JPH07270674A - Focusing detecting device - Google Patents

Abstract

PURPOSE:To obtain a focusing detecting device capable of automatically focusing at a high speed and further, with high accuracy. CONSTITUTION:The pupil of a photographic system 1-1 is divided into plural regions by a diaphragm 1-2 having plural apertures, plural bits of image information based on a luminous flux passing through plural divided regions are formed on the surface of an image pickup means 3 and when the focusing state of the photographic system is detected by plural focusing means with an image signal from the image pickup means, out of plural focusing means, first, a first focusing means 1-14 detects the focusing position of the photographic system and then, a focusing detecting system selecting means 1-16 chooses whether the focusing is executed by using a second focusing means 1-15 different from the first focusing means or the first focusing means, to detect the focusing state of the photographic system.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a focus detecting device, and more particularly to a focus detecting device suitable for a still camera, a video camera or the like for performing an autofocus using an image pickup device such as a CCD.

[0002]

2. Description of the Related Art Conventionally, a focus detection method called a hill climbing method has been known as an autofocus method in a photographing device such as a still camera or a video camera. This hill climbing method directly evaluates the sharpness of the image (subject image) and drives the focus lens (focusing lens) to change the focus to find the point where the sharpness of the image is maximum and focus on that point. The in-focus state of the shooting system is obtained as the position.

An in-focus detection apparatus using such an image sharpness evaluation method is disclosed in, for example, Japanese Patent Laid-Open No. 62-103616.
It has been proposed in the publication.

In the publication, there is provided a focus detection apparatus having a detection means for detecting the width of an edge portion of a subject image (object image) and a determination means for determining a focus state based on the size of the width. Then, the detection means detects the width of the edge portion by detecting the ratio of the gradient of the luminance change of the edge portion of the subject image and the luminance difference in the edge portion, and the width of the edge portion is small. The sharpness of the image is used to obtain the focus state of the shooting system.

FIG. 8 is an explanatory diagram showing an example of a graph of evaluation values in the hill climbing method. In the figure, the horizontal axis is the lens position of the focus lens, and the vertical axis is the sharpness evaluation value.

As shown in the same figure, when the lens position of the focus lens is shifted from the ∞ point to the N point (near point) side as shown by arrow 7-3, the sharpness evaluation value becomes a peak-like waveform A. Become. At the apex of the waveform A, that is, at the in-focus position 7-1, the evaluation value is lowered no matter which direction the focus lens is moved. At the position 7-2 of the waveform A, the evaluation value increases on the N point side and decreases on the ∞ point side.
I know the direction.

However, in this focus detection method, it is necessary to sample a number of frames to obtain the sharpness evaluation value while driving the focus lens before the focus position is reached, so it takes time to detect the focus. In addition, it is difficult to predict the in-focus position.

Also, a focus detection device utilizing an image sharpness evaluation method is proposed in, for example, US Pat. No. 4,18591. In the same issue, in the method of obtaining the focus information by comparing the sharpness of a plurality of images formed from the light flux that has passed through a plurality of parts of the taking lens, different things on the pupil plane of the taking lens using a half mirror etc. The two subject images based on the light flux that has passed through the two regions are reflected by the half mirrors, respectively made incident on the image pickup element (sensor) surface, and the phase difference is measured to obtain the focused state of the photographing system.

However, in this focus detection method, since an optical member such as a half mirror is used in the device, the entire device becomes complicated and large in size, and the image pickup element for focus detection and the image pickup means are combined. Since the relationship with the image plane is set to the same condition, there is a problem that manufacturing cost is required to obtain the positional accuracy. Further, there is a problem in the reliability of the position accuracy with time and the temperature change.

Therefore, in order to solve the above problems, the pupil of the photographing system is divided into a plurality of regions (pupil regions) by using a diaphragm having a plurality of apertures, and the divided regions are passed through. A plurality of image information based on the light flux is formed on the surface of the image pickup means, and a focus state of the image pickup system is detected by performing a correlation calculation using an image signal (image information) obtained by the image pickup means. As a result, focus detection is performed without using a special image sensor for focus detection.

[0011]

The hill-climbing method, which is an auto-focusing method used in the still camera, the video camera, etc., is performed by directly evaluating the video signal (image information) from the image sensor, and therefore the focusing accuracy is improved. Has a high price and does not require a special sensor. On the other hand, however, as described above, it takes time to sample a large number of frames to obtain a sharpness evaluation value before reaching the in-focus position, and it is difficult to predict the in-focus position. There was a problem.

Further, in the conventional focus detection method in which the focus information is obtained from the result of the correlation calculation of the image signal by using the diaphragm having a plurality of apertures, there is an advantage that sampling is performed for one frame. . However, on the other hand, when the defocus amount is small, it is difficult to detect the peak of the in-focus position. Therefore, it is necessary to make fine adjustment to obtain the detection accuracy. There is also a problem that it is difficult to keep the in-focus state by following the subject of a moving image.

According to the present invention, a special image pickup element for focus detection is provided by detecting the focus state of the photographing system by a plurality of focus means (focus detection method) using the image signal from the image pickup means. An object of the present invention is to provide a focus detection device capable of performing accurate autofocus at high speed and with high accuracy without using it.

[0014]

A focus detection apparatus according to the present invention divides a pupil of a photographing system into a plurality of areas by a diaphragm having a plurality of openings, and based on a light flux passing through the plurality of divided areas. When a plurality of image information is formed on the surface of the image pickup means and the focus state of the photographing system is detected by the plurality of focus means using the image signal from the image pickup means, among the plurality of focus means,
First, the first focus means detects the focus position of the photographing system, and then the focus detection method selection means uses another second focus means different from the first focus means to focus. Is selected or the focusing is performed by the first focusing means, and the focusing state of the photographing system is detected.

In particular, the first focusing means detects the focusing position by correlation calculation from the signals of a plurality of subject images formed on the surface of the image pickup means generated by the diaphragm, and the second focusing means. Is characterized in that the sharpness of the subject image formed on the surface of the image pickup means is taken out as an evaluation value to detect the in-focus position.

[0016]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS FIG. 1 is a schematic view of the essential parts of a first embodiment of the present invention.

In FIG. 1, reference numeral 1-1 is a photographing system (photographing lens). Reference numeral 1-2 is an aperture, which will be described later with reference to FIG.
As shown in FIG. 2, two openings 1-6 having different areas are provided.
1-7, the pupil of the taking lens 1-1 (exit pupil)
Is divided into two areas. In this embodiment, the opening 1-
The area of 7 is formed to be wider. Reference numeral 3 denotes an image pickup means, which includes an image pickup element (sensor) such as a CCD.

Reference numeral 1-3 shows the position of the image pickup means 3 when the photographing system 1-1 is in focus, and 1-4 shows the photographing system 1.
-1 shows the position of the image pickup means 3 when the rear focus state is present, and 1-5 shows the position of the image pickup means 3 when the image pickup system 1-1 is in the front focus state.

Reference numeral 1-13 is a drive means, which is a focus lens (focus lens) which constitutes the photographing system 1-1 based on a focus signal from a focus detection method selection means 1-16 which will be described later.
Is driven to focus.

Reference numerals 1-14 and 1-15 are first and second focusing means (focusing information detecting means), respectively, which use the image signal (image information) from the image pickup means 3 to detect the image pickup system. The focus state is detected respectively.

First focusing means 1-14 in this embodiment
Detects the in-focus position by correlation calculation from the signals of a plurality of subject images (images) formed on the image pickup surface of the image pickup means 3 caused by the diaphragm 1-2. The second focusing means 1-15 detects the focusing position by extracting the sharpness of the subject image as an evaluation value using the signal of the subject image formed on the surface of the image pickup means 3.

Reference numeral 1-16 is a focus detection system selection means,
The reliability is judged using the signals from the first and second focusing means 1-14 and 1-15, and the focusing means (autofocus mode) is selected from the result of the determination to focus the imaging system. It is detecting.

Each of the above means 1-14, 1-15, 1
-16 may be configured by software, and in selecting the focusing means, not only one focusing means but both may be selected simultaneously.

In this embodiment, when the image pickup system 1-1 is in focus (when the image pickup means 3 is at the position 1-13),
Two pieces of image information based on the light fluxes that have passed through the two openings 1-6 and 1-7 of the diaphragm 1-2 are imaged at the same position 1-8 on the surface of the image pickup means 3. Further, when the photographing system 1-1 is not in focus, for example, when the photographing system 1-1 is in the rear focus state (when the image pickup means 3 is at the position 1-4), the light flux passing through the opening 1-6 is imaged. At the position 1-9 on the surface of the means 3, the opening 1-7
The light flux that has passed through is imaged at a position 1-10 on the surface of the image pickup means 3 in a slightly blurred state.

Further, when the photographing system 3 is in the front focus state (when the image pickup means 3 is at the position 1-5), the light flux passing through the opening 1-6 reaches the position 1-12 on the surface of the image pickup means 3, Opening 1
The light flux passing through -7 forms an image at a position 1-11 on the surface of the image pickup means 3 in a slightly blurred state.

FIG. 2A shows the diaphragm 1-2 of FIG.
It is explanatory drawing which showed the shape of the size of the one opening 1-6.1-7, and FIG. 2 (B) is an explanatory view which showed the example of the image in the rear pin when this diaphragm 1-2 is used. It is a figure. Figure 2 (B)
2-4 is an image when passing through the opening 1-6, and 2-5 is an image when passing through the opening 1-7.

As described above, the brightness of the image on the image plane in the non-focused state is larger in the opening 1-7 than in the opening 1-7.
Since the area is larger than that of 6, the brightness of the image at the positions 1-10 and 1-11 on the surface of the image pickup means 3 is larger than that at the positions 1-9 and 1-12 on the surface of the image pickup means 3, respectively. Also becomes a bright and blurred image.

The brightness of these two images 2-4 and 2-5,
Alternatively, the direction of the focus position of the photographing system can be determined by calculating the degree of blurring of the image by the focus detection means. Note that the information used for this determination may be determined using both the brightness of the image and the degree of blurring.

As described above, in the out-of-focus state, the images formed on the image pickup surface are blurred and multiply overlapped. By calculating the autocorrelation of the image signal from the image pickup means 3 in this state, it is possible to obtain the deviation amount of the two images that overlap each other, and it is possible to obtain the information of the defocus amount from the deviation amount.

Next, the first focusing means (first autofocus mode) using this autocorrelation calculation will be described.

A signal based on the subject image formed on the surface of the image pickup means 3 is output as an electric signal having the time axis t as a variable. When this signal (original signal) is f (t), the autocorrelation function C (τ) is expressed by the following equation.

[0032]

[Equation 1] Here, T indicates the range of data used for distance measurement.

Since the actual calculation is performed on the discrete data output from the image pickup means 3, the original signal f (t) can be expressed as X (i, j) using the coordinates on the image pickup surface. Here, i represents a horizontal coordinate and j represents a vertical coordinate. When the coordinate X (i, j) is used, the autocorrelation function C
(M, n) is expressed by the following equation.

[0034]

[Equation 2] In equation (2), I represents the number of pixels in the horizontal direction of the ranging frame, and J represents the number of pixels in the vertical direction of the ranging frame. Coordinate X (i, j)
And the coordinate X (i + m, j + n) are shown in FIG.
When m and n in the equation (2) have a constant relationship and m and n are expressed as a function of K, the coordinate C (m,
n) can be represented as C (k) as a function of one variable.

Further, when the function C (k) is obtained when n is zero, it is equal to the continuous variable τ in the equation (1) replaced by the discrete variable k.

Generally, in the two-line blurred image shown in FIG. 2 (B), the correlation function C (m) is obtained from the image signal as shown in FIG. 4 (A). When m is equal to the shift amount of the overlapped images, a peak 4-1 occurs as shown in FIG. As a result, the value mp when m is other than 0 and the function C (m) has a peak becomes the interval (pixel interval) between the misaligned images on the surface of the imaging means.

The defocus amount can be obtained from this value. Further, as the defocus direction, the value mp giving the peak of this autocorrelation can be used to determine the magnitude of the brightness of the two images displaced by this pixel from the positive / negative of the average of the differences in the brightness of each pixel.

In this focus detection method, the straight line shown in FIG. 2A (the line connecting the centers of the two openings 1-6 and 1-7)
Since an image shift occurs in a direction parallel to 2-3, the apertures 1-6 and 1-7 should be installed so that the arrangement direction is not parallel to the scanning line direction of the image pickup means when the diaphragm is installed. As a result, it is possible to satisfactorily perform focus detection even on a subject having a horizontal stripe pattern, which is difficult for the conventional focus detection method.

The number of apertures of the diaphragm may be increased to three or more, and the pupil may be divided into a plurality of pieces to obtain three or more pieces of image information. Also in this case, it is sufficient to arrange the plurality of openings so that the arrangement direction of the plurality of openings is not parallel to the scanning line direction of the image pickup means.

Further, in this embodiment, as can be seen from FIG. 4A, the value of the autocorrelation function becomes maximum when m is 0, and the peak corresponding to the amount of focus shift is smaller than that. Therefore, this shift amount mp is close to zero.

That is, when the defocus amount is considerably small, the correlation value of the second peak is hidden by the peak when m is 0, which makes detection difficult. Moreover, the waveform of the data of the correlation calculation such as the subject of the periodic pattern is shown in FIG.
As shown by the respective peaks 4-2, 4-3, 4-4, 4-5, there are cases where the waveform has a plurality of peaks even when m is not 0. In such a case, accurate focus determination becomes difficult.

Therefore, in this embodiment, in such a case, as will be described later, first, an approximate focus position is obtained from the image signal of one frame by a correlation calculation, and then sharpness with high precision is obtained by the focus detection method selection means. The second focusing means (second autofocus mode) using the degree evaluation is switched.

Next, the second focusing means (second autofocus mode) for evaluating the sharpness of this image will be described.

Even if the optical path is divided by a diaphragm having a plurality of apertures, the degree of blurring is minimized during focusing, and the degree of blurring between overlapping images is increased during non-focusing. It is also clear from. By digitizing this as the evaluation value of the sharpness by the above-described conventional focus detection method, it is possible to perform the same hill-climbing focus detection as the conventional one.

Even when a diaphragm having a plurality of apertures is used, the sharpness evaluation value becomes maximum in the focused state, and in the unfocused state, just like the diaphragm having one aperture. The smaller the position, the smaller it becomes.

As an example of a subject which is poor in the method of evaluating the sharpness of the image, for example, when there is a high-luminance subject, in the hill climbing method of evaluating the sharpness, a pseudo peak occurs at a high-luminance subject. However, there is a problem that the correct focus position cannot be obtained.

However, when there is a high-brightness object having a certain threshold value or more, if the focus position is obtained by autocorrelation calculation, the focus position can be calculated from the image shift amount even in the case of a high-brightness object. Since it can be obtained, malfunction can be prevented.

Further, in the focus detection method for evaluating the sharpness of the image, the edge in the direction perpendicular to the scanning line is the object of evaluation, especially when the processing is performed by an analog circuit.
There is a problem that it becomes difficult to detect the focus by evaluating the sharpness for a horizontal stripe pattern object having only edges parallel to the scanning line direction.

Therefore, in the present embodiment, as will be described later, a plurality of focus detection methods described above are selectively used according to the focus information obtained from each focus detection means, so that subjects that are not good at each method can be separated from each other. I try to make up for each other. As a result, accurate autofocus with few malfunctions is performed.

FIG. 7 shows a concrete block diagram of the first embodiment of the present invention.

In the present embodiment, first, when performing the autocorrelation calculation, the luminance signal separation circuit 6-8 is used for the image pickup means (CC
D) The luminance signal is extracted from the signals from 6-3, and A /
The D converter 6-9 converts it into a digital signal and stores the signal X (i, j) of each pixel in the memory 6-10.

Then, the signal X (i, j) is read by the microprocessor 6-7, autocorrelation calculation is performed while changing the value of m in the above equation (2), and the motor driver 6 is based on the calculation result. The focus lens forming the photographing system 6-1 is driven by the motor 6-4 via -8 for focusing.

In the present embodiment, the defocus direction may be checked once after finding the peak. Further, the time for obtaining the information of the in-focus position after sampling one frame depends on the processing capability of the microprocessor 6-7.

Next, a calculation method based on evaluation of image sharpness will be described. In this embodiment, the ES method for evaluating the edge width of the contour of the image is used as one of the sharpness evaluation methods.

That is, using the luminance signal from the luminance signal separation circuit 6-8, the difference circuit 6-11, the differentiation circuit 6-12 and the like determine the luminance difference and the differential value of the edge portion, and the edge is calculated from the ratio of these values. The reciprocal of the width is obtained by the dividing circuit 6-13, and integrated by the integrating circuit 6-14 within a predetermined range.
Then, the integrated value is converted into a digital value by the A / D converter 6-15, read by the microprocessor 6-7, and focused by the motor 6-4 via the motor driver 6-8 at the position where the maximum value is reached. Focusing is performed by sampling while driving the lens.

5 and 6 are flowcharts of the first embodiment of the present invention. Next, the operation of this embodiment will be described according to this flowchart.

First, when the switch is turned on in step 5-1, the first autofocus mode by autocorrelation calculation is set. Next, in step 5-2, after the image data for one screen (one frame) is loaded into the memory, step 5-3
Then, the autocorrelation calculation is performed to determine the movement amount and direction of the focus lens.

Next, in step 5-4, it is determined whether or not the subject is poor in autocorrelation, and if it is determined that the subject is not good or the peak is not found near the in-focus point as described above, step 5-9. If it is determined that the subject is not a subject that is poor in autocorrelation, the process proceeds to step 2-5.
In step 5-3, the focus lens is driven by the amount of movement obtained in step 5-3.

Then, in step 5-6, it is determined whether or not the subject is poor in sharpness evaluation, and it is determined that accurate peak detection is difficult, for example, because the subject includes a subject having high brightness. In this case, the focus lens is stopped at the focal point obtained at the autocorrelation position, and the process proceeds to step 5-7. If not, the process proceeds to step 5-9, and the second autofocus mode based on the sharpness evaluation is switched to the accurate mode. Find the focal point.

As a determination method in steps 5-6, in the case of a high-brightness subject, a threshold value is determined in advance from data obtained by photographing a high-brightness subject, and the value is exceeded in the data of one frame read into the memory. It can be realized by checking with software or the like whether there is a value.

Then, in step 5-7, one frame is sampled again, and steps 5-2 and 5-3 described above are performed.
The same processing as described above is performed to determine whether or not the peak of the value mp of the correlation function really disappears. If the peak disappears, the process proceeds to step 5-8 to bring the object into focus. Repeat the steps shown in.

On the other hand, after step 5-9, the second autofocus mode is entered by evaluating the sharpness of the image.
The focus detection method for evaluating the sharpness of an image takes time because it is necessary to sample many frames while driving the lens, but in this embodiment, the focus point is known in advance by the autocorrelation calculation. The sharpness may be evaluated only before and after focusing.

In FIG. 6, ES represents the evaluation value of the sharpness obtained by using the size of the edge width of the image, and is referred to as "evaluation value ES" hereinafter.

First, in step 5-9, the calculation for obtaining the evaluation value ES is started, and then, in step 5-10, sampling of the evaluation value ES is continued while driving the focus lens. First, the driving direction is reciprocated by a predetermined amount from that point. Next, if there is a peak during this driving in step 5-11, the value of the peak ESp is stored, and the focus lens is moved to that point to obtain a focused state.

When there is no peak, the change in the evaluation value ES increases on one side and decreases on the other side as in the position 7-2 of the waveform A shown in FIG. Then you can find the peak.

Next, after reaching the in-focus state in steps 5-12 and 5-13, the change of the evaluation value ES is constantly monitored and repeated until the value of the evaluation value ES becomes smaller than the value of the peak ESp.

If the in-focus state is reached without evaluating the sharpness, the above-mentioned step 5
It is sufficient to repeat steps 7 and 5-8 to load one frame of data into the memory and monitor the result of the autocorrelation calculation.

Further, by using a microprocessor capable of parallel processing, it is possible to simultaneously monitor changes in the focus state of the photographing system by the two focus detection methods described above.

Next, in step 5-14, when the focus state of the photographing system changes, the autofocus is restarted. At this time, the threshold value ESo is arbitrarily set according to the conditions at that time, for example, half the value of the peak ESp at the time of focusing.

Then, in step 5-15, the evaluation value ES is compared with the threshold value ESo. If the changed evaluation value ES is smaller than the threshold value ESo, it is determined that the in-focus position has changed significantly, and the process proceeds to step 5-16. Advance, sharpness evaluation value ES
Calculation is stopped and the process returns to step 5-2 to return to the first autofocus mode by the autocorrelation calculation and the above steps are repeated.

On the contrary, when the evaluation value ES is larger than the threshold value ESo, it is determined that the in-focus position is slightly moved,
In the second autofocus mode for sharpness evaluation, the process returns to step 5-10 to repeat the steps described above to search for the in-focus position.

As described above, in this embodiment, by appropriately controlling each means so as to maintain the in-focus state as described above, the number of sampling frames can be reduced as compared with the conventional in-focus detection method, and Focusing detection is performed quickly and accurately by complementing each other with a plurality of focusing detection methods for subjects that are not suitable for causing malfunctions.

When it is desired to record a high-definition image signal such as still image shooting, a diaphragm 6-2 whose aperture number can be switched and an aperture switch 6-6 are used for shooting as shown in FIG. Only the diaphragm may be switched to a diaphragm having a single opening. This makes it possible to prevent overlapping of images with different backgrounds.

Although the ES method is used as one of the image sharpness evaluation methods in the present embodiment, the present invention is not limited to the above-described embodiment even if focus detection is performed using another evaluation method. It can be applied similarly.

[0075]

According to the present invention, as described above, after the in-focus position is obtained from the video signal for one frame by using the correlation calculation, the auto-focus of the sharpness evaluation system with high accuracy according to the in-focus information. Since the mode can be switched, the number of frames to be sampled can be reduced compared to the conventional focus detection method, the focus position can be reached quickly, and two focus points can be obtained by using the diaphragm having the same multiple openings. Since the detection method (autofocus mode) can be used without switching the aperture, the focus detection method can be switched instantaneously, or the correlation calculation and the sharpness evaluation value calculation can be performed simultaneously by the two focus detection methods. It is possible to achieve a focus detection device that can also perform.

Furthermore, according to the present invention, by selectively using a plurality of focus detection methods according to the focus information, it is possible to complement the subjects which are not good in each method with each other, thereby achieving accurate autofocus with less malfunction. A focus detection device that can be implemented can be achieved.

[Brief description of drawings]

FIG. 1 is a configuration diagram of a main part of a first embodiment of the present invention.

FIG. 2 is an explanatory diagram showing an aperture shape according to the first embodiment of the present invention and an example of an image when using the aperture shape.

FIG. 3 is an explanatory diagram showing a relationship between pixels on an image pickup unit surface according to the first embodiment of the present invention.

FIG. 4 is an explanatory diagram showing a graph of a calculation result of an autocorrelation function according to the first embodiment of the present invention.

FIG. 5 is a flowchart of the first embodiment of the present invention.

FIG. 6 is a flowchart of the first embodiment of the present invention.

FIG. 7 is a block diagram of a main part of the first embodiment of the present invention.

FIG. 8 is an explanatory diagram showing a graph of evaluation values of sharpness in Example 1 of the present invention.

[Explanation of symbols]

 1-1 Photographing system 1-2 Aperture 3 Imaging means 1-13 Driving means 1-14 Focusing information detecting means 1-15 Focusing information detecting means 1-16 Focusing detection method selecting means 2-1 and 2-2 Aperture Department

Front page continuation (72) Inventor Toshinobu Yamaguchi 3-30-2 Shimomaruko, Ota-ku, Tokyo Canon Inc. (72) Inventor Hiroshi Saruwatari 3-30-2 Shimomaruko, Ota-ku, Tokyo Canon Inc.

Claims (2)

[Claims] 1. A pupil having a plurality of apertures is used to divide a pupil of a photographing system into a plurality of areas, and a plurality of image information based on light fluxes passing through the plurality of divided areas are formed on an image pickup means surface. When the focusing state of the photographing system is detected by the plurality of focusing means using the image signal from the image pickup means, first of the plurality of focusing means, the first focusing means focuses the focusing system of the photographing system. After detecting the in-focus position, the in-focus detection system selecting means performs in-focus by using another second in-focus means different from the first in-focus means, or in-focus by the first in-focus means. A focus detection device characterized in that a focus state of the photographing system is detected by selecting whether to perform focus. 2. The first focusing means detects a focusing position by a correlation calculation from signals of a plurality of subject images formed on the surface of the image pickup means generated by the diaphragm, and the second focusing means. The focus detecting apparatus according to claim 1, wherein the means detects the focus position by taking out the sharpness of the subject image formed on the surface of the image pickup means as an evaluation value.