JPH01237610A - Auto focus device - Google Patents

Abstract

PURPOSE:To execute a self-correction of a coefficient by providing a correcting means for correcting the coefficient, based on a result of a focusing operation of a lens driving means and an arithmetic means. CONSTITUTION:A range finding output from a sensor unit 8 is inputted to a CPU 7. Subsequently, an operation for calculating a shift quantity from a focused position of an image pickup lens 1 by using a relative position relation of two images and a coefficient, sending-out of an output signal to a driving means for moving the image pickup lens 1, based on the shift quantity, and an operation for correcting the coefficient, based on a result of a focusing operation by lens driving and the arithmetic operation are executed by this CPU 7. That is, whenever the range finding is executed, the image shift quantity and the quantity of the lens which has been moved are stored, and after the focusing operation has been executed, the coefficient is recalculated by setting a final lens position as a target focal position, based on the stored data, and the coefficient is corrected. In such a way, the self-correction can be executed.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to an autofocus device, and more particularly to an autofocus device used in cameras and the like.

[Prior Art] As is well known, an autofocus device (hereinafter referred to as an AF device) of an eye reflex camera uses an AF sensor unit 8 having a separator lens as shown in FIG. An AF device is used that uses a correlation method to determine the amount of focus shift from the amount of shift. That is, the light transmitted through the photographic lens 1 and imaged on a surface 2 conjugate with the film is guided by a collimator lens 3 and a field mask 4 to two separator lenses 5a and 5b, and a first photoelectric conversion pixel array consisting of a plurality of photoelectric conversion pixel rows is guided. An image sensor 6 having a light-receiving element row 6a and a second light-receiving element row 6b has an image formed on the first and second light-receiving element rows 6a and 6b by the separator lenses 5a and 5b, respectively. It is. With this AF device, the amount of deviation between the two images is 1
Taking advantage of the fact that it almost corresponds to the amount of defocus, the image sensor 6 extracts the two images as a video signal,
The amount of image shift is determined from the data, and the photographing lens 1 is moved to the in-focus position based on the calculated value. That is, the amount of deviation Δ of the subject image on the image sensor 6 from the in-focus position is calculated from the output signals from the first light receiving element row 6a and the second light receiving element row 6b of the image sensor 6, and this amount of deviation and the photographing lens Using the linear approximation relationship between l and defocus IiD, defocus ff1D is determined from the amount of deviation Δ, and the photographing lens 1 is moved to the in-focus position.

However, in this AF device, the relationship between the amount of deviation Δ and the defocus mD is a linear approximation within the range of ±3 to 5 degrees relative to the in-focus position, but basically, Since they have a hyperbolic relationship, the relationship between the image shift amount Δ and the focus shift iD cannot be approximated to a straight line as shown in FIG. For this reason, the amount of movement of the photographic lens becomes so large that it becomes impossible to focus with a single focusing operation. In other words, when the photographing lens 1 is driven using the -th distance measurement result, the direct approximation Δ-
The photographic lens 1 will stop at a position that is too far from the in-focus position by the distance measurement error due to the difference in the Δ-D characteristic of the hyperbola with D characteristic.

Therefore, in order to solve this problem, a linear equation [A1.
JP-A-62-100718 discloses a solution that calculates the amount of focus deviation using the formula: A2: constant, Δ: amount of image deviation]
It is proposed by the publication No.

This varies depending on the product.

Therefore, this solution has the trouble of having to adjust the constants one by one during manufacturing, and the adjustment method is also complicated and time-consuming.

Moreover, there was also the drawback that the optimum constants varied depending on the temperature.

The purpose of the present invention is to solve these problems by
To provide an AF device in which the F device itself has a constant self-correction function.

[Means for Solving the Problems] In order to achieve the above object, the AF device of the present invention uses two images formed on an AF sensor unit having a first light-receiving element row and a second light-receiving element row. a calculation means for calculating the amount of deviation of the photographic lens from the in-focus position using the relative positional relationship and a coefficient; a drive means for moving the photographic lens based on the amount of deviation; and a combination of the driving means and the calculation means. The present invention is characterized by comprising a modifying means for modifying the coefficients based on the result of the focusing operation.

[Function] The conversion between the amount of image shift and the amount of focus shift deviates from the linear equation when the shift is large as shown in FIG. 2. However, unless the inclination is reversed, at any inclination (excluding 0 and flash), after repeating distance measurement → lens movement, the lens converges on the in-focus point.

Therefore, in the present invention, the amount of image shift and the amount of lens movement are memorized during each distance measurement, and the coefficients are recalculated based on the data stored after the focusing operation, with the final lens position set as the target focus position. Modify the coefficients by weighting /m. This is done for each AF to self-correct.

[Implementation example] The present invention will be explained below with reference to illustrated embodiments.

FIG. 1 is a schematic configuration diagram of an AF device to which the present invention is applied, in which light reflected from a subject that passes through a photographic lens 1 is divided by a half mirror 10, and one of the lights is transmitted through a shutter (not shown). The other light is guided to the film surface 9, and the other light is guided to the AF sensor unit 8 disposed at a position conjugate with the film surface 9. The configuration of this AF sensor unit 8 is similar to that explained in FIG.
The distance measurement output from is input to the CPU 7. Then, a calculation is performed to calculate the amount of deviation of the photographing lens 1 from the in-focus position using the relative positional relationship between the two images and coefficients, and a driving means (not shown) that moves the photographing lens 1 based on the amount of deviation. The CPU 7 performs operations such as sending out an output signal, and correcting the coefficients based on the results of lens driving and focusing operations based on the calculations described above.

FIG. 2 is a flowchart of a coefficient self-correction algorithm showing a first embodiment of the present invention. This algorithm repeats an image shift detection subroutine that detects the amount of shift between two images using the distance measurement signal from the AF sensor unit 8, calculates the amount of focus shift, and moves the lens, and stores the amount of focus shift and image shift each time. After focusing on the in-focus area, perform the shutter sequence and then memorize the amount of defocus.

It is composed of a coefficient correction section that self-corrects the coefficients based on the amount of image shift.

Next, the operation will be explained. At the start of AF, the image shift amount Δ is calculated and stored in Δ1. Image shift amount Δ
The focus shift mD is a coefficient that is self-corrected according to the present invention.

The focus deviation measurement is also memorized in Dl, and the lens 1 is moved by D. Next, to confirm focus, the image shift amount Δ is determined again by n1, and this is stored in the next memory. If it is not in focus, the amount of focus deviation is determined, stored in the next memory, and the lens is moved by the determined value. This is repeated until the focus is within tolerance. Once the focus is within the focus tolerance, move on to the shutter sequence. Then, after the operations necessary for photographing are completed, the coefficients are self-corrected. If the number of repetitions is two or less, it is assumed that no correction is necessary and no correction is performed. Corrections are made using the first and second distance measurement data. The true amount of defocus is calculated from the total of the memorized defocus ff1D up to that time, and the current AF is calculated from that value and the amount of image deformation.
Recalculate the coefficient of time. This coefficient is weighted by 1/m to modify the coefficient.

Note that no correction is made if it is impossible to focus. °Also, when there are an abnormally large number of repetitions, it is also effective to make corrections using data from several times before focusing. Also, the weighting is constant m
may be made smaller as the number of repetitions increases.

In this way, according to this embodiment, the coefficients are self-corrected when there is no effect on the photographing operability after the shutter sequence, so the focusing time becomes shorter over time, and it is also possible to adjust the coefficients due to changes in temperature characteristics. Effects can also be corrected, and accurate AF can be performed at high speed and with high reliability. Furthermore, since no adjustment is required during manufacturing, the product is inexpensive and has no variation in product quality. Furthermore,
By changing the weighting constant m by the number of repetitions, it is possible to quickly correct even if the coefficients deviate greatly.

3 and 4A, 4B, and 4C show a second embodiment of the present invention. A in this example
The optical system of the F device and the configuration of each unit are similar to those of the first embodiment. In this embodiment, as shown in Fig. 3, the calculation of the focus deviation ff1D is divided into each area, and linear approximation is performed within each area to calculate the focus deviation uD from the image deviation amount Δ. Flowcharts of the algorithm are shown in FIGS. 4A to 4C.

As shown in the flowchart of FIG. 4C, the algorithm for calculating the amount of focus deviation detects the amount of image deviation Δ in the same manner as in the first embodiment, and calculates the focus deviation ff1 by the linear approximation described above.
The amount of lens movement is obtained by calculating D, and the lens is moved based on this amount. After the lens movement is completed, the image deviation amount Δ is detected again, the focus deviation ff1D is calculated, and the lens movement is repeated until the lens is brought into focus. Then, the image shift amount Δ and the focus shift QD for each repetition are recorded. The configuration and operation up to this point are almost the same as those of the first embodiment.

Further, if the in-focus point is reached in one focusing operation (K-1) after completing the shutter sequence after focusing, the sequence ends there. When in-focus is achieved after multiple focusing operations, the coefficients for calculating the amount of focus deviation are self-corrected. This correction data uses the image shift amount Δ and focus shift HD each time.

The operation of the AF device of this second embodiment will be described next.
At the start, the image shift amount Δ is calculated and stored in Δ1. The focus shift mD is calculated based on the image shift amount Δ. In calculating the focus shift 2D, as shown in FIG. 4A, it is determined whether the image shift amount Δ is positive or negative. Whether the lens is extended or retracted is determined by the positive and negative signs. Furthermore, as is clear from FIG. 6, the slope differs depending on whether the amount of image shift is positive or negative (extension or retraction). As a result, calculations are performed separately into areas in the positive direction and areas in the negative direction, but since only the coefficients are different, the positive direction will be explained here as an example.

First, it is determined in which area the image shift occurs.

For example, if the image shift is in area 3 as shown in FIG. 3, the image shift amount Δ will not become negative until et, e2, and e3 are subtracted. In the area 3 routine, the out-of-focus e3 is added, the size in the area 3 is multiplied by the slope a3 in the area 3, and the out-of-focus ff1B2 up to the area 2 is added to obtain the out-of-focus ff1D. When the image shift amount Δ is negative, these signs are opposite. The focus shift flD is
This is then multiplied by a lens factor to determine the actual amount of focus shift.

After calculating the focus shift JuD, the shift mD is recorded and the lens is moved in the same way as in the first embodiment, and this is repeated until focusing. After focusing, a shutter sequence is performed and necessary operations of the CPU, such as exposing the film, are completed. After that, a focus calculation is performed on the image deviation detection bone to confirm whether or not the image is finally in focus (see FIG. 4C). Since the focus judgment is performed even if the focus is not necessarily perfect, the final focus deviation EilD is determined in order to reflect the deviation in the correction.

If the focus is achieved with -degree AF operation (K-1)
terminates the entire sequence as there is no need to modify the coefficients. If this is not the case, the inclination of the area where the Δ is present is corrected in order from the distance measurement point closest to the focal point (the one with larger K). For this correction, as shown in Figure 4B, assuming that the slope of the area with less image shift than the area to be corrected is correct, the value of the slope a that was needed this time is determined, and the coefficient a is corrected by adding a weight of 1/m. do. Further, this weighting is performed by setting the coefficient m to the number of repetitions, for example, L-3, so that the weighting is different for each area and appropriate correction can be performed. In addition, similar to the first embodiment, if there is a concern about inability or focus, such as when K exceeds a certain level, the correction may not be performed.

[Effects of the Invention] As described above, according to the present invention, (1) the formula is simple, (2) it is a means for correcting focus shift calculations that increases the calculation speed as the amount of movement of the lens is small, and self-correction is possible; .

■You can make corrections by dividing them into areas, so if you make corrections to frequently used areas first, you can finish learning faster.

■Other factors caused by the AF device, such as temperature characteristics, deviation, and aging, are also corrected each time the AF device is used, and the system is adjusted to the best condition without any adjustment during manufacturing.

■Although this is an approximation using a straight line, by performing the same approximation using a curved line, the number of areas and errors can be reduced.

It is possible to provide an AF device that exhibits remarkable effects such as, and eliminates the conventional drawbacks.

[Brief explanation of the drawing]

FIG. 1 is a schematic configuration diagram of an AF device to which the present invention is applied; FIG. 2 is a flowchart of the focus shift amount calculation means and coefficient correction means in the AF device showing the first embodiment of the present invention; FIG. , a diagram showing the areas in the characteristic curve of the amount of focus shift - the amount of image shift, and FIGS. 4A to 4C show the AF according to the second embodiment of the present invention.
Flowcharts of the focus shift amount calculation means and coefficient correction means in the device; FIG. 5 is a configuration diagram of the AF sensor unit in the AF device; FIG. 6 is a characteristic line diagram of focus shift amount-image shift amount. 1・・・・・・・・・・・・Photography lens 7...
・・・・・・・・・・・・CPU (calculation means, correction means) 8 ・・・・・・・・・・・・AF sensor unit (first light receiving element row) (second photodetector array)

Claims (1)

[Claims] (1) Passed through different optical paths that sandwich the optical axis of the photographic lens 2
Using the relative positional relationship and coefficients of the first light-receiving element row and the second light-receiving element row consisting of a plurality of pixels that receive two images, and the two images formed on these light-receiving element rows, the above-mentioned a calculation means for calculating the amount of deviation of the photographic lens from the in-focus position; a driving means for moving the photographic lens based on the amount of deviation; An autofocus device comprising: a correction means for correcting a coefficient.