JPH08152552A - Range-finding device and range-finding method - Google Patents

Abstract

PURPOSE: To select optimum range-finding data by comprehensively evaluating line-of-sight information and information concerned in the reliability of the line-of-sight information, and the range-finding information of a camera. CONSTITUTION: A multiple range-finding part 2 performs range-finding at plural points and outputs a range-finding value to a focusing area decision part 1. A line-of-sight detection part 3 detects information on the gazing point of a photographer and outputs the information on the gazing point to the decision part 1. The decision part 1 gives weight to respective focusing areas based on the distance relation of the range-finding information from the range-finding part 2, gives weight to the respective focusing areas based on the output value on the gazing point from the detection part 3, and selects the range-finding data which obtains the highest points in the sum total. In the case where the data obtains the same points, judgment is performed in a specified rule such as center priority, closest range priority or line-of-sight priority.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a distance measuring device and a distance measuring device for performing distance measurement in a plurality of focus detection areas and selecting optimum data from a plurality of distance data detected corresponding to the focus detection areas. It is about the method.

[0002]

2. Description of the Related Art Conventionally, various kinds of information are input to a camera based on, for example, an operation of a dial or a button, and the operating environment becomes complicated as the input information increases.
Utilizing line-of-sight information is an effective solution.

Here, various proposals have heretofore been made regarding a technique for applying the above-mentioned line-of-sight information to automatic focus adjustment (hereinafter abbreviated as AF). For example, JP-A-3-21
Japanese Patent No. 9218 and Japanese Patent Laid-Open No. 4-307506 disclose a technique of selecting an optimum position from a position by the line of sight or a focus point near the position (prior example 1). Then, JP-A-63-94232 and JP-A-4-94232
Japanese Patent No. 221943 discloses a technique of weighting each area and setting a focus point according to the movement of the line of sight (prior example 2).

Further, Japanese Unexamined Patent Publication No. 4-138434 discloses a technique for evaluating the reliability of the line-of-sight signal and selecting the focus point by a method that does not use the line-of-sight information when the reliability is low. (Prior example 3). Besides this, A
Regarding the F-only algorithm, a technique of setting the focus point by the multi-distance-only algorithm has been disclosed (Prior example 4).

[0005]

However, regarding the application of the sight line information to the AF, in the sight line detection by the camera,
Assuming that the photographer looks at the main subject at the time of shooting,
AF is performed based on the focus point information that the photographer is gazing at, but with the current compact camera, if you look into the viewfinder as the camera becomes smaller, you can almost see the entire viewfinder without moving your eyes. This is a problem because the eyes are often fixed near the center when no particular attention is paid.

Further, in the multi-AF algorithm,
Since the optimum focus point is selected with a high probability, under such circumstances, if the main subject is determined by adding the line-of-sight information, it may adversely affect the optimally photographed image. , It was desired to add line-of-sight information only to failed photographs.

Further, the above-mentioned prior art technique has the following problems. That is, in the above-mentioned first example, since the photographer may make an incorrect selection when the photographer does not pay attention to the main subject because it depends only on the line-of-sight information, the focus points are discretely arranged in the case of silver halide photography. However, one focus point is in a fairly narrow area, and the probability that the focus point exists at the point of interest is low, so it was doubtful that the optimum focus point could be selected.

Further, in the above-mentioned prior art example 2, particularly in a compact camera which is downsized, it is not necessary to open and move the eye when gazing at the viewfinder, so that the eye movement does not always correspond to the main subject. It was a thing. Further, in the above-mentioned prior art example 3, although it is possible to set the optimum focus point with a high probability by the multi-range finding algorithm, the intention of the photographer cannot be reflected. Further, in the above-mentioned prior art example 4, in order to improve the reliability of the line-of-sight signal, high-level detection using an area sensor is required, resulting in a high cost and a long processing time, which is a problem.

The present invention has been made in view of the above problems, and an object thereof is to comprehensively evaluate line-of-sight information, information relating to the reliability of the line-of-sight information, and distance measurement information of a camera, The optimum distance data is selected from a plurality of distance data detected corresponding to the focus detection area.

[0010]

In order to achieve the above object, a distance measuring apparatus according to a first embodiment of the present invention is a multi-point distance measuring apparatus that measures distances in a plurality of predefined focus detection areas. Means and a first weighting for the plurality of focus detection areas based on output data of the multipoint distance measuring means.
The weighting means, the eye gaze direction of the observer is detected, the eye gaze evaluation means for evaluating the reliability at the time of gaze direction detection, and the plurality of focus detection areas for the plurality of focus detection areas according to the output of the eye gaze evaluation means. Based on the added value of the second weighting means for weighting 2, the output of the first weighting means, and the output of the second weighting means, detection is made corresponding to the plurality of focus detection areas. Data selection means for selecting only one distance data from a plurality of distance data.

In the distance measuring device according to the second aspect, the second weighting means detects the line-of-sight direction of the observer based on a plurality of predefined line-of-sight detection regions, and the plurality of focus detection regions are detected. Is weighted by multiplying the reliability coefficient at the time of detection.

Further, in the distance measuring method according to the third aspect, according to a plurality of distance data from the multi-point distance measuring means, a distance weight value corresponding to each distance is given to each focus detection area, and a predetermined value is given. The eye gaze direction of the observer is determined based on the eye gaze detection area, and the eye gaze weight value based on the product with the reliability coefficient at the time of gaze detection is given to each focus detection area, and the eye gaze weight value and the distance weight value are Is evaluated, and only one distance data is selected from the plurality of distance data based on the one that gives the maximum value among them.

[0013]

That is, in the distance measuring apparatus according to the first embodiment of the present invention, distance measurement is performed in a plurality of focus detection areas defined in advance by the multipoint distance measuring means, and the first weighting means performs the distance measurement. First weighting is performed on the plurality of focus detection areas based on the output data of the point distance measuring means, and the line-of-sight evaluation means detects the line-of-sight direction of the observer and the reliability at the time of detecting the line-of-sight direction is evaluated. Then, the second weighting means performs the second weighting on the plurality of focus detection areas according to the output of the line-of-sight evaluation means, and the data selection means outputs the first weighting means and the second weighting means. Only one distance data is selected from the plurality of distance data detected corresponding to the plurality of focus detection areas based on the added value with the output of the weighting means.

In the distance measuring device according to the second aspect,
The second weighting means detects the line-of-sight direction of the observer based on a plurality of predefined line-of-sight detection regions, and performs weighting by multiplying the plurality of focus detection regions by a reliability coefficient at the time of detection. Be seen.

Further, in the distance measuring method according to the third aspect,
According to a plurality of distance data from the multi-point distance measuring means, a distance weight value corresponding to each distance is given to each focus detection area, and the line-of-sight direction of the observer is determined based on the predetermined line-of-sight detection area. A line-of-sight weight value based on the product of the reliability coefficient at the time of line-of-sight detection is given to each focus detection area, the added value of the line-of-sight weight value and the distance weight value is evaluated, and the one giving the maximum value among them is given. Based on this, only one distance data is selected from the plurality of distance data.

[0016]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS First, the concept of the present invention will be described with reference to FIGS. 1 (a) to 1 (c). A camera using the distance measuring device shown in FIG. 1A includes a multi-distance measuring unit 2 capable of measuring a plurality of points, a line-of-sight detecting unit 3 capable of detecting a gazing point of a photographer, and a plurality of focus areas. The focus area determination unit 1 determines an optimum focus area. The multi-range finding unit 2 and the line-of-sight detection unit 3 are both connected to the focus area determination unit 1.

In such a configuration, the multi distance measuring unit 2
Measures a plurality of points and outputs the measured distance values to the focus area determining unit 1. The line-of-sight detection unit 3 detects the gaze point information of the photographer and outputs the gaze point information to the focus area determination unit 1. The focus area determination unit 1 weights (gives a score) to each focus area based on the perspective relationship of the distance measurement information of the multi-distance measurement unit 2, and determines each based on the output value of the gaze point of the line-of-sight detection unit 3. The focus areas are weighted and the focus area with the highest score in each sum is selected. Further, in the case of the same score, the judgment is made based on a predetermined rule such as center priority, close-up priority, and gaze priority.

As described above, in the distance measuring device shown in FIG. 1A, each focus point is weighted according to the distance measuring information obtained by the multi-distance measuring, and the finder obtained by the line-of-sight detection. Each focus point is given a weight based on the information corresponding to the position inside, and the focus point having the highest weight value is selected.

A camera using the distance measuring device shown in FIG. 1B includes a multi-distance measuring unit 2 capable of measuring a plurality of points, a line-of-sight detecting unit 3 capable of detecting a gazing point of a photographer, and the line-of-sight detection. The reliability determination unit 4 that determines the reliability of the signal output from the unit 3 and the focus area determination unit 1 that determines an optimal focus area from a plurality of focus areas are configured. Are connected to the focus area determination unit 1, and the reliability determination unit 4 is connected to the line-of-sight detection unit 3 and the focus area determination unit 1.

In such a configuration, the multi distance measuring unit 2
Measures a plurality of points and outputs the measured distance values to the focus area determining unit 1. The line-of-sight detection unit 3 detects the gaze point information of the photographer and outputs the gaze point information to the focus area determination unit 1. The reliability determination unit 4 evaluates the reliability of a plurality of signals from the detection signal of the line-of-sight detection unit 3 and outputs the degree of reliability to the focus area determination unit 1. The focus area determination unit 1 weights (gives a score) to each focus area based on the perspective relationship of the distance measurement information of the multi-distance measurement unit 2, and determines each based on the output value of the gaze point of the line-of-sight detection unit 3. The focus area is weighted, and the reliability information (value between 0.0 and 1.0) of the reliability determination unit 4 is put on the score of each focus area weighted by the line of sight to obtain the distance measurement information. The focus area having the highest score is selected from the sum of the weighted value and the value weighted by the line-of-sight information multiplied by the reliability information. Further, in the case of the same score, the judgment is made based on a predetermined rule such as center priority, close-up priority, and gaze priority.

As described above, in the distance measuring device shown in FIG. 1B, each focus point is weighted according to the distance measuring information obtained by the multi-distance measuring, and the finder obtained by the sight line detection. Each focus point is given a weight based on the information corresponding to the position inside, and the weight of each focus point obtained from the above-mentioned line-of-sight information is multiplied by a predetermined coefficient according to the degree of reliability by judging the reliability of the line-of-sight signal. The weight of each focus point for which a new weight is set is determined by an addition value of the weight obtained from the distance measurement information and the weight obtained from the line-of-sight information in which reliability is added.

A camera using the distance measuring device shown in FIG. 1C includes a multi-distance measuring unit 2 capable of measuring a plurality of points, a line-of-sight detecting unit 3 capable of detecting a gazing point of a photographer, and the line-of-sight detection. Reliability determination unit 4 for determining the reliability of the signal output from the unit 3, photometric unit 5 for detecting the brightness of the subject, strobe unit 6, strobe control unit 7 for setting the conditions for light emission of the strobe unit 6, film Is composed of an ISO detection section 8 for detecting the ISO sensitivity and a focus area determination section 1 for determining an optimal focus area from a plurality of focus areas. The multi-ranging section 2 and the line-of-sight detection section 3 are both focus area determination section 1
The reliability determination unit 4 is connected to the line-of-sight detection unit 3 and the focus area determination unit 1, and the flash control unit 7 is connected to the IS.
The O detection unit 8, the photometry unit 5, the strobe unit 6 and the focus area determination unit 1 are connected, and the strobe unit 6 is connected to the focus area determination unit 1.

In such a configuration, the multi distance measuring unit 2
Measures a plurality of points and outputs the measured distance to the focus area determination unit 1. The line-of-sight detection unit 3 detects the gaze point information of the photographer and outputs the gaze point information to the focus area determination unit 1. The reliability determination unit 4 evaluates the reliability of a plurality of signals from the detection signal of the line-of-sight detection unit 3 and outputs the degree of reliability to the focus area determination unit 1. The photometric unit 5 measures the brightness of the subject and outputs the measured value to the strobe control unit 7. The ISO detection section 8 outputs the ISO sensitivity of the film to the strobe control section 7. The strobe control unit 7 determines the light emission condition based on the brightness information of the photometry unit 5 and the ISO sensitivity value of the ISO detection unit 8, outputs the presence / absence of strobe light emission and a strobe guide number to the focus area determination unit 1, and outputs the focus information. The amount of strobe light is controlled according to the distance measurement information of the focus area determined by the area determination unit 1. Then, the focus area determination unit 1 weights each focus area (gives a score) based on the perspective relationship of the distance measurement information of the multi-distance measurement unit 2.
Then, each focus area is weighted based on the output value of the gazing point of the line-of-sight detection unit 3, and the reliability information (value between 0.0 and 1.0) of the reliability determination unit 4 is viewed. Each focus area is weighted by the score of each focus area, and each focus area is weighted according to the strobe light emission amount from the strobe control unit 7, the guide No and the distance measurement information, and the distance measurement information is weighted. The focus area with the highest score is selected from the sum of the weighting value in step (1) and the value weighted by the line-of-sight information multiplied by the reliability information and the total weighting by the strobe light amount. Further, in the case of the same score, the judgment is made according to a predetermined rule such as center priority, close-up priority, and gaze priority.

As described above, in the distance measuring device shown in FIG. 1C, each focus point is weighted according to the distance measuring information obtained by the multi distance measuring, and the distance measuring information and the stroboscopic light emission are performed. Conditions and weighting based on the strobe distance are added, and each focus point is given weight based on the information corresponding to the position in the viewfinder obtained by line-of-sight detection, and the reliability of the line-of-sight signal is judged to determine the degree of reliability. Accordingly, the weight of each focus point obtained from the line-of-sight information is multiplied by a predetermined coefficient to set a new weight.
The weight of each focus point is determined by the added value of the weight obtained from the distance measurement information and the weight obtained from the line-of-sight information in which reliability is taken into consideration.

The present invention is basically divided into the above-mentioned three modes, and the embodiments of the present invention will be described in more detail below with reference to the drawings. Regarding the method of line-of-sight detection, the center of gravity of the red-eye image and corneal reflection (first Purkinje image)
Is adopted (Japanese Patent Publication No. 61-59132,
JP-A No. 2-5).

First, FIG. 2 shows the structure of a camera which employs the distance measuring apparatus according to the first embodiment of the present invention, and will be described. The first embodiment is used for selecting three focus areas of the camera, and one-dimensional detection is performed in the line-of-sight detection.

As shown in the figure, the LED circuit 13 for projecting a luminous flux to the eye includes a finder / line-of-sight optical system 15 and a CP.
The line sensor circuit 14 which is connected to U11 and receives the reflected light flux from the eye includes a finder / line-of-sight optical system 15 and a CPU.
11 is connected. The CPU 11 is an LE for line-of-sight detection.
It is connected to a D circuit 13 and a line sensor circuit (a device capable of detecting peaks such as SIT) 14 and also has a multi-A.
F circuit 12, lens drive circuit 16, photometric circuit 17, and 1s
It is connected to the t, 2nd release switch.

With such a configuration, the LED circuit 13 is C
In response to the control signal from the PU 11, the LEDs arranged at the positions for generating red eyes (which generate reflected light from the fundus at the sensor position) and non-red eyes are caused to emit light, and the emitted luminous flux is passed through the viewfinder / gaze optical system 15. Send it to your eyes. The line sensor 14 receives the reflected luminous flux of the projected luminous flux from the eye through the finder / line-of-sight optical system 15, and outputs a detection signal to the CPU.
Send to 11. The multi-AF circuit 12 measures distances at a plurality of points and sends distance measurement information to the CPU 11. The photometric circuit 17 sends the brightness information of the subject to the CPU 11. The lens driving circuit 16 drives the lens according to the control signal from the CPU 11.
The CPU 11 controls the line-of-sight detection system (the LED circuit 13, the line sensor 14), detects the line-of-sight direction from the obtained signal, performs reliability determination in consideration of the information of the photometric circuit 17, and determines the line-of-sight direction and reliability. Each focus area is weighted by the line-of-sight, AF is weighted by the distance measurement information from the multi-AF circuit 12, the optimum focus area is selected, and the lens drive amount is calculated by the corresponding distance measurement information. The drive circuit 16 is controlled.

Here, the finder / line-of-sight optical system 15 is used.
The detailed configuration of is as shown in FIG. That is, the light flux from the subject is guided to the eye via the beam splitter 153 including the infrared cut filter 152, the main mirror MM, the infrared half mirror HM0, and the infrared half mirror HM1. The light source LED0 of the LED circuit 13 for detecting the line of sight,
The light flux from the LED 1 is guided to the eye through the beam splitter 153 including the infrared half mirror HM0 and the infrared half mirror HM1, and the reflected light from the eye is reflected by the beam splitter 153 including the infrared half mirror HM1. The light is guided to the line sensor circuit 14 via the light cut filter 151a and the line-of-sight optical system 154. The light flux from the subject is restricted by visible light cut filters 151b and 151c disposed on the front surface (subject side) of the infrared cut filter 152 and the rear surface (eye side) of the beam splitter 153. Also,
A display LED 156 and a parallax correction liquid crystal element 155 are arranged near the primary image plane. The red-eye generating light-projecting and light-receiving axes may be displaced from the optical axis of the finder.

Light sources LED0 and LED1 for detecting the line of sight
The state of the reflected light from the eye due to is as shown in FIG. That is, FIG. 4A shows the state of the signal detected by the light emitted from the LED 0, and it can be seen that the corneal reflection signal is superimposed on the red-eye (fundus reflex) state. 4 (b) is L
The state of the signal detected by the projection of the ED1 is shown, and it can be seen that no red eye (fundus reflex) has occurred. In the present embodiment, the red eye (fundus reflex) signal is extracted by the difference processing of the signals detected in FIGS. 4A and 4B, and the center position of the pupil is obtained by detecting the center of gravity.

FIG. 5 is a view showing a masking state of the finder. As shown in the figure, a filter that transmits only the luminous flux (infrared luminous flux) for line-of-sight detection is provided outside the viewable area of the photographer. As a result, even with a camera having a high degree of freedom of the eye point, it is possible to prevent the pupil from being eclipsed by the finder and the accuracy of detecting the center of gravity of the red-eye image from being lowered. That is, the detection system can always see the entire eye of the photographer.

Hereinafter, referring to the flow chart of FIG.
A sequence of the camera using the distance measuring device of the first embodiment will be described. When the camera sequence starts (step S
1), the CPU 11 performs initialization. here,
The focus area is set to the center (step S2).
Then, the CPU 11 turns on / off the 1st release switch.
It is determined to be OFF (step S3), and if the 1st release switch is OFF, the process proceeds to step S12,
When the 1st release switch is ON, the photometric circuit 17
And the multi-focusing by the multi-AF circuit 12 (3-point ranging of FL, FC, FR) is started (step S4).

Further, the CPU 11 judges the necessity of the sight line detection (step S5), and when the distance measurement information is substantially the same for all three points, or when the sight line detection is not necessary such as when the self-timer mode is set, The flag Fp is set to "1" and the process proceeds to step S8 (step S7). On the contrary,
When the line-of-sight detection is necessary, the CPU 11 executes a later-described subroutine "line-of-sight detection" to detect the line-of-sight (step S6).

Subsequently, the CPU 11 executes a later-described subroutine "focus area selection" to perform multi-AF distance measurement and distance measurement information selection (step S8). And
The ON / OFF of the 1st release switch is determined again (step S9), and the 1st release switch is turned off.
If it is F, go to step S12 and if the 1st release switch is ON, turn the 2nd release switch ON.
/ OFF is determined (step S10). And 2n
If the d release switch is OFF, the process returns to step S9, and if the 2nd release switch is ON, CP
U11 executes a shooting sequence such as lens driving by the lens driving circuit 16, exposure determination, and exposure (step S1).
1) and this sequence is ended (step S12).

After first sight line detection in synchronism with turning on of the 1st release switch (step S3) (rough photometry is performed simultaneously), photometry by the photometry circuit 17 and distance measurement by the multi AF circuit 12 are performed. , The necessity of the line of sight may be judged. Further, it is needless to say that line-of-sight detection, photometry, and distance measurement may be performed at the same time.

Next, the sequence of the subroutine "line-of-sight detection" executed in step S6 of FIG. 6 will be described with reference to the flowchart of FIG. When this subroutine "line of sight detection" is started (step S21), the CPU 11 initializes. Here, the flags Fq, Ft, F
k, Fe1, and Fe2 are set to "0" (step S2
2). Then, the LED0 is used to emit light (light emission 0) to generate a red eye (step S23), and a subroutine "signal detection" described later is executed to perform signal detection (step S24).

Subsequently, the CPU 11 determines the state of the flag Ft related to the integration time (step S25). If the flag Ft is "1", the process proceeds to step S35.
On the other hand, when the flag Ft is “0”, the CPU 1
1 stores the detection integration time ts in t0 (step S2
6) The image signal (E (i)) is stored in D0 (i) (i: 1 to N) (step S27).

Subsequently, the CPU 11 carries out light projection (light projection 1) which does not cause the red eye by the LED 1 (step S2).
8) The subroutine "signal detection" is executed to perform signal detection (step S29). Then, the CPU 11 determines the state of the flag Ft related to the integration time (step S3
0), if the flag Ft is "1", step S3
Go to 5. On the other hand, when the flag Ft is "0", the CPU 11 stores the detection integration time ts in t1 (step S31), and the image signal (E (i)) is D1 (i).
The data is stored in (i: 1 to N) (step S32).

Subsequently, the CPU 11 executes a subroutine "signal processing" which will be described later, and determines the reliability of the detection signal and calculates the line-of-sight direction (step S33). Then, the state of the flag Fk relating to the reliability is judged (step S3
4) If the flag Fk is "0", the process proceeds to step S36. If the flag Fk is "1", the process proceeds to step S36.
In step S35, the line-of-sight reliability flag Fq is set to "1" (step S35). In this way, the display of the line-of-sight position is
ED is performed (step S36), and this sequence is ended (step S37).

The state of the line-of-sight position display in step S36 is as shown in FIG. In the figure,
The line-of-sight detection area and the focus area are shown. The viewfinder has display LEDs corresponding to each area (may also be used as another camera display) α, β, γ and warning LEDs
Is arranged.

Next, the sequence of the subroutine "signal detection (sensor control)" executed in steps S24 and S29 of FIG. 7 will be described with reference to the flowchart of FIG. When the subroutine "signal detection" is started (step S41), the CPU 11 first initializes. Here, the flag Ft related to integration is set to "0", the integration is reset, and the timer is reset / started (step S42).

Then, the integration is started (step S4
3), limiter determination of integration time is performed (step S4)
4). Here, when t> Tmax (Tmax is a predetermined value of the limiter time of integration), the process proceeds to step S49,
The flag Ft is set to "1" (step S49), and this sequence ends (step S50).

On the other hand, when t≤Tmax, the peak monitor level of the sensor is determined (step S45). Then, when the peak determination is not made, the process returns to step S44, and when the peak determination is made, the integration (integration time ts) is ended (step S46).

Then, the integration time ts is judged (step S47), and if ts <Tmin, step S4.
9, and if ts ≧ Tmin, signal E (i)
(I = 1 to N) reading and processing (AGC, A / D, etc.)
Is performed (step S48), and this sequence is ended (step S50).

The integration control may be performed once at a predetermined time, and based on the result, a different integration time from the first time may be set and performed repeatedly. However, it is advisable to continuously perform the integration time of the pair to be detected in a time division manner with substantially the same integration time (correction coefficient error range). Further, it is preferable to adjust the projected light quantity in advance so that the received light quantity of the noise component becomes substantially the same in the integration time. Further, the detection when the LED 0 emits light (light emission 0) may be performed based on the sequence of FIG. 8, and the detection when the LED 1 emits light (light emission 1) may be performed by the method shown in FIG. In the sequence of FIG. 9,
It is not necessary to detect the peak (step S5).
1 to S60).

Next, referring to the flowchart of FIG.
The sequence of the subroutine "display" executed in step S36 of FIG. 7 will be described. When each area is selected, the corresponding LED α, β, γ emits light, and when the flag Fp is “1”, the warning LED (indication that the line-of-sight information cannot be used) emits light. Then, these displays disappear when the normal camera sequence is exited due to the end of shooting or the release of the 1st release switch (steps S61 to S69).

Next, referring to the flowchart of FIG.
The sequence of the subroutine "signal processing" executed in step S33 of Fig. 7 will be described. When this subroutine "signal processing" is started (step S71), after the subroutine "ghost processing" described later is executed (step S72), variables h and r are set according to the external brightness (step S73).

This variable h is, as shown in FIG.
It is set stepwise according to the photometric value Ex. The value of the variable h decreases as the subject becomes brighter, while the variable r does not depend on the brightness of the subject. This variable r
Is related to the reflection image of the cornea and does not have to depend on the brightness of the subject, but the variable h is related to the light reflected from the fundus passing through the pupil and is greatly affected by the brightness.

Subsequently, the data D0 (i) is converted into F (i).
(I: 1 to N) (step S74), the subroutine "feature extraction 1" described later is executed, and the image F
Feature points are extracted from (i) (variable k is set to 0) (step S75). Subsequently, the state of the reliability flag Fe1 in the feature point extraction is determined (step S76), and if the flag Fe1 is "1", the process proceeds to step S84.

In step S76, the flag Fe
When 1 is "0", the previously detected MAX position x is stored in x0, and the average value Av is stored in Av0 (step S77). Then, the data D1 (i) is converted to F (i)
(I: 1 to N) (step S78), the subroutine "feature point extraction 1" is executed to extract feature points from the image F (i) (variable k is set to "1") (step S
79).

Then, the reliability flag Fe1 in the feature point extraction is determined again (step S80), and the flag F is determined.
If e1 is “1”, the process proceeds to step S84,
When the flag Fe1 is 0, the previously detected average value Av
Is stored in Av1 (step S81). Then, the difference values Avd (Avd = a × Av0−b × AV1; a, are added to the detected average values Av0 and Av1 by giving a predetermined weight.
b predetermined value) is performed (step S82). Subsequently, the suitability determination of the difference processing result Avd (determination of abnormal light based on the total value) is performed (Ka and Kb are predetermined values) (step S8).
3). If Ka <Avd <Kb is not satisfied, the process proceeds to step S84, the flag Fk is set to "1" (no reliability), and then the sequence is exited (step S93).

On the other hand, if Ka <Avd <Kb, the subroutine "feature extraction 2" described later is executed (step S85), and the reliability flag Fe2 in the feature point extraction is determined (step S86). Then, if the flag Fe2 is "1", the process proceeds to step S84, and the flag Fk for judging the reliability of the signal in the signal processing is set to "1".
After setting it to "not reliable", the sequence is exited (step S93).

When the flag Fe2 is "0", the MAX position x previously detected by the feature extraction is stored in x1 (step S87), and the eye movement is determined at x0 and x1 (step S87). S88), if | x0-x1 | <[epsilon], the process proceeds to step 84, and if | x0-x1 | <[epsilon], a subroutine "signal conversion" described later is executed, and the signal conversion is performed by the variable h. Perform (step S89).

Then, difference processing D (i) (D (i) = a * D0 (i) -b * D1) at D0 (i) and D1 (i).
(I)) is performed (step S90), and the gravity center is detected (centroid value xd) at D (i) (step S91). Then, the evaluation value x (x = xd-x0 + S; S is a correction value, which depends on the information value of the right eye or the information of the left eye or the personal information value) is returned (step S92), and the sequence is returned to the main routine (step S93).

In step S82, the difference process is performed. If the difference time is 0 or less, it is better to set the pixel processing value to 0. Next, the sequence of the subroutine "ghost processing" executed in step S72 of FIG. 12 will be described with reference to the flowchart of FIG.

When the subroutine "ghost processing" is started (step S101), the ghost data Dk0
(I) and Dk1 (i) are read (step S1)
02). The ghost data Dk0 (i), Dk1
(I) is data obtained by causing LED0 and LED1 to emit light in the dark and integrating for a predetermined integration time td, and is assumed to be recorded in advance. Subsequently, the ghost data is normalized from the data for detecting the line of sight by the integration time to obtain the difference (step S103), and the present sequence is exited (step S104).

Next, referring to the flowchart of FIG.
The sequence of the subroutine "feature extraction 1" executed in steps S75 and S79 of FIG. 12 will be described. In this sequence, a continuous value jp of a value equal to or more than a predetermined value is detected from the detected data
Also, the maximum value MAX, the maximum value position x, the average value Av, and the contrast value C are detected and evaluated.

That is, when the subroutine "feature extraction 1" is started (step S111), initialization is performed.
Here, the variables j, i, jp, S, MAX, x, C are set to "0", and the flag Fe is set to "0" (step S1).
12).

Then, the following processing is performed from i = 1 to N. That is, the detection data F (i) is compared with the predetermined value H, and the maximum value (continuous length) at which data having the predetermined value H or more continuously exists is stored in jp (steps S113, S).
114, S124 to S126, S122, S123).
Then, the total sum value S is calculated (step S115), the maximum value MAX is detected, and the maximum value position x is set (steps S117 and S118). Then, from the point where the variable i becomes 2 or more, the sum value C of contrast values (C = C + | F
(I) -F (i-1) |) is calculated (step S1).
20).

The value thus detected is evaluated as follows. That is, the continuous length jp is determined (step S127). First, in the determination of strong abnormal light, jp <L
If not, the process proceeds to step S131, and the flag Fe1 regarding the reliability of the detection signal in the feature extraction is set to "1".
Set to (unreliable). When jp <L, the contrast value C is evaluated (step S128).

Next, in the blink state detection (the contrast decreases when blinking), if C> Cp is not satisfied, the process proceeds to step S131, and the flag Fe1 relating to the reliability of the detection signal in the feature extraction is set to "1" (reliability. None), and if C <Cp, the average value Av is calculated (step S129), and the maximum value MAX and the average value Av are set.
Is evaluated (step S130).

Then, MAX-Av> ε (k) is detected in the blinking state detection (the cornea reflection does not occur at the time of blinking).
(Ε (k) is a predetermined value, k is 0 or 1, and ε (0) and ε
If (1) may not be the same or different), the process proceeds to step S131, the flag Fe1 regarding the reliability of the detection signal in the feature extraction is set to "1" (no reliability), and MAX-Av> ε ( In the case of k), this sequence is exited and the process returns to the main routine (step S133).

Next, referring to the flowchart of FIG.
The sequence of the subroutine "feature extraction 2" executed in step S85 of FIG. 12 will be described. In this sequence, the data area is narrowed by a variable r (here, r is a fixed value), and the maximum value MAX and the maximum value position x are detected from the narrowed data and evaluated. Particularly, in the non-red-eye state, the level of the white eye existing at the edge becomes high, and the maximum value may be erroneously detected. Therefore, the detection range is limited and the influence of the white eye is reduced.

When the subroutine "feature extraction 2" is started (step S151), initialization is performed. Here, the variables i, MAX, x, and S are set to "0", and the flag Fe is set.
2 is set to "0" (reliable) (step S15)
2). Then, in the range of i = 1 to N, further x0-r to x
Maximum value detection is performed up to 0 + r.

That is, the detection range is limited (step S
153 to S155, S160 to S162), maximum value MA
X, the maximum value position x is detected (steps S156 to S156).
158). Then, in the range of i = 1 to N, x0-r
The average value Av is detected up to x0 + r, and reliability determination is performed (step S163).

Subsequently, the maximum value MAX and the average value Av are compared (δ is a predetermined value) (step S164). here,
If MAX-Av> δ, the process proceeds to step S165, and if not MAX-Av> δ, the flag Fe2 is set to "1" (unreliable) (step S165).
This sequence is exited (step S166). Here, the mask processing x0-r to x0 + r is performed as shown in FIG. 27C, but the smooth mask processing (see FIG. 27) is performed.
(See (d)) may be performed (see FIG. 27 (b)).

Next, referring to the flowchart of FIG.
The sequence of the subroutine "signal conversion" executed in step S89 of Fig. 12 will be described. When the subroutine "signal conversion" is started (step S171), initialization (i = 1) is performed (step S272), and i
= 1 to N, when i <x0-h or i> x0 + h, D0 (i) and D1 (i) are set to 0, and this sequence is exited (steps S173 to S179). here,
Although the mask processing is performed in the shape as shown in FIG. 27C, a smooth mask as shown in FIG. 27D may be used. By using a smooth mask, wider information can be taken in S / N well.

Next, referring to the flowchart of FIG.
The sequence of the subroutine "focus area selection" executed in step S8 of FIG. 6 will be described. When the focus area selection is started (step S181),
The flag Fp indicating the need for the line-of-sight information is determined (step S182), and the flag Fp is "1" (no need for line-of-sight information).
In the case of, weighting (a (0) = d (0) = e (0) = 3) for each focus area a, b, e is performed (step S185). Then, the reliability coefficient k is set to "1" and the weighting of each focus area is multiplied by the reliability coefficient (a
(1) = k × a (0), d (1) = k × d (0), e
(1) = k × e (0), k = 1) (step S18
6).

In step S182, if the flag Fp is "0" (requires line-of-sight information), a subroutine "line-of-sight weighting" to be described later is executed to perform line-of-sight weighting (step S183), and the subroutine to be described later will be described. "Reliability weighting" is executed to perform reliability weighting (step S184).

Subsequently, a subroutine "distance weighting" which will be described later is executed to perform weighting by distance (step S187), and the weighting process for each focus area is performed (a
= A (1) + a (2), b = (1) + b (2), c = c
(1) + c (2)) (step S188). Then, the weight value is determined (step S189), and if the same weight does not exist in the selection conditions, the focus area having the largest weight is selected (step S19).
0).

When the same weight value exists in the selection conditions (there are a plurality of MAX weights), it is determined whether or not the central a is included (step S191). And the center a
If is included, the center a is selected (step S19
6) If the center "a" is not included, the focus areas on the closest sides of both sides are selected (steps S192, S).
193, S194).

The drive amount of the photographing lens is calculated based on the distance measurement information of the focus area thus selected (step S197), and the parallax of the viewfinder is corrected by the liquid crystal based on the distance information of the selected focus area. (Step S198), this sequence is exited (step S199). The parallax correction may be performed by a lab process by recording trimming (parallax) information on the film in accordance with the finder optical system by magnetic recording or the like, instead of performing it by the finder optical system.

Next, referring to FIG. 19 and FIG.
The sequence of the subroutine "line-of-sight weighting" executed in step S183 of FIG. FIG. 19 shows a state of the line-of-sight detection area in which line-of-sight weighting is performed.
FIG. 19A shows the divided areas adopted in this embodiment,
19 (b) and 28 (a) show other examples of division.

In this subroutine "line-of-sight weighting", the focus areas are weighted according to the detection areas A to F shown in FIG. The weighting factors a (0), b (0), c (0) when each area is selected are as shown in the following table (steps S201 to S21).
5).

[0075]

[Table 1]

Next, referring to the flowchart of FIG.
The sequence of the subroutine "reliability weighting" executed in step S184 of FIG. 18 will be described. In this subroutine, the flag Fq of the reliability of the line-of-sight information is determined. If the flag Fq is "1" (no reliability), the reliability coefficient k is set to 0.5, and the flag Fq is "0". In the case of (reliable), the reliability coefficient k is set to "1" (steps S221 to S224). Then, the weighted value of the line of sight of each focus area is multiplied by the reliability coefficient (a
(1) = k × a (0), b (1) = k × b (0), c
(1) = k × c (0)), this sequence is exited (steps S225 and S226).

As for reliability, a plurality of flags (for example, flag Fk (image detection image abnormality flag) and flag Ft (integration abnormality flag) instead of flag Fq are evaluated separately, and different flag states are combined to obtain different reliability values. A plurality of reliability coefficients k1 and k2 may be set, and an example of weighting by the line of sight in the case of the arrangement shown in Fig. 28 (a) is shown in Fig. 28 (b).

Next, referring to the flowchart of FIG. 22,
The sequence of the subroutine "distance weighting" executed in step S187 of Fig. 18 will be described. When the subroutine "distance weighting" is started (step S23)
1) Convert each distance measurement information into a distance block value (almost the same distance has the same block value) (step S23).
2). FIG. 23 shows how the distance is converted into the block value. The block values are La, Ld, and Le corresponding to each focus area (a, d, e). The relationship between the block values and each focus area weight value are set, and this sequence is exited (steps S233 to S259). The setting table is shown in FIG.

Then, as shown in FIG. 25, the liquid crystal arranged near the primary image forming plane of the finder shields the light except for the photographing area according to the distance measurement information of the focus area thus selected.

As described above, the first embodiment sets the optimum distance measuring point based on the distance measuring information, the line-of-sight information, and the information including the reliability of the line-of-sight signal with a simple algorithm. can do.

Although the embodiments of the present invention have been described above, it is needless to say that the present invention is not limited to these embodiments and that various improvements and modifications can be made. For example, the line-of-sight detection method may use an area sensor to improve reliability. Then, the reliability judgment is performed by blinking and moving the eyes, but it is more preferable to combine the judgment of the contact between the camera and the face (the eye is pressed against the camera finder) and the judgment of the eye glasses. Further, if the calibration is performed for each individual, the detection accuracy is further improved. In addition, display the calibration during the sequence (at the start of the sequence, or 1st
Immediately after detecting the line of sight).

The detection area is further divided into a passive type two-dimensional wide field of view A using an area sensor.
F may be used, or active scan multi-AF may be used. Further, the integral control of the sensor may be a fixed type for a predetermined time. Further, as the visual axis detection method, another known method may be used.

Further, when the weights are the same in step S189 of FIG. 18, the focus area may be selected (priority to the near focus or priority to the line-of-sight position) instead of the center priority.
The priority order may be center → gaze position. Then, the flash generation is determined (from ISO sensitivity and brightness information), and when the flash fires, it is assumed that the brightness of the subject will be based on the distance measurement information and the flash guide number value. Weighting a
(3), d (3), and e (3) may be performed to select the focus area in the final addition processing.

Incidentally, FIG. 26 shows the structure of the second embodiment in which a strobe is taken into consideration and will be described. In the second embodiment, a strobe circuit 18 and an ISO sensitivity reading switch are added to the configuration of FIG. The weighting value may be set as follows. It should be noted that the subject brightness indicates a value with respect to the optimum exposure.

[0085]

[Table 2]

According to the above embodiment of the present invention, the following constitution can be obtained. (1) Multi-point distance measuring means for measuring a plurality of focus detection areas,
A line-of-sight detecting means for detecting the line-of-sight direction of the photographer, and a selecting means for selecting only one piece of distance-measuring data based on the outputs of the multi-point distance-measuring means and the output of the sight-line detecting means. Characteristic distance measuring device. (2) Multi-distance measuring means for measuring a distance between a plurality of focus points, distance-distance weighting means for giving a weight value to each focus point according to the distance-measuring information detected by the multi-distance measuring means, and a photographer's A line-of-sight detection unit that detects the line-of-sight direction and a line-of-sight detection block corresponding to the line-of-sight signal detected by the line-of-sight detection unit are determined, and line-of-sight weighting that gives a weight value to each focus point according to the corresponding block. A distance measuring device, wherein only one focus point is selected on the basis of a means and a sum of the distance measuring weight value and the line-of-sight weight value. (3) The distance measuring apparatus according to (2), wherein the distance measuring weighting means gives a larger weight value as the distance from the multi distance measuring means is shorter. (4) The distance measuring device according to (2), wherein the distance measuring weighting means gives the same weight value when a plurality of distance measuring data are within a predetermined range. (5) The distance measuring apparatus according to (2), wherein the line-of-sight weighting unit weights each focus point for the line-of-sight detection block shown in the viewfinder. (6) Multipoint distance measuring means for measuring a plurality of focus points, distance measuring weighting means for giving a weight value to each focus point according to distance information detected by the multipoint distance measuring means, and photographing The line-of-sight detection means for detecting the line-of-sight direction of the person and the line-of-sight detection block corresponding to the line-of-sight signal detected by the line-of-sight detection means are determined, and a weight value is given to each focus point according to the corresponding block. Line-of-sight weighting means, detection reliability determination means for evaluating the reliability of the data obtained by the line-of-sight detection means, and predetermined output to the line-of-sight weighting means according to the degree of reliability detected by the line-of-sight detection means. Based on the reliability coefficient means for setting the weight value multiplied by the coefficient to each focus point, and the added value of the distance measurement weight value and the line-of-sight weight value, only 1 is set. Distance measuring apparatus and selects the focus point. (7) The reliability determining means determines the reliability at the time of detecting the line of sight as 2
The distance measuring device according to (6), wherein the reliability coefficient means sets a coefficient according to the reliability evaluation value. (8) Multi-point distance measuring means for measuring a plurality of focus points, distance measuring weighting means for weighting each focus point according to distance information detected by the multi-point distance measuring means, and strobe device And strobe weighting means for determining the light emission conditions of the strobe device and the amount of light reached at the time of light emission, and weighting each focus point, a line-of-sight detection means for detecting the line-of-sight direction of the photographer, and the line-of-sight detection means. The line-of-sight detection block corresponding to the line-of-sight signal detected by the line-of-sight detection unit is determined, and the line-of-sight weighting unit that weights each focus point according to the corresponding block, and the reliability of the data obtained by the line-of-sight detection unit. A detection reliability determination means for evaluating
In accordance with the reliability degree detected by the line-of-sight detection means, reliability coefficient means for weighting the output of the line-of-sight weighting means by a predetermined coefficient at each focus point, the distance measurement weight value and the line-of-sight A distance measuring device characterized in that only one focus point is selected on the basis of a value added to a weight value. (9) The distance measuring apparatus according to (8), wherein the strobe weighting means performs weighting according to the degree of underexposure based on the distance measurement information and the strobe light emission amount.

[0087]

According to the present invention, a plurality of plural objects detected in correspondence with the focus detection area based on the distance measurement information, the line-of-sight information, and the information including the reliability of the line-of-sight signal by a simple algorithm. It is possible to provide a distance measuring device and a distance measuring method for selecting optimum distance measuring data from distance data.

[Brief description of drawings]

FIG. 1 is a conceptual diagram of a camera using a distance measuring device of the present invention.

FIG. 2 is a diagram showing a configuration of a camera using the distance measuring apparatus according to the first embodiment of the present invention.

FIG. 3 is a diagram showing a detailed configuration of a finder / line-of-sight optical system in FIG.

FIG. 4 is a light source LED0, L for line-of-sight detection of the first embodiment.
It is a figure which shows the state of the detection signal by the reflected light from the eye by ED1.

FIG. 5 is a diagram showing a masking state of a finder.

FIG. 6 is a flowchart showing a camera sequence of the first embodiment.

FIG. 7 is a flowchart showing a sequence of a subroutine “line-of-sight detection” executed in step S6 of FIG.

8 is a flowchart showing a sequence of a subroutine "signal detection (sensor control)" executed in steps S24 and S29 of FIG.

9 is a flowchart showing a sequence of the modified example of FIG.

10 is a flowchart showing a sequence of a subroutine "display" executed in step S36 of FIG.

FIG. 11 is a diagram showing various display states of a finder by LEDs.

FIG. 12 is a flowchart showing a sequence of a subroutine “signal processing” executed in step S33 of FIG.

FIG. 13 is a flowchart showing a sequence of a subroutine “ghost processing” executed in step S72 of FIG.

FIG. 14 is a diagram for explaining a method of setting variables h and r.

15 is a flowchart showing the sequence of a subroutine "feature extraction 1" executed in steps S75 and S79 of FIG.

16 is a flowchart showing a sequence of a subroutine "feature extraction 2" executed in step S85 of FIG.

17 is a flowchart showing a sequence of a subroutine "signal conversion" executed in step S89 of FIG.

FIG. 18 is a flowchart showing a sequence of a subroutine “focus area selection” executed in step S8 of FIG.

FIG. 19 is a diagram showing a state of a line-of-sight detection area in which line-of-sight weighting is performed.

20 is a flowchart showing a sequence of a subroutine "line-of-sight weighting" executed in step S183 of FIG.

FIG. 21 is a flowchart showing a sequence of a subroutine “reliability weighting” executed in step S184 of FIG.

22 is a flowchart showing a sequence of a subroutine "distance weighting" executed in step S187 of FIG.

FIG. 23 is a diagram for explaining how a distance is converted into a block value.

FIG. 24 is a diagram showing an appearance of a setting table.

FIG. 25 is a diagram showing how a liquid crystal arranged near the primary image forming surface of the finder shields light other than the photographing area according to the distance measurement information of the selected focus area.

FIG. 26 is a diagram showing a configuration of a second exemplary embodiment with a strobe taken into consideration.

FIG. 27 is a diagram for explaining mask processing.

FIG. 28 (a) shows the arrangement of focus areas,
(B) is a figure which shows the example of weighting with respect to the focus areas a to c of (a).

[Explanation of symbols]

1 ... Focus area determination unit, 2 ... Multi distance measuring unit, 3 ...
Line-of-sight detection device, 4 ... Reliability determination unit, 5 ... Photometric unit, 6 ... Strobe unit, 7 ... Strobe control unit, 8 ... ISO detection unit, 1
1 ... CPU, 12 ... Multi AF circuit, 13 ... LED circuit, 14 ... Line sensor circuit, 15 ... Viewfinder / line-of-sight optical system, 16 ... Lens drive circuit, 17 ... Photometric circuit.

─────────────────────────────────────────────────── ─── Continuation of the front page (51) Int.Cl. ⁶ Identification code Office reference number FI technical display area G03B 7/28

Claims (3)

[Claims] 1. A multipoint distance measuring means for performing distance measurement in a plurality of predefined focus detection areas, and a first for the plurality of focus detection areas based on output data of the multipoint distance measuring means. First weighting means for performing weighting, line-of-sight evaluation means for detecting the line-of-sight direction of the observer, and evaluating reliability when detecting the line-of-sight direction, and the plurality of focus detection areas according to the output of the line-of-sight evaluation means A second weighting means for performing a second weighting with respect to the first weighting means, and an added value of the output of the first weighting means and the output of the second weighting means. Only 1 out of multiple detected distance data
A distance measuring device comprising: data selecting means for selecting one distance data. 2. The second weighting means detects the line-of-sight direction of an observer based on a plurality of predefined line-of-sight detection regions, and the reliability coefficient at the time of detection for the plurality of focus detection regions. The distance measuring device according to claim 1, wherein weighting is performed by multiplying by. 3. A distance weighting value corresponding to each distance is assigned to each focus detection area in accordance with a plurality of distance data from the multi-point distance measuring means, and the line of sight of an observer is determined based on a predetermined line of sight detection area. The direction is determined, a line-of-sight weight value based on the product of the reliability coefficient at the time of detecting the line-of-sight is given to each focus detection region, and the added value of the line-of-sight weight value and the distance weight value is evaluated,
A distance measuring method characterized in that only one distance data is selected from the plurality of distance data based on the one giving the maximum value among them.