JP2517432B2 - Autofocus device - Google Patents

Description

The present invention relates to an autofocus device used in a video camera.

(B) Conventional technology In an autofocus device for a video camera, the method of using the image signal itself from the image sensor for evaluating the focus control state has essentially no parallax and has a depth of field. It has many advantages such as being able to focus accurately with a shallow subject or a distant subject. Moreover, a special sensor for autofocus is unnecessary, and the mechanism is extremely simple.

Japanese Patent Laid-Open No. 63-215268 (H04N 5/232) discloses an example of the autofocus device as described above.
The outline of this prior art will be described below with reference to FIGS. 2 and 3. FIG. 2 is an overall circuit block diagram of the autofocus circuit according to the conventional art. Lens (1)
The image formed by becomes an image signal by the image pickup circuit (4) including the image pickup element, and the luminance signal therein is input to the focus evaluation value generation circuit (5).

The focus evaluation value generating circuit (5) is configured as shown in FIG. 3, for example.

The luminance signal passes through a high-pass filter (HPF) (5c) to separate only high-frequency components, and amplitude detection is performed by a detection circuit (5d) at the next stage. This detection output is the A / D conversion circuit (5
The signal is converted to a digital value in e), only the focus area signal set in the center of the screen is extracted by the gate circuit (5f), integrated by each field by the integrating circuit (5g), and the focus evaluation of the current field is performed. Value is available. At this time,
The vertical and horizontal sync signals separated from the luminance signal by the sync separation circuit (5a) are input to the gate control circuit (5b) to set the focus area. In the gate control circuit (5b), a rectangular focus area is set in the center of the screen based on the vertical and horizontal sync signals and a fixed oscillator output, and a gate opening / closing signal that allows passage of the brightness signal only in this area range is set. Is supplied to the gate circuit (5c).

The focus evaluation value generation circuit (5) configured as described above always outputs the focus evaluation value for one field.

Immediately after starting the focusing operation, the first focus evaluation value is held in the maximum value memory (6) and the initial value memory (7). afterwards,
A focus motor control circuit (10) rotates a focus motor (3), which is a stepping motor, in a predetermined direction to rotate a focus ring (2) supporting the light receiving lens (1) to receive the light receiving lens. Displace (1) in the optical axis direction and monitor the output of the second comparator (9). Second
The comparator (9) compares the focus evaluation value after driving the focus motor with the initial evaluation value stored in the initial value memory (7), and outputs the magnitude.

The focus motor control circuit (10) controls the focus motor drive circuit (31) to rotate the focus motor (3) in the first direction until the second comparator (9) outputs a large or small output. , When the output that the current focus evaluation value is larger than the initial evaluation value is output, the rotation direction is maintained as it is, and when it is determined that the current evaluation value is smaller than the initial evaluation value, The rotation direction of the focus motor is reversed, and thereafter the output of the first comparator (8) is monitored.

The first comparator (8) compares the current maximum focus evaluation value held in the maximum value memory (6) with the current evaluation value, and the current focus evaluation value is stored in the maximum value memory (6). Two comparison signals (S1) that are larger than the content (first mode) and have decreased below a preset first threshold value (second mode)
Output (S2). Here, the maximum value memory (6) is updated based on the output of the first comparator (8) when the current evaluation value is larger than the content of the maximum value memory (6), and the value is always updated. The maximum focus evaluation value up to is retained.

(13) indicates the position of the focus ring (2) supporting the lens (1), receives the focus ring position signal detected by the motor position detection circuit (30), and stores the focus ring position as the lens position. Similar to the maximum value memory (6), the position memory is updated based on the output of the first comparator (8) so as to always hold the focus ring position when the maximum evaluation value is reached.

The motor position detection circuit (30) is
The focus ring consists of an up-down counter that is reset at the start of the focusing operation and counts up or down by setting the step amount of the focus motor (3), which is a stepping motor, as positive in the near point direction and negative in the far point direction. The position signal has this count value.

The focus motor control circuit (10) monitors the output of the first comparator (8) while rotating the focus motor (3) in the direction determined based on the output of the second comparator (9), and the focus evaluation value The focus motor (3) is reversely rotated at the same time when the second mode is instructed such that is smaller than the first evaluation value (M) set in advance as compared with the maximum evaluation value.

Due to the reverse rotation of the focus motor (3), the moving direction of the light receiving lens (1) changes, for example, from a direction approaching the image sensor to a direction away from the image sensor, or vice versa.

After this reverse rotation, the contents of the position memory (13) and the current ring position are compared by the third comparator (14) and when they match, that is, the focus ring (2), that is, the lens (1), the focus evaluation value. The focus motor control circuit (10) functions to stop the focus motor (3) when the position returns to the maximum position. At the same time, the focus motor control circuit (10) outputs a lens stop signal (LS).

The change in the lens position in the so-called hill climbing focusing operation described above is shown in FIG.

(11) is a fourth memory for holding the focus evaluation value at that time at the same time as the lens stop signal (LS) is issued after the focusing operation by the focus motor control circuit (10) is finished. The content held in the fourth memory (11) is compared with the current focus evaluation value by the fourth comparator (12), and the current focus evaluation value is preset in comparison with the content of the fourth memory (11). When it becomes smaller than the second threshold value, it is determined that the subject has changed, and the subject change signal is output.
Upon receiving this signal, the focus motor control circuit (10) repeats the hill-climbing focus operation to follow the change of the subject.

This autofocus system is excellent in focusing accuracy and adaptability to a wide range of subjects, but is weak for subjects such as a wall without a pattern, in which a high-frequency component is unlikely to occur in the luminance signal even during focusing. .

That is, during such a focusing operation of the subject, the evaluation value does not drop from the maximum value of the obtained evaluation values to the first threshold value or more, and the lens position having the maximum value is in focus forever. The position cannot be determined, the lens continues to be displaced in the entire area between the infinity point and the near point, and the motor cannot be stopped.

In order to prevent such a situation from occurring, even if the evaluation value does not drop from the maximum value to the first threshold value or more,
If the lens scans the entire area between the infinity point and the near point once,
It is possible to stop the motor by unconditionally returning the lens to the position where it takes the maximum value or returning it to the initial position where the focusing operation started. In this case, the motor stoppage is emphasized rather than the reliability of the focusing operation itself, and the probability that the final lens position is in focus is significantly lower than that of a normal subject with sufficiently high contrast,
This is unavoidable.

(C) Problems to be Solved by the Invention As described above, when shooting a subject in which a high-frequency component is unlikely to occur in the luminance signal even during focusing, the distance is far from the lens stop position after the focusing operation is completed. Even if a subject having a sufficiently large contrast enters the focus area, the blur is too large and the evaluation value does not change enough to exceed the second threshold, and the focusing operation for a new subject is not restarted. The problem arises.

(D) Means for Solving the Problems The present invention takes out the amount of high-frequency components of a luminance signal in a region set on the screen as a focus evaluation value, and performs imaging so that the focus evaluation value becomes maximum. An autofocus device that performs a focusing operation by driving a system, detects a brightness difference in the area, and restarts the focusing operation when a predetermined change occurs in the brightness difference after the focusing operation ends. It is characterized by Furthermore, it is characterized in that fuzzy inference is used to judge the change of the brightness difference.

(E) Operation Since the present invention is configured as described above, not only the focus evaluation value but also the brightness difference is used to determine whether or not to restart the focusing operation. It is possible to quickly find a change in the above and restart the focusing operation.

(F) Embodiments Embodiments of the present invention will be described below with reference to the drawings.

FIG. 1 is a circuit block diagram of this embodiment. The first
In the figure, the same parts as those in FIGS. 2 and 3 of the prior art are designated by the same reference numerals and the description thereof will be omitted.

The luminance signal in the video signal output from the image pickup circuit (4) is input to the focus evaluation value generation circuit (50).

As shown in Fig. 5, the focus evaluation value generation circuit (50) has two types of HPF with different cutoff frequencies, 200KHz and 600KHz.
(5c) (50c), switching circuit (5h) that selects and outputs these HPF (50c) outputs for each field, switching circuit (5h)
A detection circuit (5d) that amplitude-detects the output, an A / D converter (5e) that converts this detection output to a digital value, a sync separation circuit (5a) that separates the sync signal from the luminance signal, and the separated sync signal and Based on the fixed oscillator output, as shown in FIG. 6, only the passage of the A / D converted value of the luminance signal portion corresponding to the relatively small first area (A) set in the center of the screen is set. The gate circuit (5
The second gate opening / closing signal that includes only the first area (A) and allows only the A / D converted value of the luminance signal portion in the second area (B) having a larger area than this area to be included in f). Gate control circuit (50f)
b), The digital values of the high frequency components of the luminance signal in each area output from the gate circuits (5f) (50f) are added over one field, and as a result, digital integration is performed, and this integrated value is calculated for one field. It is composed of integrating circuits (5g) and (50g) which output the first and second focus evaluation values (Va) (Vb) respectively and are reset for each field.

Here, since the switching circuit (5h) alternately selects one of the HPF (5c) and (50c) outputs for each field by the output of the sync separation circuit (5a), the field where HPF (5c) is selected Then, from the integration circuit (5g), the digital integrated value (V) of one field of the high frequency component of 200KHz or higher which is the cutoff frequency of the HPF (5c) of the luminance signal in the first area (A)
1) is output as the first focus evaluation value (Va), and at the same time, from the integrating circuit (50g), the digital integration value (V1 for one field of the high frequency component of 200KHz or higher of the luminance signal in the second area (B) is also generated. ′) Is output as the second focus evaluation value (Vb). In the next field, the first focus evaluation value (Va)
Is the digital integration value (V2) for one field of the high frequency component above 600 KHz which is the cutoff frequency of the HPF (50c) of the luminance signal in the first area (A), the second focus evaluation value (Vb)
Becomes the digital integrated value (V2 ') of the high frequency component of 600 KHz or higher, and the same repetition is performed thereafter.

The first and second focus evaluation values (Va) thus obtained
(Vb) is input to the separation circuits (51) and (52), respectively. The separation circuits (51) and (52) both switch the switching operation of the switching circuit (5h) on a field-by-field basis.
The separation timing is controlled by the switching control signal (SC) that is generated from a) and is created based on the vertical synchronization signal, and the separation circuit (51) integrates the first focus evaluation value (Va) using each HPF. Separated into values (V1) (V1 ') and evaluate each (V1)
Output as (V1 '). Similarly, the separation circuit (52) uses the second focus evaluation value (Vb) as the integral value (V2) (V
2 ') and output each as an evaluation value (V2) (V2'). Therefore, the evaluation value (V1) (V1 ') (V2) (V2')
Will be updated every two fields.

Next, the evaluation values (V1) and (V2) are respectively stored in the initial value memory (7).
(57), subtraction circuit (70) (80), memory (24) (58),
It is input to the relative ratio calculation circuits (25) (59) and the switching circuit (20). On the other hand, the evaluation values (V1 ') (V2') are input to the relative ratio calculation circuits (25) (59), respectively.

In order to start the focusing operation, the initial value memory (7) (57) first evaluates the evaluation value (V1) at the time of starting the rotation of the focus motor (3) in a predetermined initial direction.
While the motor (3) rotates in the initial direction and the lens (1) is displaced, the evaluation value (V1) is obtained when two fields have elapsed after the motor was started.
Since (V2) is updated, the subtraction circuit (70) (80) in the subsequent stage
Then, the value obtained by subtracting the contents of the initial value memory (7) (57) from the new evaluation value is used as the change amount (ΔV1) (ΔV2) of the evaluation value (V1) (V2), and the focus motor control circuit Output to (100).

The memories (24) and (58) hold the evaluation values (V1) and (V2) obtained after the motor (3) is started for two fields and input them to the relative ratio calculation circuits (25) and (59) in the subsequent stage. .

The relative ratio calculation circuit (25) (59) is composed of a divider (61), a memory (62) and a subtractor (63) as shown in FIG. Every time (V1 ') (V2') is updated, the ratio with the latest evaluation value (V1) (V2) held in the memory (24) (58), V1 '/ V1, V2' / V2 Is calculated as a relative ratio (r1) (r2).

Here, the relative ratio (r1) is 1 when HPF (5c) is used.
It is the ratio of the integrated value (V1) for the field and the integrated value (V1 ') when using the HPF (50c). Both integrated values and the focus ring position (lens position) when the same subject is used. The relationship is as shown in Fig. 7. That is, the cutoff frequency is high
The integrated value at HPF (50c) has a steep peak, and the integrated value at HPF (5c) with a low cutoff frequency has a gentle peak.
Therefore, if the relationship between the relative ratio and the degree of blurring of the subject (the amount of displacement or the amount of displacement from the lens position when focusing) is shown in a graph, a monotonically decreasing characteristic curve as shown in FIG. 8 is obtained.

This is a kind of normalized state quantity because the state quantity of the relative ratio is a function value that can express the in-focus state (degree of blur) of the subject in the same way as the focus evaluation value, and is expressed as a ratio. Therefore, it has the property of being less susceptible to the environment in which the subject is placed. For example, when the illuminance of the subject changes, the absolute value of the focus evaluation value changes,
There is no significant change in the relative ratio. Usually, since the above-mentioned property is independent of the type of subject, it is possible to use this relative ratio as a parameter of the degree of blur.
When the monotonically decreasing characteristic curve in FIG. 8 is made to correspond to the lens position and the focus ring position immediately, the focus position is changed to a near-point and ∞ point side in a substantially linear manner as shown by the alternate long and short dash line in FIG. A characteristic diagram can be obtained.

The relative ratios (r1) (r2) obtained by the divider (61) are input to the memory (62) and the subtractor (63). Memory (6
2) holds the input relative ratio for two fields, delays it, and supplies it to the subtractor (63).
In 3), the previous relative ratio stored in the memory (62), that is, the relative ratio of two fields before is subtracted from the latest relative ratio obtained by the divider (61), and this subtracted value is used as the relative ratio (r2) ( r2)
Is output to the focus motor control circuit (100) as a change amount (Δr1) (Δr2). This change (Δr1) (Δr
It goes without saying that 2) can take a negative value when the initial rotation direction of the motor is opposite to the focusing direction.

The switching circuit (20) is a focus motor control circuit (10
Area selection signal (Sa) output from (0) selects the area to be used for focusing operation and the both focus evaluation value (V1)
Select one of (V2).

The focus evaluation value selected by the switching circuit (20) is supplied to the maximum value memory (6) and the first comparator (8), and a signal for executing hill climbing control is supplied to the focus lens as in the conventional example. The position memory (1
3) and the third comparator (14) perform exactly the same operation as in FIG. 2 of the conventional example.

However, the focus evaluation values obtained from the switching circuit (20) are all updated every two fields, and the comparison operation of the first comparator (8) for the hill climbing operation is performed every two fields. It

The zoom position detection section (64) sends a signal indicating the current focal length (Z) of the lens according to the wide-angle to telephoto zoom area of the zoom mechanism having a known zoom lens mounted on the video camera. Output to the control circuit (100).

The change amount detection circuit (81) has a function of detecting a change amount after the end of the focusing operation with respect to the evaluation values (V1) and (V2), and specifically, as shown in FIG. 5 memory (8
2) (83), subtractor (84) (85), and selection circuit (86)
It has. The fourth memory (82) stores, as (W1), the evaluation value (V1) when the lens stop signal (LS) is issued from the focus motor control circuit (100), that is, immediately after the end of the focusing operation. Similarly, the fifth memory (82) stores the evaluation value (V2) when the lens stop signal (LS) is issued as (W2). The subtractor (84) receives the evaluation value (V1) and the retained data (W1) of the fourth memory (82) as input, and subtracts the subtraction value of V1-W1 from the evaluation value (V1) immediately after the end of the focusing operation. Amount (Δ
It is output to the selection circuit (86) in the subsequent stage as W1). Similarly, the subtractor (85) uses the evaluation value (V2) and the fourth memory (83).
The holding data (W2) is input, and the subtraction value of V2-W2 is output to the selection circuit (86) in the subsequent stage as the change amount (ΔW2) of the evaluation value (V2) immediately after the end of the focusing operation. The evaluation value (V
1) Since (V2) is updated every two fields even after the motor is stopped, the change amounts (ΔW1) (ΔW2) also change every two fields. The selection circuit (86) changes the amount (ΔWk) (k = 1 or 2) according to the area switching signal (Sa) sent from the focus motor control circuit (100) to the switching circuit (20).
One of the two is selected and output to the focus motor control circuit (100). Therefore, if the first area (A) is selected as the focus area during the focusing operation, the change amount (ΔW1) is changed, and if the second area (B) is selected, the change amount (ΔW2) is changed. Is selected.

The average brightness detection circuit (65) is, as shown in FIG. 11, a detection circuit (65a), an A / D conversion circuit (65b) and an integration circuit (65c).
The amplitude signal of the luminance signal corresponding to the entire screen is detected by the detection circuit (65a), and the detected output is A / D converted by the A / D conversion circuit (65b). ), All the A / D conversion data for one field are integrated and digitally integrated, and the integrated value is the average luminance (IR
It outputs to the control circuit (100) as S).

The contrast detection circuit (66) is configured as shown in FIG. 13, and has a horizontal contrast in the first area (A).
That is, the brightness difference is detected. Here, the operation of the contrast detection circuit (66) will be described. First,
In order to detect the contrast, the first
The area (A) is subdivided into 16 small areas (Nij) (i, j = 1 to 4) of the same area, which are arranged in 4x4 vertically and horizontally, and the luminance signal is separated by the separation circuit (67). Is divided into each small area, and 16 digital integrators (Ki
j) (i, j = 1 to 4).

Each of the digital integrators has exactly the same structure as the average brightness detection circuit (65) shown in FIG. 11, and the brightness signals corresponding to each small area are digitally integrated to obtain an integrated value (Fij). Is derived. These integrated values (Fij) are
The four small areas arranged in the horizontal direction are input as one set to the Max-Min calculation circuit (Li) (i = 1 to 4). That is, the integrated value (Fij) (j = 1 to 4) is stored in the Max-Min calculation circuit (L1), and similarly, the integrated values (F2j) (F3j) (F4j) are respectively stored in Max-M.
Input to the calculation circuit (L2) (L3) (L4).

The Max-Min calculation circuit (Li) selects the maximum value and the minimum value among the four integrated values of the small areas arranged in the horizontal direction, performs the subtraction of (maximum value-minimum value), and performs this subtraction. The value (Gi) is output to the maximum value detection circuit (68) in the subsequent stage.

The maximum value detection circuit (68) functions to take out the maximum value among the subtraction values (Gi) for four rows, and outputs this maximum value as the contrast (ΔE1). Therefore, after all, the contrast (ΔE1) corresponds to the digital value of the luminance difference itself of the row having the largest horizontal luminance difference in the first area (A) for one field.

The contrast (ΔE1) thus obtained is supplied to the change amount detection circuit (87), and the change amount of the contrast from the contrast immediately after the end of the focusing operation is calculated. That is,
As shown in Fig. 23, the lens stop signal (L
The value of the contrast (ΔE1) when (S) is emitted is (E
E) and contrast (ΔE1) obtained for each field thereafter, Δ (ΔE1) = ΔE1−EE
Is calculated, and this Δ (ΔE1) is output to the control circuit (100) as a change amount of the brightness contrast in the first area (A) from the end of the focusing operation.

The focus motor control circuit (100) executes a hill-climbing focusing operation based on the outputs of the first comparator (8) and the third comparator (14) as in the conventional example. In addition, the output from the subtraction circuits (70) (80), that is, the change amount (ΔV1) (ΔV2) from the initial value of the evaluation values (V1) (V2), the relative ratio calculation circuits (25) (59) Output, that is, the evaluation value (V1) for the first area (A), the change amount (Δr1) (Δr2) from the initial value of the relative ratio (r1) (r2) of each area, and the contrast detection circuit (66) Output, that is, the direction of the in-focus point is determined by fuzzy inference based on five kinds of data of the brightness contrast (ΔE1) in the first area (A), and the evaluation value (from the separation circuit (51) ( V1), the output from the average brightness detection circuit (65), that is, the average brightness of the entire screen (IR
Area selection is executed by fuzzy inference based on four types of data: S), the output from the zoom position detection circuit (64), that is, the current focal length (Z) and contrast (ΔE1). Furthermore, the amount of change in contrast (Δ (ΔE
1)), the focal length (Z), and the amount of change (Wk) in the focus evaluation value after focusing on the focus area (where k = 1 or 2), based on three types of data, confirmation of the subject change and The focus motor restart decision is executed by fuzzy inference.

Next, the processing for direction control, area selection, and restart decision by the above fuzzy inference will be described in detail.

First, the direction determination process is executed according to the flowchart shown in FIG. 20, and the fuzzy reasoning used at this time is to evaluate the evaluation value (V1), the change amount (ΔV1) (ΔV2), the change amount (Δr1) and the contrast (ΔE1). As a variable, a value of 0 to 1 is taken as the conclusion part, and the parameter (di) is set so that the direction of the in-focus point becomes the current traveling direction of the lens when it becomes larger and the opposite direction when it becomes smaller.

Here, the rule is [rule (1)] "if ifV1 is large and ∆r1 is large then d1 = 1.0" [rule (2)] "if deltaV1 is large and ∆r1 is not then d2 = 0.7" [rule (3)] "if V1 is small and ΔV2 is small and ΔE1 is small then
d3 = 0.2 ”is set.

 Next, each rule will be described.

[Rule (1)] is defined by a membership function as shown in FIGS. 14 (a) and 14 (b). Figure 14 (a) shows "Δ
The membership value of condition (1) of the rule (1) "V1 is large" is a membership function for the change amount (ΔV1) that is an input variable. As the change amount (ΔV1) increases, the membership value ( u11) is a function including a monotonically increasing straight line, and from this function the membership value (u11) according to the amount of change (ΔV1) can be obtained.

FIG. 14 (b) shows the membership value of the rule (condition (2) of 1) that "Δr1 is large", which is a membership function for the input variable change (Δr1), and the change (Δr1) is It is a function including a monotonically increasing straight line in which the membership value (u12) increases as it increases.
From this function, the membership value (u12) corresponding to the change (Δr1) can be obtained.

The rule (1) takes into consideration the case where both the focus evaluation value and the relative ratio in the first area (A) increase, and in this case, there is a focal point in the current movement direction of the lens (1). Since there is a high possibility, select the first area (A),
The conclusion part (d1) is set as d1 = 1.

[Rule (2)] is defined by the membership function as shown in FIGS. 15 (a) and 15 (b). Figure 15 (a) shows "Δ
V1 is large ”is the membership function of the condition (1) of the rule (2), which is a membership function for the input variable amount (ΔV1), and the membership value (u21) is obtained from this.

FIG. 15 (b) shows the membership value of the condition (2) of the rule (2) that “Δr1 is not large”, and the membership value (u2
2) is a function that includes a monotonically decreasing straight line, and from this function, the membership value (u2
2) is obtained.

Rule (2) is when the focus evaluation value and the relative ratio change tendency are different, and there is a possibility that the in-focus point is in the direction opposite to the current lens movement direction, and the conclusion section (d2) is slightly smaller than d1. It is set to d2 = 0.7.

[Rule (3)] is defined by the membership function as shown in FIGS. 16 (a), (b) and (c). Figure 16 (a)
Is a function including a monotonically decreasing straight line in which the membership value (u31) decreases as the evaluation value (V1) increases, indicating the membership value of the condition (1) of the rule (3) that “V1 is small”. From this function, the membership value (u31) according to the evaluation value (V1) can be obtained.

Figure 16 (b) shows the rule that "Δr2 is small" (3)
The membership value of condition (2) of
It is a function that includes a monotonically decreasing straight line where the membership value (u32) decreases as 2) increases. From this function, the membership value (u32) according to the variation (ΔV2) can be obtained.

Figure 16 (c) shows the rule that "ΔE1 is small" (3)
Shows the membership value of condition (3), and as the contrast (ΔE1) increases, the membership value (u3
3) It is a function that includes a monotonically decreasing straight line that becomes smaller. From this function, the membership value (u33) according to the contrast (ΔE1) can be obtained.

Rule (3) considers the case where the focus evaluation value in the first area (A) is low and the change cannot be taken out. At this time, if the contrast of the first area (A) is small, the first area (A) is matched. If the subject to be focused does not exist, the direction is determined by the change of the focus evaluation value of the second area (B), and the change amount of the focus evaluation value in the second area (B) is negative. Further, assuming that the focal point is likely to be in the direction opposite to the current movement direction of the lens, the conclusion part (d3) is set as low as d3 = 0.2 in order to easily reverse the movement direction of the lens.

A direction determining process of calculating the direction determining parameter (D) from each of the above rules and finally determining the direction from this parameter will be described with reference to the flowchart of FIG.

When each membership value (uij) (i, j: integer) is calculated according to each input variable in STEP (100), then STEP (101)
At, the smallest membership value for each rule is calculated as the success rate (Ui) of each rule. For example, regarding rule (1), since u11> u12 according to FIG. 14, the degree of success (U1) is U1 = u12, and rule (2)
As shown in Fig. 15, since u21> u22, the degree of success (U2) is U2 = u22. Further, regarding rule (3), since u31 <u33 <u32 is obtained from FIG. 15, the degree of success (U3) is U
3 = u31.

Based on the degree of success (Ui) of each rule obtained in this way, ST
In EP (102), the parameter (D) for direction determination is Is calculated by This formula (1) means that the weighted average of each conclusion part is calculated according to the degree of satisfaction of each rule.

In step (103), the direction is determined by the obtained parameter (D). Specifically, if D ≧ 0.5, the current lens movement direction has the in-focus position, so the motor (3 ), The control command for controlling the drive is issued, and D <
If 0.5, the focus position is in the direction opposite to the current direction, so a control command for reversing the motor (3) is immediately issued to the focus motor drive circuit (31). In this way, 5
Taking into consideration one of the two factors, the direction of the in-focus point is determined with high accuracy, and the hill-climbing focusing operation described above is executed while rotating the motor (3) in this determination direction to move the lens.

Next, the area selection process is executed according to the flowchart shown in FIG. 21, and the fuzzy inference used at this time is
Focus evaluation value (V1), contrast (ΔE1), focal length (Z) and average brightness (IRS) are used as input variables, and a value of 0 to 1 is taken as the conclusion part, and when it is small, the first area (A), large area Then, the membership value (ai) is set, which means that the second area (B) is used.

Here, the rule is [rule (4)] "if (V1) is large then a4 = 0.0" [rule (5)] "if (V1) is medium and and (IRS) is small then a5 = 0.
8 ”[Rule (6)]“ If (V1) is small and (ΔE1) is large and (Z) is not small then a6 = 0.3 ”Next, each of the above rules will be described.

[Rule (4)] is defined by the membership function as shown in FIG. 17, and this FIG. 17 shows the membership value of the condition (1) of the rule (4) that “V1 is large” and the focus evaluation value. It is a membership function for (V1). As the evaluation value (V1) increases, the membership value (u4
1) is a function that includes a monotonically increasing straight line, from which the membership value (u4) according to the evaluation value (V1)
1) is obtained.

In the rule (4), when the focus evaluation value (V1) is large, the area (A) is prioritized assuming that the subject exists in the first area (A). Therefore, the conclusion part (a4) is a4 = 0. .
Set to 0.

[Rule (5)] is defined by the membership function as shown in FIGS. 18 (a) and 18 (b). Figure 18 (a) shows condition (1) of rule (5) "(V1) is medium".
Is a mountain-shaped function that shows the membership value of, and has the maximum evaluation value (V1) and the membership value (u51) according to the evaluation value (V1).
Is found.

FIG. 18 (b) shows the membership value of the condition (2) of the rule (5) that "(IRS) is small", and the membership value (u5 increases as the average brightness (IRS) increases.
2) is a function that includes a monotonically decreasing straight line, and the membership value (u
52) is obtained.

Rule (5) considers the case where the focus evaluation value (V1) in the first area (A) is not too large and the average brightness of the screen is dark, and at this time the S / N ratio of the video signal is poor and the focus evaluation is performed. Since the reliability of the value is low, the conclusion part (a5) is set to a5 = 0.8 so that the second area (B) is easily prioritized as the focus area in order to capture more information.

[Rule (6)] is defined by a membership function as shown in FIGS. 19 (a), (b) and (c), and FIG. 19 (a) shows the condition of rule (6) that “(V1) is small”. It is a function including the monotonically decreasing straight line showing the membership value of (1), and the membership value decreases as the evaluation value (V1) increases. From this function, the membership value (u61) corresponding to the evaluation value (V1) ) Is required.

FIG. 19 (b) shows the membership value of the condition (2) of the rule (6) that “(ΔE1) is large”, and the membership value (u62) increases as the contrast (ΔE1) increases. Is a function containing a straight line,
From this function, the membership value (u62) corresponding to the contrast (ΔE1) can be obtained.

FIG. 19 (c) shows the membership value of the condition (3) of the rule (6) that "(Z) is not small". The membership value (u6 increases as the focal length (Z) increases.
3) is a function that includes a monotonically increasing straight line, and from this function the membership value (u6
3) is obtained.

Rule (6) considers the case where the focus evaluation value of the first area (A) is small, but the contrast in the first area (A) is high, and the case where the focal length is short and the depth of field is deep is considered. Except for this, the conclusion part (a6) is set to a6 = 0.3 so that the first area (A) is relatively prioritized as being blurred or having some subject.

An area selection process of calculating the area selection parameter (Y) from the above-mentioned rules and finally selecting the area from this parameter will be described with reference to the flowchart of FIG.

At STEP (200), the condition (j) of rule (i) (i = 4 to 6, j = 1 or 2or) is calculated from the value of the input variable and the membership function.
When the membership value (uij) in 3) is calculated, next ST
In EP (201), the minimum membership value for each rule is calculated as the satisfaction rate (Ui) for each rule. For example, for rule (4), the degree of success (U4) is U4
= U41, and for rule (5) from Fig. 18, u51
Since <u52, the degree of success (U5) is U5 = u51, and further, regarding the rule (6), u61 <u62 <u63 from FIG. 19, so the degree of success (U6) is U6 = u61.

Based on the degree of success (Ui) of each rule obtained in this way, ST
Parameter (Y) for area selection in EP (202) Is calculated by This formula means that the weighted average of each conclusion part is calculated according to the degree of satisfaction of each rule. In STEP (203), the area is selected by the obtained parameter (Y). Specifically, if Y ≧ 0.5, the second area (B) is selected as the focus area, and if Y <0.5, the first area (A )
Is selected.

In this way, a highly accurate area selection is made in consideration of four factors, and the focus evaluation by the switching circuit (20) is performed in order to execute the hill-climbing focus operation with the focus evaluation value corresponding to this. A value selection is made.

The area switching process is also executed during the focusing operation and the monitoring operation of the subject change after reaching the focusing point. In addition, immediately after the actual switching of the area, the switching circuit (20)
The focus evaluation value obtained from Eq. (1) causes a temporary large fluctuation. Therefore, in order to prevent malfunction due to this fluctuation, the same comparison result is continuously detected in the comparison operation of the first comparator (8) during the hill climbing focus operation. The comparison result is output as valid only when it is obtained.

The restart determination process is executed according to the flowchart shown in FIG. 24. The fuzzy inference used at this time is
The amount of change (Δ (ΔE1)), the amount of change (ΔWk) and the focal length (Z) are used as input variables, a value of 0 to 1 is taken as the conclusion part, and restart is performed when there is a change in the screen when it becomes large. , It is a parameter (hi) that maintains the stopped state when it becomes smaller.

Here, the rule is [Rule (7)] "if Δ (ΔE1) is large in the positive direction and Z is not small the
n h7 = 0.8 ”[rule (8)]“ if Δ (ΔE1) is not large in the positive direction and ΔWk of the focus area is not small then h8 = 0.8 ”[rule (9)]“ If absolute value of if Δ (ΔE1) is Not large and ΔWk of focus area is small then h9 = 0.0 ”[Rule (10)]“ If Δ (ΔE1) is large in the positive direction and ΔWk of focus area is not small then h10 = 1.0 ”is set.

 Next, each rule will be described.

[Rule (7)] is defined by the membership function as shown in FIGS. 25 (a) and 25 (b). Figure 25 (a) shows "Δ
(ΔE1) is larger in the positive direction ”, which is a membership function for the change amount (Δ (ΔE1)) that is the input variable and indicates the membership value of the condition (1) of the rule (7), and the change amount (Δ (ΔE1) This is a function that includes a monotonically increasing straight line in which the membership value (u71) increases as ΔE1)) increases in the positive direction. From this function, the membership value (u71) corresponding to the change amount (Δ (ΔE1)) can be obtained. .

FIG. 25 (b) shows the membership value of the condition (2) of the rule (7) that "Z is not small", which is the membership function for the input variable focal length (Z). ) Becomes larger (in the TELE direction)
It is a function that includes a monotonically increasing straight line whose membership value (u72) increases as the focal length (Z) increases.
The membership value (u72) according to is calculated.

According to the rule (7), when the contrast of the brightness of the first area (A) is larger than that at the time of focusing, and the focal length is long and the depth of focus is shallow, the subject may have entered the first area (A). The conclusion part (h7) is set to h7 = 0.8 so that it is very high and restarting is likely to occur.

[Rule (8)] is defined by the membership function as shown in FIGS. 26 (a) and 26 (b). Figure 26 (a) shows "Δ
(ΔE1) is not large in the positive direction. ”It shows the membership value of condition (1) of rule (8) and is the membership function for the input variable change amount (Δ (ΔE1)). The value (u81) is obtained.

FIG. 26 (b) shows the membership value of the condition (2) of the rule (8) that "the ΔWk of the focus area is not small" and is selected as the focus area of the first or second areas which are the input variables. Membership value (u82) increases as the absolute value of the change amount (ΔWk) (k = 1 or 2) increases, and the member corresponding to the change amount (ΔWk) is greater than this function. Ship value (u82) is calculated.

Rule (8) considers the case where the contrast of the brightness of the first area does not increase, but there is a change in the focus evaluation value of the focus area. In this case, pay attention to the change of the evaluation value. , The conclusion section (h8) is set to a slightly higher value h8 = 0.8 so that a change in the subject is likely to occur and restarting is likely to occur.

[Rule (9)] is defined by the membership function as shown in FIGS. 27 (a) and 27 (b). Figure 27 (a) shows "Δ
The absolute value of (ΔE1) is not large ”, the membership value of condition (1) of rule (9) is shown, and the amount of change (Δ (Δ
Membership value (u91) as it approaches zero in E1))
Is a mountain-shaped function that increases, and the membership value (u91) corresponding to the amount of change (Δ (ΔE1)) can be obtained from this function.

FIG. 27 (b) shows that "ΔWk in the focus area is small".
Indicating the membership value of the condition (2) of the rule (9), the change amount (Δ) of the one selected as the focus area of the first or second area which is the input variable.
This is a mountain-shaped function in which the membership value (u92) increases as the absolute value of Wk) (k = 1 or 2) decreases. From this function, the membership value (u92) according to the amount of change (ΔWk) is obtained. The membership value (u92) corresponding to the amount of change (ΔWk) can be obtained from this function.

The rule (9) takes into consideration the case where both the brightness contrast of the first area (A) and the focus evaluation value of the focus area have not changed much from the time of focusing, and in this case, change to the subject. Since it is unlikely that the occurrence of the error occurs, the conclusion part (h9) is set to h9 = 0.0 so as to easily hold the stop operation.

[Rule (10)] is defined by the membership function as shown in FIGS. 28 (a) and 28 (b). Figure 28 (a) shows "Δ
(ΔE1) is larger in the positive direction ”is the membership function for condition (1) of rule (10), which is the membership function for the change amount (Δ (ΔE1)) that is the input variable. (U101) is obtained.

FIG. 28 (b) shows the membership value of the condition (2) of the rule (10) that "ΔWk of the focus area is not small", and is selected as the focus area of the first or second areas which are the input variables. As the absolute value of the change amount (ΔWk) (k = 1 or 2) of the
This is a valley-shaped function in which the membership value (u102) increases, and the membership value (u102) corresponding to the amount of change (ΔWk) is obtained from this function.

Rule (10) considers the case where the brightness contrast of the first area is increased and the focus evaluation value of the focus area is changed. In this case, the contrast is higher than that of rule (8). The probability that a change in the subject will occur as much as is increased, and the conclusion part (h10) is set to a slightly higher value h10 = 1.0 so that it is easy to restart even if the subject changes almost certainly ing.

The process of calculating the restart decision parameter (H) from the above-mentioned rules and finally deciding whether or not to restart according to this parameter will be described with reference to the flowchart of FIG.

When each membership value (uij) (i, j: integer) is calculated according to each input variable in STEP (300), then STEP (301)
At, the smallest membership value for each rule is calculated as the success rate (Ui) of each rule. For example, regarding rule (7), since u71 <u72 according to FIG. 25, the degree of success (U7) is U7 = u72, and rule (8)
From Fig. 26, since u81> u82, the degree of success (U8) is U8 = u82, and rule (9) is 27
From the figure, u91 <u92, so the degree of success (U9) is U9 = u91, and regarding rule (10) from Figure 28, u101>
Since u102, the degree of success (U10) is U10 = u102.

Based on the degree of success (Ui) of each rule obtained in this way, ST
Parameter (H) for determining restart in EP (302) Is calculated by This expression (3) means that each conclusion part is weighted and averaged according to the degree of satisfaction of each rule.

At STEP (303), it is determined whether or not to restart based on the obtained parameter (H). Specifically, if H ≧ 0.5, it is determined that the screen has changed, and the initial value memory (7) (57),
The maximum value memory (6), position memory (13), etc. are reset, and the control circuit (100) issues various control commands in order to redo a series of hill climbing focusing operations from the beginning. If H <0.5, the subject changes. As a result, the motor (3) is kept stopped. In this way, after taking into consideration the three factors, the restart is determined with high accuracy, and the focusing operation that quickly follows the change of the subject becomes possible.

Further, the division of regions, rules, etc. are not limited to those in this embodiment. Needless to say, the operation of this embodiment can be executed by software using a microcomputer.

(G) Effect of the Invention According to the present invention as described above, it is possible to detect a subject change after the end of the focusing operation regardless of the focused state,
It is possible to restart the focusing operation with high accuracy and appropriateness. Furthermore,
By using the fuzzy inference for this determination, it is possible to comprehensively determine the restart of the focusing operation including the brightness contrast and other factors.

[Brief description of drawings]

1 and 4 to 27 relate to an embodiment of the present invention,
FIG. 1 is an overall circuit block diagram, FIG. 4 is an explanatory diagram of hill climbing focusing operation, FIG. 5, FIG. 10, FIG. 11, FIG. 13, FIG.
FIG. 23, FIG. 23 is a circuit block diagram of a main part, FIG. 6 is an explanatory diagram of area setting, FIG. 7 is a relationship diagram between a lens position and a focus evaluation value, and FIG. 8 is a relationship diagram between a relative ratio and a blur degree, FIG. 9 is a relational diagram of the lens position and the focus evaluation value and the relative ratio, FIG. 12 is an explanatory diagram of area division, FIG. 14, FIG. 15, FIG. 16, FIG.
FIG. 18, FIG. 19, FIG. 25, FIG. 26, FIG. 27, and FIG. 28 are views showing membership functions, and FIG. 20, FIG. 21, FIG. 2 and 3 are circuit block diagrams of a conventional example. (1) …… Lens, (50) …… Focus evaluation value generation circuit,
(81) (87) …… Change amount detection circuit, (100) …… Focus motor control circuit, (4) …… Imaging circuit.

Claims (2)

(57) [Claims] 1. A focusing operation is performed by extracting an amount of a high frequency component of a luminance signal in an area set on the screen as a focus evaluation value and driving an image pickup system so that the focus evaluation value becomes maximum. An autofocus device for performing a focus detection operation, wherein the brightness difference in the area is detected, and the focus operation is restarted when a predetermined change occurs in the brightness difference after the focus operation ends. 2. The autofocus device according to claim 1, wherein fuzzy inference is used to determine the change in the brightness difference.