JPH0364847B2 - - Google Patents

Description

Detailed Description of the Invention The present invention projects an object image onto a pair of photoelectric conversion devices placed before and after a predetermined focal plane with a predetermined optical path difference, and converts the photoelectric conversion output from the photoelectric conversion devices into The present invention relates to a focus detection method that detects a focus state based on an evaluation function value obtained by calculation according to a predetermined evaluation function.

Such a focus detection method is conventionally known, and its principle will be explained with reference to FIGS. 1 and 2.

In FIG. 1, light reflected from an object 1 is condensed by an objective lens 2, divided by a half mirror 3, and incident on light receiving element arrays _D1 and _D2 . The light-receiving element rows D ₁ and D ₂ are arranged so as to have a predetermined optical path difference δ in the optical axis direction before and after the planned focal plane 4, and a blurred image corresponding to the amount of defocus with the objective lens 2 is formed on these light-receiving element rows. It is formed. Light receiving element row D
The photoelectric conversion outputs of 1 and D2 are A/D converted by the A/D conversion circuit 5, and then arithmetic processed by the CPU 6 based on a predetermined evaluation function. The in-focus and out-of-focus states are determined and displayed on the display elements 7a to 7d of the display unit 7, respectively. As an evaluation function, if x _o is the A/D conversion value of the output of the n-th element in the light-receiving element array, then, for example, 〓 ⁿ |x _o −x _o-1 | or |x _o −x _o-1 | The maximum value of , the sum of the maximum value of |x _o −x _o-1 | and the next largest |x _o −x _o-1 |, etc. are used. FIG. 2 shows the relationship between the evaluation function values F ₁ and F ₂ obtained by calculating the outputs of the light receiving element rows D ₁ and D ₂ based on the evaluation function and _the amount of defocus. When focused on, the evaluation function value F ₁ corresponding to this light receiving element row becomes maximum, and when focused on light receiving element row D ₂ , the evaluation function value corresponding to this light receiving element row becomes maximum.
When F ₂ becomes maximum and the intended focal plane 4 is in focus, the evaluation function values F ₁ and F ₂ become equal. Therefore,
Absolute value Δ of the difference between evaluation function values F ₁ and F ₂ = |F ₁ −F ₂
When this is smaller than a predetermined reference value A ₁ , it can be determined that the focus is in focus. Furthermore, when Δ≧A ₁ and F ₁ ≧F ₂ , the front focus state is determined, and when Δ≧A ₁ , it is determined that the focus is in focus.
When F ₁ <F ₂ , it can be determined and displayed that the camera is in a rear focus state, and focusing can be controlled manually or automatically based on this determination result. However, as the amount of defocus increases and the blur becomes larger, both F ₁ and F ₂ become smaller, and Δ=|F ₁
−F ₂ | also becomes small, and there is a risk that an in-focus state may be detected even at an out-of-focus position.

The simplest way to prevent this is to change the magnitude of Δ= _{|F 1 − F 2 |} This means that it is determined that focusing is not possible regardless of the condition. That is, as shown in FIG. 2, if the evaluation function values F ₁ and F ₂ are smaller than a predetermined threshold B, it is determined that focusing is impossible. By doing this, the amount of blur, that is, the amount of defocus increases, and Δ=|
A state in which F ₁ −F ₂ | becomes smaller than A ₁ will no longer be mistakenly determined to be "in focus."

However, if this is done, the range in which focus can be detected becomes narrower. This is because the evaluation function value F ₁
The relationship curve between F _{1 and F 2} and the amount of defocus differs depending on the subject, and depending on the subject, as shown in Fig. 3, Δ may change even if the evaluation function values F ₁ and F ₂ become smaller than the out-of-focus determination threshold B. Although it is possible to sufficiently detect the front focus and rear focus states without becoming smaller than the focus judgment reference value A ₁ , the above method is able to detect "in focus" based only on F ₁ and F ₂ <B, regardless of the size of Δ. This is because it is judged as 'impossible'.

Therefore, the method of determining inability to focus by comparing a predetermined inability to focus determination threshold B with the evaluation function values F ₁ and F ₂ has the drawback of unduly narrowing the range in which focus can be detected.

The purpose of the present invention is to eliminate this drawback, and the present invention projects an object image onto a pair of photoelectric conversion devices arranged before and after a predetermined focal plane with a predetermined optical path difference. Absolute value of the difference between evaluation function values F ₁ and F ₂ obtained by calculating the photoelectric conversion outputs from a pair of photoelectric conversion devices according to respective predetermined evaluation functions |
In a focus detection method that detects the in-focus state by comparing F ₁ −F ₂ | with a focus determination reference value A ₁ , the evaluation function value F ₁ , corresponding to a pair of photoelectric conversion devices
F ₂ and the absolute value of the difference |F ₁ −F ₂ | are F ₁ , F ₂ <B |F ₁ −F ₂ |<A ₂ B: Predetermined in-focus determination threshold A ₂ : Focus It is characterized in that it is determined that focusing is impossible when the reference value is equal to or slightly larger than the determination reference value _A1 .

In this way, in the present invention, the evaluation function values F ₁ and F ₂
In addition to the condition that is smaller than the in-focus determination threshold B, the absolute value Δ of the difference between the evaluation function values F ₁ and F ₂ = |
The system is designed to determine "unable to focus" only when the conditions that F ₁ - F ₂ | is equal to or slightly larger than the focus judgment reference value A ₂ are smaller than the reference value A ₂ . In this way, it is determined that the in-focus condition is actually determined to be below the level shown by B' in FIG. Focus detection becomes possible between QQ', and autofocus control can be performed over a correspondingly wider range. Moreover, the value of B′ is not fixed like B, and the value of |F ₁ −F ₂ | is actually
Since it is determined when A is smaller than ₂ , the actual in-focus determination level B' changes according to the evaluation function value curve that changes depending on the subject, and an appropriate focus detectable range can be obtained accordingly. Become.

Next, a preferred example of the method of the present invention will be explained with reference to the flowchart of FIG. In this example, first, in the judgment block 11, it is checked whether |F ₁ −F ₂ | is larger than or smaller than the reference value A ₂ which is slightly larger than the focus judgment reference A ₁ , and |F ₁ −F ₂ |≧A ₂ In this case, it is next checked in judgment block 12 whether F ₁ >F ₂ , and if F ₁ ≧F ₂ , it is determined that the front pin state is present.
When F ₁ < F ₂ , it is determined that the rear pin state is present. On the other hand, |
_{When F 1 −F 2 | _} It is judged that it is impossible to focus. On the other hand, when F ₁ , F ₂ ≧B, it is further checked in decision block 14 whether |F ₁ −F ₂ |<A ₁ , and when |F ₁ −F ₂ |<A ₁ , the focus is determined. If |F ₁ −F ₂ |≧A ₁ , it is further checked in decision block 12 whether F ₁ ≧F ₂ , and if F ₁ >F ₂ , the front pin state is determined. When ₁ < F ₂ , it is determined that the rear pin state is present.

Therefore, according to this method, as shown in FIG. 5, an out-of-focus determination output can be obtained in the region where |F ₁ −F ₂ |<A ₂ and F ₁ , F ₂ <B, and in V. , F ₁ ,
With F ₂ < B | F ₁ − F ₂ | ≧ A ₂ and F ₁ , F ₂ ≧ B |
F ₁ −F ₂ ｜≧A ₂ and in the region where F ₁ ≧
When F ₂ (region), the front pin state determination output is obtained, when F ₁ < F ₂ (region), the rear pin state determination output is obtained, and when F ₁ , F ₂ ≧B, |F ₁ − F ₂ ｜＜A ₁
A focus determination output is obtained in a certain area.

FIG. 6 is a block diagram of an apparatus for carrying out the method of FIG. 4. The outputs of the light receiving element arrays D ₁ and D ₂ are selected by a changeover switch 21 controlled by a control circuit 20 . First, when the output of the light-receiving element row D ₁ is selected, the output signal x _o of each light-receiving element is sequentially amplified by the amplifier 22 and A/D converter 23 receives the A/D converter 23.
After being D-converted, it is stored in the memory 24. After the output signals of all the elements of the light receiving element array D ₁ are stored in the memory, the output signals of the adjacent light receiving elements x _o ,
The shift circuit 25 and the subtraction circuit 2 calculate the difference x _o −x _o+1 between x _o+1.
Obtained by 6. Next, the absolute value of x _o −x _o+1 is calculated in the absolute value circuit 27, and the adder 28 calculates the absolute value of all the light receiving elements |x _o −x _o+1 | sum Σ|x _o −x _{o +1}
| is calculated to obtain the evaluation function value F ₁ , and this is stored in the F ₁ area of the memory 29 . Next, an example of a light receiving element
The output of D ₂ is selected and the same calculation is performed to obtain the evaluation function value F ₂ , which is stored in the F ₂ area of the memory 29. Next, the control circuit 20 stores data F ₁ , F ₂ stored in the F 1 , F ₂ areas of the memory 29 .
Data stored in F ₂ and constant memory 30
Read A ₁ , A ₂ , B and read these data F ₁ , F ₂ ,
Based on A ₁ , A ₂ , and B, the front focus, in-focus, back focus, and out-of-focus states are determined according to the flowchart in Figure 4, and such control circuits can be easily created using a microprocessor. realizable. The judgment output is latched by the latch circuit 32 of the display circuit 31, and the corresponding element (in this example, a light emitting diode) 33 is latched by the latch circuit 32 of the display circuit 31.
a, 33b, 33c, or 33d is emitted to display the corresponding state, and focus adjustment can be performed manually or automatically based on the determination result.

[Brief explanation of drawings]

Figures 1 and 2 are diagrams explaining the principle of a known focus detection method, Figure 3 is a diagram explaining the principle of the method of the present invention, and Figure 4
5 is a flowchart of a preferred embodiment of the method of the present invention, FIG. 5 is a determination characteristic diagram thereof, and FIG. 6 is a configuration diagram of an embodiment of an apparatus for carrying out the method of FIG. 4. D ₁ , D ₂ ... Light receiving element example, F ₁ , F ₂ ... Evaluation function value, B ... Inability to focus judgment threshold, A ₁ ... Focus judgment reference value, A ₂ ... Equal to A ₁ ? Slightly larger reference value, 20... Control circuit, 21... Changeover switch, 22... Amplifier, 23... A/D converter, 24... Memory, 25... Shift circuit, 26
... Subtraction circuit, 27 ... Absolute value circuit, 28 ... Addition circuit, 29 ... Memory, 30 ... Constant memory,
31...Display circuit, 32...Latch circuit, 33a
~33d...display element.

Claims (1)

[Claims] 1. An object image is projected onto a pair of photoelectric conversion devices arranged with a predetermined optical path difference before and after a planned focal plane, and the photoelectric conversion output from the pair of photoelectric conversion devices is Absolute value of the difference between evaluation function values F ₁ and F ₂ obtained by calculating each according to a predetermined evaluation function |F ₁ −F ₂ |
In _the focus detection method that detects the in-focus state by comparing _the
The absolute value of the difference between F ₂ |F ₁ −F ₂ | is the focus judgment reference value A ₁
If |F ₁ −F ₂ |≧A ₂ , then compare F ₁ and F ₂ , and if F ₁ ≧F ₂ , _the front pin state, F When ₁ < F ₂ , it is determined that the rear focus is on, and when |F ₁ - _{F 2} | < A ₂ , then F ₁ and F ₂ are compared with the focus judgment threshold B, and F ₁ , F ₂ < When B, it is determined that it is impossible to focus, and when F ₁ , F ₂ ≧B, further |F ₁ −
Compare F _{2 |} with the focus judgment reference value A ₁ , and when |F _{1 −F 2} |<A ₁ , it is determined that the state is in focus _; Compare F ₁ and F ₂ ,
A focus detection method characterized by determining a front focus state when F ₁ ≧F ₂ and a back focus state when F ₁ <F ₂ .