JPS61166509A - Focus detector - Google Patents

Abstract

PURPOSE:To improve the reliability of a focus detector by setting a reliability discrimination level in accordance with the operation condition of a charge storage photosensor array. CONSTITUTION:An image sensor 20 consisting of the charge storage photosensor array and a temperature sensor 21 using, for example, a thermistor which detects the operation temperature of this image sensor 20 are provided. A discrimination level controller 25 reads in a reset time and a gain outputted from a sensor controller 23 and the operation temperature of the image sensor 20 outputted from the temperature sensor 21 and set discrimination levels PL, CL, and YML in accordance with a setting table. Experimentally determined numerical values are given to this table, and these numerical values are compared with a contrast value C, a correlation level given signal YM, and a picture element signal peak value P.

Description

DETAILED DESCRIPTION OF THE INVENTION [Industrial Field of Application] The present invention relates to a focus detection device such as a camera equipped with a charge storage type photosensor array.

[Problem to be solved by the invention] The first lens of the photographic lens is symmetrical to the optical axis.
The subject light beams that have passed through the and second regions are respectively re-imaged to create two images, and the mutual positional relationship of these two images is determined to determine the amount of deviation of the imaging position from the expected focal position and its deviation. A focus detection device has already been proposed that determines the direction (whether the imaging position is in front of or behind the expected focal position, that is, whether it is in the front or in the back). With such a focus detection bag, this optical system has a predetermined focal plane (4) behind the photographic lens (2) or further behind this plane (the device has a condenser lens (6), and further behind it a refocusing lens). An image sensor (20) is provided with imaging lenses (8) and (IO), and on the imaging surface of each re-imaging lens is a photo sensor array (12) and (14) each having a CCD as a light-receiving element, for example.
The arrangement is as follows. Each photosensor array (12), (+
4) As shown in Fig. 6, in the case of so-called front focus, in which the image of the object to be focused is formed in front of the intended focal plane, the images above are close to the optical axis (18) and close to each other;
On the other hand, in the case of rear focus, they are far from the optical axis (18).

When in focus, the distance between two corresponding points of the two images is a specific distance defined by the configuration of the focus detection optical system. Therefore, in principle, the correlator (
16), the focus state can be determined by detecting the distance between the two images.

However, the focus state detected as described above is, for example, when the subject itself has low contrast LI? There may be cases where the temperature is not high (Xf, nertK), and the contrast value is compared with a preset reliability judgment level to determine the in-focus state. A reliability determination method for determining whether or not a focus detection device is reliable and preventing malfunctions of the focus detection device is proposed in Japanese Patent Laid-Open No. 140408/1983.

By the way, the photosensor array used in the above-mentioned type of focus detection device is an accumulation type photosensor array, and when the integration time is set long to correspond to a dark subject, or when the operating temperature of the photosensor array is When the value is high, the dark output increases, and due to this increase in the dark output, noise that causes errors in calculation of the in-focus state also increases. for example,
Even if there is a subject with low contrast, the contrast value will be higher than the judgment level due to the original signal, and the in-focus state will be judged based on this contrast value.
The focus detection device malfunctions.

Conventionally, focusing on this characteristic of the charge storage type photosensor array, in order to prevent malfunctions in the focus detection operation, a cut level has been set for each pixel signal, and this cut level has been set for operation iWIL'V. A system has been proposed in which the noise contained in the pixel signal is captured by changing it in minutes and before the pixel signal is processed (Japanese Unexamined Patent Publication No. 57-72113, U.S. Patent No. 4.437゜743).

However, the pixel signals output from the photosensor array are the original light output with dark output and noise superimposed on it, making it extremely difficult to set the above cut level. However, if the cut level itself is changed, the validity of the cut level itself becomes a problem, and in some cases, even though focus detection operation (focus judgment) is possible, the output level of the pixel signal may be lower than the cut level. This makes it impossible to determine focus.

Therefore, the technical problem of the present invention is to determine the reliability of the in-focus state obtained by processing each pixel signal of a charge storage type photosensor array, such as integration time, operating temperature, etc.
Furthermore, the purpose is to take into consideration the operating conditions of the photosensor array such as gain, set the optimum reliability judgment level, and remove fluctuations caused by the operating conditions to make accurate reliability judgments. .

[Means for Solving the Problems] Therefore, the present invention uses a pair of re-imaging lenses arranged symmetrically with respect to the optical axis of the photographing lens to extract the first image from the subject light beam that has passed through different parts of the photographic lens. ．． a focus detection optical system that forms a second image; a charge storage type photosensor array installed on the imaging surface of the re-imaging lens; and the first photo sensor array of the photosensor array. a focus state detection means for detecting a focus state by arithmetic processing of each pixel signal corresponding to the second image; a determination level setting means for setting a determination level for determining reliability of a detection signal; and whether or not a focused state detected based on the determination level set by the determination level setting means is reliable. The focus detection device includes reliability determination means for determining the reliability determination means.

In the present invention, the operating condition of the photosensor array means any one or a combination of two or more of integration time, operating temperature, and gain.

In addition, the reliability judgment level can be set in various ways depending on the reliability requirements, and as explained below, it can be set to one of the contrast value, correlation level, correlation value 1 pixel signal peak value, and differential signal peak value. It is set corresponding to one or a combination of two or more of these data.

[Example] The present invention will be explained in detail below.

First Embodiment FIG. 1 shows the basic circuit configuration of a focus detection device according to a first embodiment.

As shown in FIG. 1, the focus detection circuit includes an image sensor 20 made of a charge accumulation type photosensor array, and a temperature sensor 21 using, for example, a thermistor, for detecting the operating temperature of the image sensor 20.

The image sensor 20 is driven and scanned by a sensor drive circuit 22, and pixel signals are sequentially output from each photosensor. Timings such as start and end of driving of the sensor drive circuit 22 are controlled by an integral time/auto gain controller (AGC) 23 (hereinafter simply referred to as a sensor controller). Although not specifically shown, the sensor controller 23 has a built-in photodiode that monitors the brightness of the subject, and sets the integration time for the pixel signal according to the brightness of the subject, and within the preset maximum integration time. It also has an auto gain controller function that switches the gain to 2x, 4x, 8x, etc. when the human power level of the pixel signal does not reach a preset threshold.

The pixel signal obtained according to the integration time and gain set by the sensor controller 23 is input to the next stage arithmetic processing circuit 24, and the arithmetic processing circuit 24 calculates the defocus amount DF as described below. , contrast value C1, correlation level signal YM, and pixel signal peak value P are calculated.

Further, the integration time and gain set by the sensor controller 23 are transmitted to the image sensor 2 by the judgment level controller 25 to which the output of the temperature sensor 21 is input.
The determination level controller 25 determines the reliability of each of the contrast value C1 correlation level signal YM and pixel signal peak value P calculated by the arithmetic processing circuit 24, as described below. Set judgment levels CL, YML, and PL.

In order to judge the reliability of the differential gear IjDF obtained by the arithmetic processing circuit 24, the contrast value C1
The correlation level signal YM and the pixel signal peak value P are the first . Second. In the third comparison circuit 26°27.28, the corresponding judgment level CL.

Compare with YML and PL. These first. The second and third comparison circuits 26, 27, and 28 each have C<CL.

When YM≧YML and P<PL, “High” is output to the next-stage OR circuit 29, and “Low” is output otherwise. Only when the outputs of the circuits 26 to 28 are all "L ov", "Low" is output as the reliability determination signal S, and even if at least one of the first to third comparison circuits 26 to 28 is "High" For example, it outputs "High".

The AP/FA controller 30 to which the defocus amount DF and reliability determination signal S are input determines that the defocus amount DF is correct (reliability AP operation (
While performing automatic focusing) or FA display, when reliability judgment signal S is °High', defocus amount DP
As the reliability of the value is low, the necessary manual operation is indicated without using this value.

Next, the calculation of the defocus amount DF, etc. performed by the arithmetic processing circuit 24 will be explained.

Now, in FIG. 5, the outputs, that is, the pixel signals of each pixel a-a+n corresponding to the illuminance distribution of the first image formed on one side (12) of the image sensor 20 by the re-imaging lens 8 are expressed as 3+. pixel b-b corresponding to the illuminance distribution of the second image formed by the other re-imaging lens IO
Let the pixel signal of +n+8 be 'b'''-fb+n+8.

Here, pixel a-a+rl.

bb+n+8 are the 1st and 2nd positions of the image sensor, respectively.
If it is called a region, if it is detected in which part of the second region the first image of the first region most closely matches, the first image of the first region is determined. The image interval between the second images can be determined, and the defocus IDF can be calculated.

Specifically, a total of nine sets of consecutive (n+1) pixels in the second region (b-b+n), (b+ l -b+
n+ 1), -, (b+8 to b+n+8), and each set of pixel signals (Ib-1b+n), (Ib+t-
1b+n+II (Ib+a to Ib+n+8) as the pixel signal 1a- of a total of (n+1) pixels a-a+n in the first area.
1a+. The output with the highest degree of matching is found. To detect such a political change, for example, (,Σ l 1b+i+j Ia+H1)1=0 is calculated for nine combinations (j=0.l,...,8), and the minimum value is found. can be done.

for example! ,~1. The ones that most match +o are Ib+4~l
If b+n+4, then the set of pixels (b+4 to b+n
+4) and (a-a+n) are the first . This is the image interval of the second image, and the number of pixels is (b+4
-a), so if the pixel pitch is d, the image distance x is x = (b+4-a) x d -...(
1).

Now, the first design point when focusing. The distance between the second images is i
o, the defocus amount DF at this time is DF = Cx((b+4-a)xd-i,o) -
It is found in (2). Here, C is a constant specific to the focus detection optical system. Note that this defocus amount DF also includes information on the defocus direction, and if DF is positive, it indicates rear focus, and if it is negative, it indicates front focus.

Here, the first and second images are colored on the image sensor 20.
The method to detect the coincidence of images and find the image interval more accurately is as follows.
The present applicant is Japanese Patent Application No. 58-2622, Japanese Patent Application No. 511! −
It has already been proposed in the patent application No. 113936,
Moreover, since this is not the subject of this application, it will not be discussed in detail here.

Next, a contrast value C1 correlation level signal YM is used for determining the reliability of the above-mentioned differential gear dregs amount DF.

To explain the pixel signal peak value P, first, the pixel signal peak value P is defined as the largest value of all the pixel signals I, ~Ia+n, Ib~'b+n+8 used for calculating the defocus IDF.

That is, P = wax(+,,...1' a+n, I b
, "', I h+n+8)...(3) Moreover, the contrast value C is defined by the following formula.

Next, the correlation level signal YM is now H(4)=Σl Ia+k lb++(-1
1...(5)k=0゜L=! ,...,9 When we define the -growth function, Hain(4) = 5in(H(1), H(2),
..., H(9))...(6) It is defined as a value obtained by normalizing the minimum value of H(i,) obtained as follows by the contrast value C.

YM=Hmin(L)/C= (7) The reason why the contrast value C is used for normalization is that the -growth function H(4) depends on the contrast value.

Note that the correlation level signal YM obtained here is obtained in units of pixel pitches, but in reality, the coincidence function is often minimized at an intermediate position between pixels. Therefore, the minimum value H*1n(i
) can be calculated by interpolation in order to find the position L.sub.in where .

This interpolation calculation method is described in detail in Japanese Unexamined Patent Application Publication No. 126517/1983 filed by the present applicant, so it will not be described in detail here.

Next, a method for setting reliability determination levels CL, YML, and PL performed by the determination level controller 25 will be described.

This judgment level controller 25 has a built-in setting table as shown in Table 1, and the sensor controller 25 has a built-in setting table as shown in Table 1.
3, and the operating temperature of the image sensor 20 output from the temperature sensor 21, and set each judgment level PL, CL according to the setting table.

Set YML. The numerical values given in the table are experimentally determined numerical values, and these numerical values are compared with the contrast value C1 correlation level signal YM1 pixel signal peak value P, respectively.

[Margin below] To explain a little about the table shown in Table 1, the integral time T- is 50 m5 or less, 100 m5 or less, 20
There are a total of 3 stages of 0m5 or less, and for the gain/gain, if the predetermined output level is not reached even when the maximum integration time Tm is set to Ta2 = lOOms, the gain is set to 2.4.
．． Switch to 8x in order, and at lOlooI the gain is 8.
If it is still insufficient, set the maximum integration time Tm to T+a3 =
Switch to 20 Oms and perform integration.

In this case, the gain can be switched from 1 to 8 times in the same way. Since integration at 100 m5 is usually given priority, when the maximum integration time is set to 200 m5, the gain can be switched between 4 times and 8 times. It's going to be a long time ago.

Note that the maximum integration time was set in three stages above, but
In that case, the time setting value is not limited to this and is not limited to
Further, it may be performed in two stages.

In addition, the operating temperature is set in two stages, high and low, at 40°C, but this is simply a decision made in consideration of the operating characteristics of the thermistor used as the temperature sensor 21. However, the operating temperature range may be divided more finely.

In Table 1, a total of 18 combinations of determination levels PL, CL, and YML are set for the three operating conditions.

In the first embodiment described above, a total of three quantities, namely, the contrast value C, the contrast level signal YM, and the pixel signal peak value P, were used simultaneously to determine the reliability of the defocus amount DF, and the determination criteria were set to be the strictest. or 1 amount seems to be either 2
The reliability may be determined based on a combination of two quantities.

<Second Embodiment> In the second embodiment shown in FIG.
The pixel signal sent from the sensor controller 23 is delayed by, for example, 4 pixels by the delay circuit 31, and the difference between the pixel signal delayed by 4 pixels and the pixel signal directly output from the sensor controller 23 is calculated in the difference circuit 32. taken,
The difference signal is input to the arithmetic processing circuit 33.

In this case, the arithmetic processing circuit 33 converts the difference signal into the !th!
Defocus ff1DF is calculated by regarding it as a pixel signal in the case of the embodiment. In other words, each pixel signal summer 8~!
,+. , Ib~■b+. +8 is +4 to Dk=Ik-1
(However, k=a-a+n-4, b-b+n-4)・Da～oa+n-4・Db～Db+n-
4 is the data to be processed, and Ik in the first embodiment
is replaced with Dk and the number of pixels is decreased by 4, and the defocus amount DF, the correlation level signal YM, and the differential signal peak value SP are calculated and output.

Arithmetic processing based on such a difference signal is a method of detecting the correlation or degree of coincidence between the luminance changes of the first and second images, so it is possible to eliminate error factors caused by aberrations of the focus detection optical system, etc. It is advantageous in this respect. It also makes sense to cut low frequency components to absorb the imbalance between the first and second images due to manufacturing variations in the optical system, and to cut the dark output component of the sensor output, that is, the DC component. be. Additionally, high-frequency components that can cause erroneous distance measurements are somewhat cut out.

II. The constant level controller 34 is equipped with a tilt pull for setting the determination level as shown in Table 2, and adjusts the difference according to the integral time and gain output from the sensor controller 23 and the operating temperature from the temperature sensor 21, respectively. Signal peak value SP.

Judgment level SPL for correlation level signal YM.

Set YML.

[Margins below] Table 2 These differential signal peak values SP and correlation level signals YM are shown in Table 1. It is compared with the judgment levels SPL and YML in the second comparison circuits 35 and 36, and SP<SPL, YM≧
When YML, “High” is output respectively, and SP≧
When SPL and YM<YML, "Low" is output, respectively.

When the reliability determination signal S output from the OR circuit 37 is "SOV", the AF/FA controller 30 determines that the defocus amount DF is reliable, and selects AP or F.
While performing operation A, when signal S is "High",
The obtained defocus ff1DF is ignored as unreliable, and manual operation is instructed, for example.

<Third Example> 1st. In the second embodiment, the focus detection circuit is constituted by a hardware circuit, but a microcomputer may be used to perform focus detection and reliability determination by arithmetic processing by a program.

In that case, the system includes an image sensor 20, a temperature sensor 2, a sensor drive circuit, an integral time control circuit, and an auto gain control circuit.

Image sensor 2 with functions such as A/D conversion circuit
0 and the temperature sensor 2■ into digital signals, and an interface circuit 38 that outputs the integration time and gain as digital values, a microcomputer 39, and a defocus amount DF and reliability calculated by the microcomputer 39. The AF/FA controller 40 receives the sex determination signal S as an input.

The microcomputer 39 uses the pixel signal input via the interface circuit 38 and the operating temperature of the image sensor 20 as basic data to execute arithmetic processing according to the flowchart shown in FIG.

When the control is started, two flags HEATF and LCONF are set to "0°" in step O1. This flag is set to "B" when the temperature is high, and "0°" is set when the temperature is low.Furthermore, the flag LCONF is a flag for determining whether or not the defocus IDF is reliable. "Vi" is set when focus cannot be detected during contrast, etc. Then,
In step 02, the integration time of the image sensor 20 driven by the interface circuit 38 is counted. The integration time is set by the interface circuit 38 in accordance with the brightness of the object, as if the integration time is multiple times, and the microcomputer 39 detects the integration time by counting the integration time. In step 03, the pixel signal after the integration time has passed is read, in step 04, the gain at that time is included, and in step 05, the temperature data (operating temperature) detected by the temperature sensor 21 is read. .

Next, in step 06, a defocus amount DF is calculated using the pixel signals output from the image sensor 20. Since this calculation is similar to that described in detail in the first embodiment, further explanation will be omitted. In step 07, a contrast value C is similarly calculated from the pixel signal.

After the above calculation, processing for determining reliability will be performed below.

First, in step 08, it is determined whether the temperature data is equal to or higher than the preset temperature TD, and if it is, in step 09, the integral time Tll1 is compared with the preset first set time Ta1l. Ru. In step IO, a flag (IEATF) or "BI" is set to indicate that the operating temperature is high.In step I!, the judgment level P for the pixel signal peak value P is set.
M is set as L.

On the other hand, when it is determined in step 08 that the operating temperature is lower than the set temperature fTD, N is set as the determination level PL for the pixel signal peak value P for low temperature in step I2. Note that M>N. Further, the numerical values M and N set for the determination level PL may be set in multiple stages depending on the integration time and gain.

In step 13, the pixel signal peak value P from the image sensor 20 is compared with the determination level PL, and if it is larger than the determination level PL, it is determined that the NS ratio of the pixel signal is good, so step 15 Proceed to. On the other hand, if it is smaller than the determination level PL, it is considered that the NS ratio is poor, so in step 14, a flag LCONF indicating the quality of the contrast is set to B.

In the next steps 15 to 29,
Contrast judgment level OL is determined by operating temperature, maximum integration time (Ti2.Ta3), and gain (XI, X2
．． Selectively set according to X4).

Currently, the maximum integration time for the image sensor 20 is 1 serving 2. There are two types of Tm3 (TI2=100ms
, Ta3 = 200m5). Step 21, 22゜2
3, when the operating temperature is low and the maximum integration time is Tm2, the contrast determination levels CL are set to a, b, and c (a<b<C) according to the gain values of 1 and 2.4, respectively. In addition, the maximum integration time is Tm3 when the operating temperature is low.
When , the gain value l.

In step 24, 25, and 26, the contrast determination level CL is set according to 2.4, and b, c, d (b<
c<d).

On the other hand, when the operating temperature is high and the maximum integration time is Ta2, the contrast determination level CL is set to a, c, and d in accordance with the gain, respectively, in steps 24, 25, and 26, which are the same as above. Further, when the operating temperature is high and the maximum integration time is Ta3, the gain values 1, 2, .
4, the contrast judgment level CL is set to c, d, e(
Set c<d<e) (steps 27, 28, 29)
.

Next, in step 30, the contrast value C calculated from the pixel signal is compared with the contrast determination level CL set in steps 15 to 29 above, and if the contrast value C is small, it is determined as low contrast and step 3! The flag LCONF indicating low contrast is set to "B".

In the next steps 32 to 36, the correlation level signal determination level YML is selectively set based on the gain, integration time, and operating temperature. The gain is (X2).

In the case of (×4), set YML as X regardless of other factors. Similarly, when the maximum integration time is Ta2 and when the operating temperature is high, YML is set to X in step 36. on the other hand,
When the gain is (×l), the maximum integration time is T+++2, and the operating temperature is low, YML is set in step 35.
Set to Y.

When the correlation level signal determination level YML is set as described above, the correlation level signal YM obtained in the previous step is
and the correlation level signal judgment level YM are compared, and YM
>YML, it is assumed that there is no significant correlation, and LCONF is set to "Bi" in step 38.

After performing the reliability determination as described above, in step 39, the previously determined defocus amount DF is output to the AP/FA controller 40, and in step 40, the value of the flag LCONF is output to the reliability determination signal S. Output as . When the reliability determination signal S is "0", the AP/FA controller 40 assumes that the manually input defocus jlDF is a reliable value, and performs A F/FA control based on this def-drag amount DF. Execute. Further, when the reliability determination signal S is "bi", the defocus IIDF is not a reliable value, and therefore, for example, manual operation is instructed without performing AF or FA control.

[Effects of the Invention] As is clear from the above explanation, according to the present invention, the reliability determination level can be set according to the operating conditions of the charge storage type photosensor array, so that the reliability of the reliability determination level itself can be improved. Therefore, the reliability of this type of focus detection device can be improved.

[Brief explanation of the drawing]

FIG. 1 is a block diagram of a focus detection circuit according to a first embodiment of the present invention, FIG. 2 is a block diagram of a focus detection circuit according to a second embodiment of the present invention, and FIG. 3 is a block diagram of a focus detection circuit according to a second embodiment of the present invention. FIG. 4 is a flowchart of the program calculation executed by the microcomputer in FIG. 3, FIG. 5 is an explanatory diagram showing the focus detection optical system, and FIG. 6 is an explanatory diagram of the focus detection circuit according to the third embodiment. FIG. 2 is an explanatory diagram schematically showing the detection principle. 20...Image sensor, 2t...Temperature sensor,
23...Sensor controller, 24.33...Arithmetic processing circuit, 25.34...Judgment level controller, 26-28
; 35, 36... Comparison circuit, 29.37... OR circuit.

Claims (17)

[Claims] (1) A focus detection optical system that forms first and second images from the subject light flux that has passed through different parts of the photographic lens using a pair of re-imaging lenses arranged symmetrically with respect to the optical axis of the photographic lens. , a charge storage type photosensor array installed on the image forming surface of the re-imaging lens, and arithmetic processing of each pixel signal corresponding to the first and second images of the photosensor array to determine the in-focus state. a focus state detection means for detecting a focus state; and a judgment level setting means for detecting an operating condition of the photosensor array and setting a judgment level for determining the reliability of a focus state detection signal according to the detected operating condition. A focus detection device comprising: and a reliability determination unit that determines whether or not the detected in-focus state is reliable based on the determination level set by the determination level setting unit. (2) The operating condition of the photosensor array is the operating temperature of the photosensor array, and the reliability determination level is set according to the detected operating temperature. focus detection device. (3) A patent claim characterized in that the operating condition of the photosensor array is the integration time of pixel signals of the photosensor array, and the reliability determination level is set according to the integration time of the detected pixel signals. The focus detection device according to item 1. (4) The operating condition of the photosensor array is a gain with respect to a pixel signal of the photosensor array, and a reliability determination level is set according to the gain with respect to the detected pixel signal. The focus detection device according to item 1. (5) The operating condition of the photosensor array is a combination of any two of the operating time of the photosensor array, the integration time of the pixel signal, and the gain for the pixel signal, and the detected operating time and the integration time of the pixel signal 2. The focus detection device according to claim 1, wherein a reliability determination level is set according to a combination of any two of a gain for a pixel signal and a gain for a pixel signal. (6) The operating condition of the photosensor array is a combination of the operating time of the photosensor array, the integration time of the pixel signal, and the gain for the pixel signal, and the operating condition for the detected operating time, the integration time of the pixel signal, and the gain for the pixel signal is 2. The focus detection device according to claim 1, wherein a reliability determination level is set according to a combination of gains. (7) The focus detection device according to any one of claims 1 to 6, wherein the determination level is set for a contrast value. (8) The focus detection device according to any one of claims 1 to 6, wherein the determination level is set for the correlation between the first and second images. (9) The focus detection device according to any one of claims 1 to 6, wherein the determination level is set to a pixel signal peak value. (10) The focus detection device according to any one of claims 1 to 6, wherein the determination level is set for a differential signal peak value. (11) The focus detection device according to any one of claims 1 to 6, wherein the determination level is set for both the contrast value and the correlation. (12) The focus detection device according to claim 11, wherein the determination level is further set for a pixel signal peak value. (13) The focus detection device according to claim 11, wherein the determination level is further set for the differential signal peak value. (14) The focus detection device according to any one of claims 1 to 6, wherein the determination level is set for both a contrast value and a pixel signal peak value, respectively. (15) The focus detection device according to any one of claims 1 to 6, wherein the determination level is set for both the contrast value and the differential signal peak value, respectively. (16) The focus detection device according to any one of claims 1 to 6, wherein the determination level is set for both the correlation and the pixel signal peak value. (17) The focus detection device according to any one of claims 1 to 6, wherein the determination level is set for both the correlation and the differential signal peak value.