaphragm til! (2
), the light is photoelectrically converted by the imaging circuit (3) and output as a captured video signal. This imaged video signal is amplified by a variable gain amplifier (4) and sent to a video circuit, and is also amplified by a variable gain amplifier (4) and sent to a video circuit.
3), is supplied to the integrator (80).

L P F (22) extracts the low frequency component of the luminance signal in the captured video signal and outputs it to the subsequent switching circuit (26).

The synchronization separation circuit (23) extracts vertical and horizontal synchronization signals from the imaged video signal, and the subsequent switching control circuit (25) combines these vertical and horizontal synchronization signals with the CC of the imaging circuit (3).
Based on the fixed oscillator output used to drive D, the third
A switching signal for screen division over six areas (A1) to (A6) in the figure is issued.

The switching circuit (26) receives the switching signal and switches each region (
AI) to (A6) are switched sequentially, L P F
(22) The output is time-divided for each region by this switching circuit (26) and supplied to integration circuits (31) to (36), respectively.

The integrating circuits (31) to (36) all perform A/D conversion of the output of the switching circuit (26) as shown in FIG.
(27), this A/D switching output and the subsequent latch circuit (
28) A digital integrator consisting of an adder (29) that adds the output and a latch circuit (28) that latches the added output, and the low frequency component of the luminance signal in the corresponding area is sampled at a predetermined sampling rate. A/D converted at regular intervals,
This A/D converted data is integrated over one field period. Here, the integration circuit (31) outputs the integrated value for one field of the low frequency component of the luminance signal within the area (A1) to the memory (41), and similarly thereafter
2) The integrated value for one field of the luminance signal in (A3) (A4) (A5) (A6) is determined by the integration circuit (32) (33
)(34,)(35)(36) respectively memory(42)
It will be output to (43) (4, 4) (45) (46). The latch circuit (28) is reset for each field, and each memory retains the data immediately before the reset of each latch circuit, and the data is updated for every l field.

By the way, the areas (AI) to (86) have the areas (Sl) to (S6), respectively, area (A1) is located at the center of the screen as shown in FIG. 3, and area (A2) is located at the area ( It is located on the outer periphery of A1). Furthermore, an area (A2) is formed around this area (A2).
3) to (A6) are arranged.

When the integration for one field, which is one screen, is completed, the latest integrated value for one field in each area held in the memory (41) to (46) is the brightness evaluation value of each area. (Y
l) to (Y6) are the subsequent simple average circuits (68),
Each normalization circuit and each weighting are output to the circuit.

The normalization circuits (51) to (56) convert the brightness evaluation values (Yl) to (Y6) in each region into respective areas (Sl) to (S6
), and the brightness evaluation value per unit area of each region is normalized brightness evaluation value (vl) to (V6) (however, V1=
Y1/Sl, V 2 =Y 2/S2, . . .

The priority determination circuit (57) determines each normalized luminance evaluation value (Vl
) to (6), the priority (weight) of each area is determined. The priority determination process in this priority determination circuit (57) is executed according to a flowchart as shown in FIG. Specifically, the following six rules are used.

[Rule (1)] rif Vl and V2 are close and V 1 and v3 are not close then area (AI) (A2) priority j [Rule (2)] rif Vl and v2 are not close and V 1 and ■3 are close then Area (AI), (A3) Priority” [Rule (3)
CorifVl and ■2 are not close andV 1 and V3 are not close then area (AI) priority" [Rule (4)] rirVl and V2 are close andV 1 and ■3 are closeth
en area (A1), (A2), (A3) priority" [Rule (5)] "100 max(Vi) (i=1 to 6) is small then
``All areas have the same priority'' [Rule (6)] Prioritize areas (A1) where rif max (V i ) is not small and the simple average value is small.'' These rules are as shown in Figures 6 to 11. , “
Conditions such as “close” and “small” are “V2/VIJ
It is defined by a membership function for each input variable such as rmax (Vi)J, and has the priority (wik) of each area as a conclusion part. In addition, the inference is the usual m
The in-max method is used.

Next, each rule will be explained in detail.

[Rule (1)] is defined by membership functions such as those shown in FIGS. 66a and 66b. Figure 6(a) shows rV
This is a membership function for the input variable (V2/Vl) that indicates the degree to which condition (1) of rule (1) "l and v2 are close" is satisfied. That is, the normalized brightness evaluation value (vl) of area (A1) and the normalized brightness evaluation value (v2) of area (A2)
), the input variable is V2/Vl, and V2/V1 . The membership value (u, . . . ) is determined by substituting the input variable (V 2 /V 1 ) in the latest field into the chevron-shaped membership function that takes a maximum value when ˜1. Note that when V2/V1=1, the membership value (u, .) is maximum.

Figure 6(b) shows the input variable (V3/
is a membership function for Vl). That is, in order to determine the degree to which the normalized brightness evaluation value (Vl) of the area (A1) and the normalized brightness evaluation value (U3) of the area (A3) are not close, the input variable is set to V3. /Vl
Then, the input variable in the latest field (
V3/Vl) by substituting the membership value (
u, , ) can be found. Incidentally, when V3/V1=1, the membership value (u, ,) becomes the minimum. In this way, Figure 6 (
a) By (b), the membership value (u, , ) (u.

, ) will be calculated. Note that this calculation is based on the second
This corresponds to S T E P (100) in the flowchart in the figure.

The membership value (u, ,) (u, ,) is ST
In E P (101), the minimum value of both, that is, the smaller membership value is selected as the degree of fulfillment (Ul) of rule (1). In the example of FIG. 6, since a ll<u +t, U1=u is set.

The operations of S T E P (100) and (101) described above are also executed for the remaining five rules.

[Rule (2)] is defined by valley-shaped and mountain-shaped membership functions as shown in Figure 77a) and (b), and as in the case of Figure 6, the rule that rVl and U2 are not close (
2) Membership value (u, , for condition (1))
) is closer to (a), and rVl and U3 are closer.'' Membership value (u
o) is determined from (b), and in 5TEP (101), the smaller membership value (u, , ) (u, , ) is selected as the degree of fulfillment (U2) of rule (2). In the example of FIG. 7, since u@)>ust, U2=uxt is set.

[Rule (3)] is defined by a valley-shaped membership function as shown in Figures 8(a) and (b), and as in the case of Figure 6, rule (3) states that rVl and U2 are not close. The membership value (Ul,) for condition (1) is (a
), and the rule that rVl and U3 are not close (
3) Membership value (u, , for condition (2))
) is determined from (b), and in S T E P (101), the smaller membership value (u, , ) (u, , ) is selected as the degree of fulfillment (U3) of rule (3). 8th
In the example shown in the figure, ust<ust, so U 3 ” u
s is set to 1.

[Rule (4)] is defined by a mountain-shaped membership function as shown in Figures 99a and (b), and the membership value for condition (1) of rule (4) that "vl and U2 are close" is The membership value (u,',) for condition (2) of rule (4), ``u,,) is closer than (a), and rVl and U3 are closer,'' is determined from (b), and S T E
Membership value (u.

, )(U, ,) is smaller than the probability that rule (4) holds true (U
4) is selected. In the example of FIG. 9, Ull>uJ
t, so U 4 = u 4 t is set.

[Rule (5)] is] As shown in Figure 1, the maximum value (max (Vi) of all normalized luminance evaluation values (Vl) to (U6)
) (where i=1 to 6) is the input variable, and this max
It is defined by a membership function shown by a simple decreasing straight line indicating the degree to which (Vi) is small, and once max(Vi) is determined, the membership value (U, , ) is uniquely determined. still,
This membership value (u, ,) becomes smaller as max (Vi) becomes larger. S T E P (1
In 01), since there is only one membership value for rule (5), U5=u, 1 is set after rule (5) is established (U5).

[Rule (6)], as shown in Figures 11(a) and (b), is a membership function with a simple increasing straight line with max (Vi) as an input variable, as in rule (5), and a total normalized luminance evaluation. It is defined by a membership function of a simple decreasing straight line with values (vl) to (U6) as input variables. That is,
In the membership function of FIG. 11(a), rmax(
Condition (1) of rule (6) that “Vi) is not small”
In order to determine the degree to which max(Vi) is not small in , if max(Vi) is determined as an input variable, the membership value (U, , ) can be determined. Note that this membership value (u, ,) becomes smaller as max (Vi) becomes smaller. In addition, in the membership function of FIG. 11(b), input variables are used to determine the degree to which the simple average value (Z, ) is small in condition (2) of rule (6) that "the simple average value is small". If the simple average value is determined as , then the membership value (U, . . . ) can be determined. Note that this membership value (u,,) becomes smaller as the simple average value becomes larger. S T E P (
101), the membership value (u,,) and (121
) to determine the degree of establishment of rule (6) (U6
) is set as U6=ust.

As described above, it is determined in STE I) (102) that the calculation of the degree of fulfillment (Ui) (i = 1 to 6) for all rules in STE P (100) (101) has been completed. Then, in STE P (103), the priority (Wk) (k=1 to 6) for each area is calculated. This priority (Wk) is calculated by weighted averaging of the conclusions based on the degree of establishment of each rule, as shown in the following equation.

In this formula (A), wik is the priority for each area regarding each rule, and is determined individually for each rule.

For example, regarding rule (1), "area (A1), (
In order to numerically indicate "prioritize A2)," the priorities (W, , ) to (
w,,) is w,,=w,! = 3 w , 3= w , ,: w , , = w , , = 1
is set in advance. That is, the priority of area (AI) (A2) for rule (1) with respect to other areas is set to three times. Note that this priority setting is based on experiments conducted in advance.

Regarding rule (2), in order to indicate "areas (AI) and (A3) are given priority" as the conclusion part, the priority of each area (W, ) to (W□) is W鵞,=W, , = 3 W Kotodama = w, 4 = w, , = w, , = 1 are set in advance.

Regarding rule (3), "give priority to area (A1)"
In order to show as a conclusion, the priority of each area (W,,)
〜(W3.) is w3. =3 W3亥” W s I= W s + = W s l
= W x * = 1 is set in advance.

Regarding rule (4), "area (A1), (A2),
In order to show "Prioritize (A3)" as the conclusion part, the priority of each area (w,,) to (W4゜) is W + + =
W t s = W 4 s = 3W + + =
It is preset that W port=W,,=1.

Regarding rule (5), "All areas have the same priority"
In order to show as a conclusion, the priority of each area (W,,)
〜(W,) is W s1 = Vl 51 = W S 3 = W g
4=W65=W5@”l is set in advance.

Regarding rule (6), "give priority to area (AI)"
In order to show as a conclusion, the priority of each area (W,,)
〜(was) is w,,=3 w,,:: w,,: w,,= w,,=
w , , : is preset as 1. Note that the simple average value (Z, ) is calculated by a simple average circuit (68) as described later.

Considering the priority (Wk) considering all rules using the priority of each area in each rule set in this way using the examples shown in Figures 6 to 11, for the area (AI),
Formula (A) becomes. In this formula (B), ``u ++'3+u **・3+u 3+'3''11
,! -3+u, 1・1+u,・3 “ut+”u! Goose+u 3++u +t”
Since u s++u st, the priority (Wl) of area (A) is W t=(3ut+"3uis+3us+"3tL*"
ui+”3ust)/(u++”uts”us++u4
O priority (W, ) to (W
6) is calculated similarly. The priority (Wk) of each area determined by fuzzy inference for all rules is sent to the weighting circuits (61) to (66). The weighting circuits (61) to (66) perform so-called priority processing by weighting based on the priorities (Wl) to (W,) for each area. That is, each brightness evaluation value (Yl) to (Y4) is multiplied by the priority of the corresponding area (W, ) to (W,) to yield Yi-Wi (i=1
~6) is calculated. All of the luminance evaluation values weighted in this way are supplied to the weighted averaging circuit (67). The weighting is performed by the average circuit (67), and the weighting is performed by the circuits (61) to (6).
6) Divide the added value of the output by the sum of the products of each priority and area, and output the weighted average value (Z,). That is, calculate. Note that Si (i=1 to 6) indicates the area of each region.

The simple averaging circuit (68) adds up all of the respective brightness evaluation values (Yi) and calculates the added value as the area (S) of the entire screen.

+S, +...S,) to obtain the simple average of the entire screen. Note that this simple average value (Zl) is calculated based on each brightness evaluation value (
Y i ) is weighted by circuits (61) to (66) with priorities (W, ) to (W,) all set to "1", and weighted by averaging circuit (67) using equation This is the same value as that calculated in C).

The simple average value (zl) and weighted average value (2,) calculated as described above are input to the divider (69), and m=zl
/zl is performed, and this division value (m) is input to the gain control circuit (70) and target level control circuit (71).

The gain control circuit (70) supplies a target level (P) to a comparator (5) that controls the gain of the variable gain amplifier (4). This target level (P) obtains the optimum exposure for the image capture screen when m=1, that is, when the simple average value (Zl) and the weighted average value Z,) are equal and the brightness distribution of the image capture screen is not considered. It follows the division value (m) which is a correction value so as to always satisfy P = m P .

Therefore, in the end, the weighting in exposure adjustment is the average value (Z,
) becomes the optimal target level (Po).
) will change.

The comparator (5) integrates the captured video signal over a sufficiently long time constant (for example, one field period) to obtain an integral value that indicates the brightness level of the corresponding field! (90) The output is compared with the target level (P), and this comparison output is supplied to the variable gain amplifier (4) so that the integrated output reaches the target level (P).
By controlling the gain so as to match , AGC is applied to the video signal in consideration of weighting and processing.

The target level control circuit (71) supplies a target level (Q) to a comparator (72) that controls the aperture amount of the aperture mechanism (2), and this target level (Q) is equal to the target level (
Similarly to P), when the division value (m) satisfies the condition of m=1, it is set to the optimal target level of Q = q o, and between it and the division value (m), Q = m q , As a result, the target level (Q) changes so that the average weighting value (Z,) always matches the optimal target level (qo) during exposure adjustment. .

The comparator (72) integrates the target level (Q)! (80
) output, and this comparison output is compared with the aperture mechanism (
2), and based on this comparison output, the aperture mechanism (2)
the aperture mechanism (2) so that the integral output indicating the brightness level of the relevant field matches the target level (Q).
The aperture amount is controlled. Note that the time constant of the integrator (80) is equal to that of the integrator (90), and is set so that the aperture mechanism (2) does not follow instantaneous changes in the captured video signal.

As mentioned above, the variable gain amplifier (4) and the aperture mechanism (2)
The target level (
P) (Q) is a variable gain amplifier (4) because the weighting that has been processed changes according to the average value (Z, ).
The weighting process is fully taken into consideration for the electrical exposure adjustment by the diaphragm III (2) and the optical exposure adjustment by the iris III (2). For example, if the simple average value (Zl) of the entire screen is "120", the average value ( Z,) is "100", sufficient brightness is obtained over the entire screen, but rules (1) to (6)
), if we focus only on the area that must be prioritized, sufficient brightness will not be obtained and the central area will be dark, and the division value (m) will be m = 1.2 Therefore, the target levels (P) and (Q) are respectively P=mP,,Q=
As a result, the variable gain amplifier (
The gain of step 4) also increases, the aperture amount of the aperture mechanism (2) also decreases, and optimal exposure adjustment is performed for the priority area.

Next, how the rules (1) to (6) affect exposure adjustment will be explained for each rule. Rules (1) to (4) are the basics of priority processing, and when the brightness evaluation values of areas (AI), (A2, and A3) are close to each other, they work to increase the priority of that area. do.

For example, as in the prior art, the areas (A1.) (A2) (A3) with the highest probability of containing the subject are simply given the same priority as the areas (A4) (A5) (A6), When the subject (S) is photographed in a backlit situation as shown in FIG. 5, a bright background such as the sun appears only in the area (A2), making it impossible to correct the subject (S) appropriately. Therefore, if rules (1) to (4) are applied, both areas (A1) and (A3) are dark and only area (A2) is bright, so the normalized brightness evaluation values (Vl) (V2) (V3) include V1 #V: lV2 holds, conditions (1) and (2) of rule (1), condition (2) of rule (3), and rule (4)
) is difficult to hold, so only the probability of rule (2) being satisfied is extremely high, and the priority of areas (A1) and (A3) becomes high, and the objects that fall within these areas (A1) and (A3) (S), and the optimum exposure condition for this subject (S) is obtained. These series of rules are particularly effective when backlit.

Rule (5) corresponds to the case where the entire screen is dark, and when the maximum value of the normalized brightness evaluation value is not large, priority processing is not performed and the average value of the screen is used as the representative value. In addition, as a rule when the entire screen is dark, the following rule (5)'
It is also possible to substitute rule (5).

[Rule (5)’] [If the aperture is wide open.

Then all areas have the same priority.'' This rule (5) is based on the aperture mechanism (2) that reduces the darkness of the image capture screen.
If the aperture is quite open, that is, the aperture is quite large, the imaging screen is assumed to be dark, and the priority is set the same for all areas, making unnecessary corrections possible. It works to suppress The membership function of rule (5) with this degree of openness as an input variable is illustrated in Figure 12, and the priority of each area is w,...
=w, , =w, , =w, , =w, , =w□=1.

At this time, to detect the opening degree of the aperture, the driving voltage value for operating the aperture mechanism (2) is set as shown in Fig. 14 (A/D converter (2)
00) and feeding it back to the priority determination circuit (57), or the aperture mechanism (2) is driven by a stepping motor whose rotor position can be counted, and this motor is opened. Various methods can be considered, such as making the degree of opening correspond to the number of steps in the direction and further providing a separate sensor to detect the degree of opening. In addition, wringer 111
The aperture amount (2) changes in inverse proportion to the voltage value of the comparator (72). That is, the degree of opening changes in proportion to the voltage value. Note that it is also possible to use both rules (5) and (5)' as rules corresponding to the case where the entire screen is dark.

Rule (6) corresponds to the case where there is a so-called abnormal brightness part, such as an extremely high brightness part such as a light source, in the screen, and since the abnormal brightness part is in one of the areas, normalization is Although the maximum value of the brightness evaluation value (Vi) is not small, the simple average value (Z
When l) is small and the entire area other than the area where the abnormal brightness portion is present is dark, the area (A1) is given priority unconditionally.

Therefore, if the photographer positions the abnormal brightness area within the area (A1) at the center of the screen in order to photograph the abnormal brightness area under a dark background, the normalized brightness evaluation value (V
The maximum value of i) is (Vl), which is a large value, but the simple average value (Z,) is small, and priority is given to the area (AI). Therefore, the exposure is adjusted so that the brightness level of area (A1) is at the optimal level, that is, the abnormal brightness part in area (A1) is in the optimal exposure state, and as a result, it is possible to photograph the abnormal brightness part. Become. At this time, the areas (A2) to (A6) become dark with extremely low brightness due to the above-mentioned exposure adjustment, but this is unavoidable since the highest priority is given to photographing the abnormally bright parts.

In addition, when the abnormal brightness area exists in any area other than the area (AI) and another object with lower brightness than the abnormal brightness area exists in the area (A1), a backlight condition will occur, and the normalized brightness evaluation value ( The maximum value of Vi) becomes the normalized brightness evaluation value of the area where the abnormal brightness part exists and is large, and because it is a backlight condition, the area other than the area where the abnormal brightness part exists becomes dark, and the simple average value (Z, ) becomes smaller, and in this case also the area (A
1) has a higher priority. Therefore, the exposure adjustment is performed so that the influence of the abnormal brightness portion is reduced and the main subject in the area (A1) is brought to the optimum exposure state.

Note that the division of regions and the setting of each rule are not limited to this embodiment, and various forms can be considered. In addition, the switching circuit in Figure 1 (
It goes without saying that the operations of the divider (26) to (69) can be processed by software using a microcomputer.

Further, in the above embodiment, the optimal target level (Po) of the imaging screen without considering the brightness distribution is set in advance, and the weighting is a division value that is the ratio of the simple average value (Z, ) to the average value (Z, ). An integrator (90) that multiplies the optimal target level (P, ) by (m) as a correction value to indicate the brightness level of the captured video signal.
Optimum exposure control is achieved by comparing with the output, but as shown in Figure 13, the optimal target level (Po)
The weighting is the average value (2,) and the target level memory (91)
The target level (P6°) stored in (Poo
is a value obtained by digitizing the target level (P, )) at a comparison tW (92), and based on this comparison result, the gain of the variable gain amplifier (4) is controlled, and the aperture mechanism (2 ) It is also possible to perform electrical and optical exposure adjustment by controlling the aperture amount. For example, with weighting, when the average value (2,) is smaller than the target level (P@'), it is assumed that the image capture screen is underexposed compared to the optimal exposure state after considering the brightness distribution, and Zt = P, ° The gain of the variable gain amplifier (4) is increased so that the aperture mechanism (
2) The aperture amount is reduced to increase the brightness, and conversely, the weighting is set to the average value (when Z is larger than the target level (peo), it is assumed that the exposure is overexposed compared to the optimal exposure state, and Zt=
The brightness can be reduced by lowering the gain of the variable gain amplifier (4) and increasing the aperture amount of the aperture mechanism (2) so that P, ゛.

(G) Effects of the Invention As described above, according to the present invention, the brightness distribution of the image capturing screen can be detected extremely easily, and the brightness of the image capturing screen does not change suddenly depending on the brightness distribution, resulting in smooth exposure. Adjustment is possible.

[Brief explanation of drawings]

1 to 12 relate to an embodiment of the present invention, in which FIG. 1 is an overall circuit block diagram, FIG. 2 is a flowchart, FIG. 3 is an explanatory diagram of screen division, and FIG. 4 is a main circuit. The block diagram, Figure 5 is a diagram showing an example of the imaging screen, and Figure 6 is the rule (
Figure 7 is an explanatory diagram of rule (2), Figure 8 is an explanatory diagram of rule (3), Figure 9 is an explanatory diagram of rule (4), and Figure 10 is an explanatory diagram of rule (5). FIG. 11 is an explanatory diagram of rule (6), FIG. 12 is an explanatory diagram of rule (5)', and FIGS. 13 and 14 are circuit block diagrams of other embodiments of the present invention. It is. (31) (32) (33) (34) (35) (36)・
... Integration circuit, (61) (62) (63) (64) (6
5) (66)... Weighting circuit, (70)... Gain control circuit, (71)... Target level control circuit, (57
)...priority determination circuit

Claims (1)

[Claims] (1) Brightness level detection means that divides an imaging screen into a plurality of regions and detects the brightness level of each region as a brightness evaluation value, and detects the brightness distribution of the imaging screen in any two regions of the plurality of regions. An automatic exposure adjustment device comprising: a brightness distribution detection means that uses fuzzy inference using a ratio of brightness evaluation values of as an input variable; and an exposure control means that performs exposure control according to the output of the brightness distribution detection means.