JP2001174691A - Range-finder for camera - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide the range-finder of a camera by which a range-finding point is efficiently obtained at high speed while inexpensively constituting it by using a simple arithmetic control circuit such as an ordinary one-chip microcomputer. SOLUTION: This range-finder possesses a CPU1 provided with functions as a first splitting means to split an image plane into the M-number of points, a luminance measuring means to measure a luminance distribution at each of the M-number of points, a selecting means to select the N-number of points out of the M-number of points based on the luminance measured result of the luminance measuring means and a second splitting means to respectively split the N-number of points selected by the selecting means more at the time of deciding the main object position.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a camera distance measuring device.

[0002]

2. Description of the Related Art In an AF camera, a function of measuring a distance between a plurality of points (points) on a screen and performing focusing has conventionally been adopted as a multi-AF function. In a camera having a multi-AF function, the time required for distance measurement increases as the number of distance measurement points increases.
Japanese Patent Application Publication No. 142490 discloses that a scene is divided.

[0003]

SUMMARY OF THE INVENTION The above-mentioned Japanese Patent Application Laid-Open No.
According to the method described in Japanese Patent Application Laid-Open No. 142490, it takes a considerable amount of time to obtain the distance distribution. For example, when performing distance measurement for one point,
In the case of a so-called passive type AF for detecting a subject image signal from a pixel signal, complicated calculations such as appropriate integration control of pixel data, correlation calculation, and interpolation calculation are required.
Moreover, these operations cannot be performed at the same time, and must be performed in order. Therefore, in order to measure the distance of many points, the above processing has to be repeated many times.

The present invention has been made in view of such a problem, and an object of the present invention is to provide a high-speed, low-cost configuration using a simple arithmetic control circuit such as a normal one-chip microcomputer. An object of the present invention is to provide a camera distance measuring device capable of efficiently finding a distance measuring point.

[0005]

In order to achieve the above object, a distance measuring apparatus for a camera according to a first aspect of the present invention comprises: a first dividing means for dividing a screen into M points; Luminance measuring means for measuring the luminance distribution of each point, selecting means for selecting N points out of the M points based on the luminance measurement result of the luminance measuring means, and determining the main subject position In doing so, there is provided a second dividing unit for further dividing each of the N points selected by the selecting unit.

According to a second aspect of the present invention, in the camera distance measuring apparatus according to the first aspect, the luminance measuring means measures a luminance distribution when auxiliary light is projected from the camera.

According to a third aspect of the present invention, in the camera distance measuring apparatus according to the first aspect, each of the N points has a plurality of distance measuring points. The position of the main subject is determined based on the position.

A distance measuring device for a camera according to a fourth aspect of the present invention is a camera for determining the position of a main subject in a screen, wherein the sensor array determines the luminance distribution in the screen.
The screen is divided into M points on the basis of the output of the sensor array, a distance measurement operation is performed on a limited N points among the M points, and a plurality of distance information obtained by the distance measurement operation is obtained. Deciding means for deciding the position of the main subject on the basis of

A distance measuring apparatus for a camera according to a fifth aspect of the present invention, in the fourth aspect, further comprises a switching means for switching the number of points for which a distance measuring operation is performed among the divided M points. ing.

[0010]

Embodiments of the present invention will be described below in detail with reference to the drawings. To clarify the distance measuring method according to the present embodiment, a distance measuring method according to the related art will be described first.

FIG. 28 shows the basic configuration of the distance measuring device. In FIG. 28, 1 is a CPU, 2a and 2b are sensor arrays, 3a and 3b are light receiving lenses, 5 is a subject, 34 is an A / D converter, 35 is an arithmetic unit, and 33a and 33b are image patterns.

A conventional distance measuring method will be described below with reference to FIGS. 29 (a), (b) and (c).

A in the screen as shown in FIG.
When measuring the distance between the three points B and C, first, for the point A, the subject distance is obtained by performing integral control, correlation operation and interpolation operation (step S1 in FIG. 29A).
00, S101, S102). Next, the same process is performed for point B (steps S103 and S1 in FIG. 29A).
04, S105). Next, the subject distance is calculated for point C in the same procedure (step S10 in FIG. 29A).
6, S107, S108). Finally, a main subject distance is selected from points A, B, and C (step S109).

Here, the integral control means the sensor array 2a,
In detecting the image signal of the subject 5 using 2b,
This is a technique of storing a voltage signal within a dynamic range of a processing circuit when an output current of a pixel forming the sensor arrays 2a and 2b is stored in a capacitor and converted into a voltage signal.

Next, the correlation operation will be described. FIG. 30 is a diagram for explaining the calculation principle of distance measurement.
FIG. 30A shows the relationship between the focus and the shift amount, and FIG.
(B) is a diagram showing a signal output for each sensor portion of the right sensor R, FIG. 30 (c) is a diagram showing a signal output for each sensor portion of the left sensor L, and FIG. 30 (d) is adjacent to the left and right sensor outputs. FIG. 7 is a diagram showing a relationship between outputs of the sensor portion.

FIG. 30 shows the correlation operation for each pixel.
(B) When the image signal as shown in FIG. 30 (c) is obtained, the relative displacement amount X of the image position on the sensor array 2a side and the image position on the 2b side is obtained in pixel pitch units. This takes a difference for each pixel data of the two sensor arrays 2a and 2b,
The result of adding these is obtained, and then the sensor array 2b
Is shifted by only one pixel, and the difference between the pixel data is added again in the same procedure. The above processing is repeated, and when the two image signals match, it is determined that the degree of correlation is high,
An approximate value of the deviation amount X is obtained. When shifted by several pixels, the two data coincide, the added value becomes the MIN value, and the degree of correlation is the best. However, if the added value does not become 0, it is considered that there is a shift amount equal to or less than the pixel pitch,
An interpolation operation is further performed.

The above-described complicated arithmetic control is shown in FIG.
In repeating the flow according to the flow of (a), for example, when measuring a distance of as many as 30 points as shown in FIG. 29 (c), it takes about ten times as long as the case of FIG. 29 (b). In other words, the photographer may miss a photo opportunity.

Next, a method for calculating the relative position difference between the light patterns in the passive type distance measuring apparatus will be described with reference to FIG.
This will be described with reference to FIG.

In FIG. 29, the difference between the positions of the light receiving lenses 3a, 3b (base line length B) indicates that the sensor arrays 2a, 2b
The relative position difference x of the distribution of light incident above changes depending on the distance L of the subject 5. Assuming that the focal length of each of the light receiving lenses 3a and 3b is f, the subject distance L is obtained by the following equation.

L = B × f / x (Equation a) Each sensor of the sensor arrays 2a and 2b outputs a current signal according to the amount of incident light. These are converted to an A / D converter 34.
, And is input to the arithmetic unit 34. The calculation unit 34 calculates the relative position difference x by the correlation calculation.
Is detected to calculate the image shift amount. The result is input to a calculation control means (CPU) 1 composed of a one-chip microcomputer or the like, and is calculated based on the above (formula a).
Is required. The above is the basic principle of the passive type triangulation and the general device configuration.

The above described shift amount calculating function generally comprises two procedures as described later, but these may be incorporated in the CUP 1 as a control program. When focusing a camera using such a technique, the CP
If U1 controls the operation of the camera and appropriately controls a lens for photographing or focusing via an actuator such as a motor, a camera with an automatic focus (AF) function can be provided.

By the way, in order to calculate the image shift amount,
An operation step (that is, a correlation operation) for examining how much the image is shifted by the unit of the sensor pitch in the two line sensors 2a and 2b is required. Then, an operation step (hereinafter, referred to as an interpolation operation) for calculating the shift amount more precisely with a finer resolution is required.

For example, the waveform 33a is displayed on the sensor array 2a.
When the light is incident in the pattern shown in FIG. 30, the magnitudes of the outputs R1 to R6 of the sensors have a distribution 33a as shown by a bar graph in FIG. Here, 'R' indicates the right sensor, 'L' indicates the left sensor, and the suffixes 1 to 6 attached thereto indicate, for example, the absolute position of the sensor with respect to the optical axis of the light receiving lens. If the same signals as the outputs R1 to R6 are output from the outputs L1 to L6 of the left sensor, the relative position difference x is 0, and the subject distance L to be obtained is “infinity”.

If the subject is at a "finite distance", the left sensor L shifted by the number S of sensors determined from the above x and the sensor pitch SP has a signal shown in FIG.
As a result, output signals having values similar to the outputs R1 to R6 are obtained.

The value F on the vertical axis in the graph of FIG.
F (i) is obtained according to the following equation.

FF (i) = Σ | R (i) −L (i) | (Expression b) That is, the output of the corresponding L sensor is subtracted from the output of a certain R sensor, and the absolute value thereof is calculated as The result added for each sensor may be used as the FF. That is, first, R
The absolute value is obtained by subtracting L (i) from (i), i is changed by a certain width, and these are added.

Next, one sensor of Ri or Li is set to 1
If the difference is obtained in the same manner as the adjacent sensor which has been shifted earlier by the unit, FF (i + 1) can be expressed by the following equation.

FF (i + 1) = Σ | R (i + 1) −L (i) | (Expression c) As described above, the values of the FFs are sequentially shifted (hereinafter, referred to as SIF).
T amount), the SIFT amount at which FF, which is the sum of the difference between R and L, has the minimum value (Fmin) is considered to be the position where the best correspondence is obtained. Is obtained as the above S. The outline of the processing procedure relating to the correlation operation has been described above.

Further, when the output distribution of both sensor arrays is illustrated taking the above S into consideration, as shown in FIG. 30 (b), each of the sensors shifted by S on the L side has a corresponding subscript on each of the R side. An output similar to that of the sensor is obtained.

FIG. 30 shows details of the interpolation operation.
This will be described with reference to (b) to (d). The actual image shift amount on the two sensor arrays does not completely correspond to the sensor pitch, and more precisely, the image shift amount must be detected with an accuracy smaller than the pitch. Therefore, it is necessary to perform an interpolation operation. R and L in FIGS. 30 (b) and 30 (c) respectively represent a part of sensor outputs constituting the sensor arrays 2a and 2b in FIG. FIG.
(D) shows a graph in which the “correlation calculation” has already been completed, and the data has been shifted by the above S to make it easy to compare. That is, L0 to L4 are exactly Ls to Ls +
4 should be described, but this S is omitted to avoid complication in description.

Here, it is assumed that light shifted by x on the basis of R is still incident on the L sensor after shifting by S. At this time, for example, the light incident on R0 and R1 is mixed and incident on the L1 sensor, and similarly, the light shifted by x on the basis of the R is sequentially incident on each L sensor. It can be seen that (L1 to L3) are expressed as shown in [Equation 1] of FIG.

Fmin and a value FF of the FF obtained by shifting the shift amount from Fmin in the plus direction and the minus direction.
When 1 and F + 1 are expressed using the outputs of the respective Rn and Ln, they are expressed as [Equation 2] in FIG. further,
When [Expression 2] is expanded using [Expression 1], the value Fmin,
Each of F-1 and F + 1 is represented as [Equation 3] in FIG.

In addition, 式 | R0−R1 |
When + | R1−R2 | + | R2−R3 | 表現 is expressed as (ΣΔR), the displacement x is obtained by the calculation shown in [Equation 4] of FIG. 27 without depending on (ΣΔR).
This is the "interpolation operation".

These calculations are performed by the calculation unit 35 in FIG. 28, but may be performed by a calculation control means (CPU) 1 such as a one-chip microcomputer according to a predetermined program.

Based on the values S and x thus obtained,
CPU 1 calculates the amount of extension of the focusing lens,
If controlled, an autofocus (AF) camera can be provided.

However, in measuring a large number of ranging points, it is very important to determine which ranging point to perform the above-described calculation for the ranging.

An embodiment of the present invention which overcomes the above-mentioned conventional problems will be described below. FIG. 1A shows a configuration of a distance measuring apparatus to which the present embodiment is applied. FIG.
In (a), 2a and 2b are light receiving elements, 3a and 3b are light receiving lenses, 4 is a focusing unit, 5 is a subject, 7 is an integrating unit, and 8 is a selecting unit. The control unit (CPU) 1 has the functions of an A / D converter 1a, an integration control unit 1b, and a correlation operation unit 1c.

For example, in a composition such as that shown in FIG. 1B, it is necessary to perform correct distance measurement and focusing on a point indicated by a star in the screen. Correlation calculation and interpolation calculation must be performed by properly capturing the image. However, if the user changes the composition and holds the camera in the composition as shown in FIG. 1C, the position of the ☆ mark in the screen is at the upper part of the screen, here, the periphery of the sun. Therefore, the camera reacts to the brightness of the sun, increasing the probability that the image of the person to be photographed cannot be obtained correctly.

As described above, if the main subject position is not correctly determined, it is difficult to take a photograph with the correct focus, no matter how many points in the screen can be measured.

FIG. 2 is a diagram mainly showing a detailed configuration of the integration section of FIG. The sensors S1 to S4 constituting the sensor arrays 2a and 2b are driven by a bias circuit 20 as a power supply, and each output a signal current corresponding to the amount of received light. This signal current is guided to the integration amplifiers A1 to A4 when the integration start / end switch 7a is ON.
When the reset switch 7b is in the OFF state,
A voltage signal corresponding to the integration amount is output to the output of each of the integration amplifiers A1 to A4. This result is stored in the A /
The data is read by the D converter 1a, and the correlation operation is performed by the correlation operation unit 1c.

However, since the amount of light entering each of the sensors S1 to S4 varies depending on the brightness of the scene, the color of the object, and the reflectance, the amount of light integrated by the integration amplifiers A1 to A4 having a limited dynamic range. In order to keep the value at an appropriate value, an accurate integration control technique is required. For example, when the integration time is too short, the integration result becomes flat and no difference can be obtained. However, when the integration time is too long, the integration result becomes uniform due to saturation of the circuit.

As is clear from the above description of the correlation calculation, if the image changes little, it is difficult to correlate the two images obtained by the two sensor arrays, and as a result, correct distance measurement cannot be performed. .

Therefore, a technique of monitoring the integration result in real time and terminating the integration when the level reaches an appropriate level is used. That is, the output of the sensor to be monitored is determined depending on which one of the plurality of switches 7c is turned on. FIG. 3 is a timing chart when the switch 7c is turned on to perform integral control. When light is incident on each of the sensors S1 to S4, first, the reset switch 7b is turned on to reset the outputs of the integration amplifiers A1 to A4 to a reference level, and then the integration start / end switch 7a is turned on, and the reset switch 7 is turned on.
When b is turned off, integration is started at the timing of T1.

The maximum integration value detection circuit 6 includes a plurality of integration amplifiers A1 to A4 which are input by sequentially switching a plurality of switches 7c.
Is a circuit that detects the largest one of the integrated values of Therefore, when the A / D selection switch 8 is connected to the maximum integration value detection circuit 6, the output having the largest integration amount is selected and input to the A / D converter 1a of the CPU 1. Therefore, the CPU 1 calculates this maximum integrated output as shown in FIG.
As shown in (a), the A / D converter 1a is sequentially operated and monitored, and at time T2 when this maximum integrated value does not exceed the dynamic range of the circuit, the integration start / end switch 7 is turned on.
If a is turned off, the integrated output of each sensor does not exceed the dynamic range. After integration stops, each sensor S
If the A / D selection switch 8 is switched and controlled to perform A / D conversion by taking in the integrated outputs of 1 to S4, CP
U1 can monitor each sensor output sequentially.

The image signal thus obtained is shown in FIG.
As shown in (b), dark places show low output and bright places show high output according to the incident state of light. With such a technique, the distance measuring device of the camera can obtain an appropriate image signal.

Further, the CPU 1 controls a plurality of switches 7c.
Is selectively operated so that only the output of a specific integration amplifier is connected to the maximum integration value detection circuit 6, so that among the sensors S1 to S4, as shown in the example of FIG. Even if the sun's light enters S1, this sensor S1
Is not connected to the maximum integration value detection circuit 6, the integration control is not performed by the strong light of the sun.
Conversely, when direct sunlight from the sun is incident during integration monitoring, integration control ends before an image signal of the subject is obtained, and accurate distance measurement often cannot be performed. The same can occur in a circuit that takes the difference between the maximum value and the minimum value or a circuit that detects the average value, instead of the maximum integration value detection circuit 6.

Further, in the configuration shown in FIG.
Using the light beam indicated by the dotted line from a, it is possible to measure the distance of points other than the point C at the center of the screen with respect to the subject 5, that is, the points L and R shifted in the base line length direction.

Also, as shown in FIG.
By adding one sensor array 2a and 2b arranged behind the sensors 3a and 3b in the direction perpendicular to the base line length direction, one part in the direction U and the part D in the direction perpendicular to the base line length direction are measured. The distance becomes possible. In other words, it is possible to measure the distance of many points in the screen in the schematic sensor array monitor area (distance-measurable area) as shown in FIG.

If this idea is further extended, as shown in FIG. 5, instead of one or three line sensors,
So-called area sensor 2 in which sensors are arranged continuously
By using a and 2b, the entire screen can be monitored. For example, as shown in FIG. 29C, the number of measurable points can be increased to 30 or more.

If the number of distance measurement points is increased by such a measure, accurate distance measurement can be performed irrespective of the position of the main subject on the screen. Even in the case of a composition as shown in FIG. 3A, even when a person is present at the edge of the screen, accurate focusing can be performed. However, as described above, the wider the ranging area is, the more often unnecessary measurement is performed at the time of ranging, and the probability of erroneous ranging increases accompanying this.

Therefore, for example, in the scenes shown in FIGS. 1B and 1C, if the position of the subject 5 is measured from the beginning and focused thereon, other points will be erroneously measured. It does not cause out of focus or time lag. Therefore, in the present embodiment, FIG.
As described later, first, the luminance distribution of each point in the screen (in this case, divided into M points by the function of the dividing unit of the CPU 1) obtained from the integration control result of the entire screen is measured. Next, based on the luminance distribution at this time, N points out of the M points are selected as high priority areas.
Next, a distance distribution is obtained by further dividing each of the N points and measuring the distance, and the position of the main subject is determined based on the obtained distance distribution. Then, interpolation calculation is performed only for the main subject position. According to such a distance measuring method, the distance measuring operation is completed before a point having a low priority is measured, so that high-speed distance measuring can be performed. Here, there are many cases where the brightness of the main subject is higher than that of the main subject above the center of the screen, such as the sky and the sun. Therefore, in the present embodiment, it is helpful to determine the main subject based on such information.

FIGS. 7 and 8 are flowcharts for explaining the details of the procedure for determining the priority. Here, considering that the photographer makes the composition horizontal or vertical, first, in step S1, the position of the upper portion of the screen is detected.

FIG. 9 is a diagram for explaining vertical / horizontal detection or upward detection of a camera. As shown in FIG. 9, a case 51 containing a fluid conductive material 52 such as mercury is built in the camera, and any one of a plurality of electrodes (ini) 53 inserted into the case 51 is selected. The vertical and horizontal detection of the camera can be performed by the CPU determining whether the short circuit occurs due to the fluid conductor. That is,
If (b) and (c) are short-circuited as shown in (a), it can be determined as horizontal, and if (b) and (b), it can be determined as vertical, and
As shown in (c), if b and d, it can be determined to be upward.

After checking the position of the upper part of the screen by the above-described method, the process proceeds to step S2. In step S2, based on the determination result at this time, when the screen is divided into nine (1 to 9) as shown in FIG. 10A, whether the block is on the upper side, whether the block is on the upper side,
The determination method is switched depending on whether or not the block is above. In each case, the idea is the same: the top of the screen is empty, and the main subject is at the bottom,
Furthermore, since the sequence is configured in the center of the screen based on the idea that the probability of the existence of the main subject will be high, the description of the flow will be made here, assuming an example of a horizontally long composition in which the camera is normally held (the upper part). Perform In such a composition, since the area has a high probability of being empty, the average luminance of the area in step S3 is obtained, and this is used as a criterion for determining the empty area. That is, in the composition as shown in FIG. 10B, this brightness information is used when it is determined that the additional portion is also empty.

However, if the top block is empty,
In this case, it is not possible to judge whether or not the distance measurement is necessary. Therefore, in step S4, the predetermined luminance BVS is compared with the average luminance BVU of the uppermost block. Since the priority coefficient P is 1
× 1 (step S5). The first numeral 1 of 1 × 1 is the value of the weight given to each of the parts of FIG. 10A. The four corners of the screen and the middle of the upper part are set to 1 because the frequency of the main subject is low. Areas other than those parts are assigned a weight of 2, and the center is given a weight of 3, since the existence probability is extremely high.

When the number 1 after 1 × 1 is empty,
The weighting is increased by 1 depending on the necessity of the distance measurement and 1 or 2 or 6 in other cases, depending on the luminance and its change. That is, when the area is empty and has a low existence probability, the priority coefficient is 1 × 1 = 1. Therefore, when there is a mountain in a part of the sky as in the scene of FIG. 10C, step S4 is branched to N and the process proceeds to step S6, where the priority coefficient is set to 1 × 2. 1x2 first half number 1
Does not change because it is position dependent.

After the priority coefficient of the block is determined, the next block (here) is determined. In step S7, a change in the luminance is checked. There is a change in the scene as shown in FIG. 10B, and there is no change in the scene as shown in FIG. If there is a change, step S8 is branched to Y and the process proceeds to step S9, where the luminance of each block is compared with the average luminance BVU at the upper portion of the screen previously obtained. Therefore, the weight of 2 is given, and if different, the weight of 6 is given. The above is the processing from step S9 to step S17.

In the scene as shown in FIG. 10B, step S9 is branched to N, and the flow advances to step S10 to set the priority to 2
× 6 = 12. Further, since the luminance distribution is the same in the scene of FIG. 10C, step S8 branches to N, and the process proceeds to step S20. Here, since there is no large change in luminance, the same weight 2
Is fine. Considering the weighting of the position, weighting of 3 × 2 = 6 is performed in step S20, and weighting of 2 × 2 = 4 is performed in step S21.

Then, weighting of 6 is performed on the assumption that there is a high probability that an object is present in the remaining block at a lower position than the remaining blocks, based on the luminance determination so far. Furthermore, since the weight of the position is multiplied, 1 ×
For 6 = 6, the weighting is 2 × 6 = 12 (step S22, step S23). Therefore,
Are 12, and 6 are 6, and these are distance measuring points with high priority.

After step S17, the following processing is performed.
In the scene as shown in FIG. 10 (b), the weights are 12, 6, and 4, respectively. For the lowermost stage, the remaining predictions of the luminance determination so far are performed with respect to the position weights in steps S18 and S19. Is multiplied by 3. This gives 1 × 3 =
3 is weighted by 2 × 3 = 6. Therefore, the order of priority is higher.

In step S24, based on the weighting results of the priority coefficients P1 to P9, in the scene of FIG.
In the scene (c), priorities are assigned in the order of...

As described above, by analyzing the distribution of the position and the luminance in the screen in consideration of the probability that the subject exists, the points to be distance-measured (9 to 9 in FIG.
The division can be made finer).

Incidentally, also in the case where step S2 is not branched to step S3 (when branched to step S25 or S26), the concept is the same, so that the description here is omitted. Even if the number of distance measurement points is increased based on such a concept, it is possible to narrow down the points at which actual distance measurement is performed (four points in the embodiment of FIG. 6).

On the other hand, in a scene as shown in FIG.
Step S2 in FIG. 7 branches to step S3, and step S8 in FIG. 8 branches to N, and the process proceeds to step S20. In this case, a hatched area shown by (a) in FIG. 11 is a distance measurement candidate area. Therefore, only the selected area is measured with emphasis. For each area, the distance measurement is performed by dividing the area into smaller parts. here,
As shown in FIG. 11A, three distances are measured for each area.

If the focusing as in the present embodiment is not performed, a distance measurement of 9 areas × 3 points = 27 points is required. In this embodiment, the focusing is limited to 4 × 3 = 12 points. Thus, it can be seen that the speeding up has been achieved. Here, when the distance measurement only for the correlation operation is performed for the above 12 points, a distance distribution as shown in the three-dimensional graph of FIG. 11B is obtained. Such a distance distribution clarifies the face part and the body part of the subject as shown in FIG. 11C, so that the CPU may focus on the center of the face as the main subject. Although such a more accurate determination of the main subject cannot be determined only from the brightness distribution alone, in the present embodiment, it is realized by a distance measurement procedure as shown in FIG.

Step S201 in FIG. 6 is a step for performing integral control so that the entire screen has an appropriate contrast. In the next step S202, M number of steps are performed by using the procedure shown in FIGS. The priority is determined based on the luminance distribution of each area. Step S203, S20
In step 4, S205 and step S210, four distance measurement candidate areas are selected, and distance measurement (correlation calculation) is performed for each area at every three points. As a result, a distance distribution as shown in FIG. 11B is obtained (step S206). In the next step S207, the main subject position is determined based on the relationship between the face and the body part (step S207). Here, the center of the area in FIG. In the next step S208, interpolation calculation is performed on the main subject position data, and in step S209, focusing is performed.

As described above, in the present embodiment, in addition to the brightness information, finer distance distribution information is used, so that the position of the main subject can be determined quickly and accurately for focusing. . If the CPU has a high speed and has a sufficient time lag, the interpolation calculation may be performed in step S204. This is because the distance measurement points have already been narrowed down in step S202.

Instead of using the vertical / horizontal composition determination method shown in the flowcharts of FIGS.
2. An optimum ranging area may be selected from the output of the area sensor as shown in FIG.

For example, in a horizontal composition as shown in FIG.
Since the upper half is empty, there is a high probability that the main subject exists somewhere in the lower half. In order to determine this automatically, as shown in FIG. 12 (b), the area sensor is considered based on the x and y coordinates, and the output values of pixels having the same y value along the x direction are added and displayed as a graph. Luminance distribution in the case and FIG.
As shown in (c), the output luminance values of the pixels having the same x value along the y direction are added, and the added luminance distribution when the graph is displayed may be compared. In FIG. 12B, since the sky and the earth are separated, a large luminance change represented by ΔBV is observed.
In FIG. 12 (c), the sky and the ground melt and a monotonous change in luminance occurs. The portion where this large change in luminance is detected as yB.

On the other hand, in a vertical composition as shown in FIG.
Conversely, a large change in ΔBV is seen in the added luminance distribution added in the y direction, and only a monotonous change is seen in the added luminance distribution obtained by adding the output values of the sensor areas having the same y value in the x direction ( (FIG. 13 (b), (c)). Here, as in FIG. 12, a portion where a large change in luminance occurs is defined as xB.

Using such characteristics, the following FIG.
If a distance measuring point to be prioritized is selected in the flow as shown in FIG. 15, a priority distance measuring area is set in a portion excluding the empty space as shown in FIGS. 16 (a) and 16 (b). Can be. When xB and yB cannot be detected,
The central area of the screen as shown in FIG. 16C, which has a high probability that the main subject exists, may be set as the priority area.

The flow shown in FIGS. 14 and 15 corresponds to the flow shown in FIG.
This is a flow for determining the distance measurement priority area (focusing priority area) from the nine divisional areas as shown in FIG. 7, and can be applied to an application with a larger number of divisions as in the flow of FIGS. 7 and 8 described above. There is, however, a simplification here to simplify the explanation.

Step S30 in FIG. 14 is an integration control step for obtaining an image signal of the entire screen. Step S3
1. Steps S32 and S33 are steps for obtaining a transverse luminance distribution such as a horizontal composition of a screen as shown in FIG. 16A. In the example of FIG. 12, data added in the x direction is obtained. Since there are three divisions in the vertical direction, the addition results of the upper, middle, and lower stages are BV1, BV2, and BV3. These comparisons are performed in steps S34 and S35. Since a place where the luminance greatly changes beyond a predetermined luminance difference is considered to have crossed the boundary with the sky, the distance measurement priority of a portion above it is lowered to lower the distance measurement priority. Raise the priority of the person. However, FIGS. 7 and 8
The weight of the screen position is considered in the same way as described above. When step S34 is branched to y, only the top stage is empty, so the middle stage is weighted by 6 according to the luminance distribution, and the lower stage is weighted by 3, which is multiplied by the weight of the position of each area. Can be When step S35 is branched to y, not only the upper stage but also the middle stage is empty, so the uppermost stage has the lightest weight (step S35).
50). The middle stage is weighted by 2 (step S5).
2, S53), the lower stage is weighted by 6 (step S5).
3, S54).

If there is no large luminance distribution exceeding .DELTA.BV0 in steps S34 and S35, the luminance difference between the upper and middle stages and the luminance difference between the upper and lower stages are determined in step S36. The larger one is stored as ΔBVY. This is for re-determining the vertical and horizontal composition later.

Steps S37 to S3 in FIG.
Reference numeral 9 denotes a portion in which luminance data addition in the y direction is performed as shown in FIG. 13 in order to determine whether or not the composition is a vertical composition as shown in FIG. 16B. In steps S40 and S41, as in steps S34 and S35, the luminance difference between the added values is compared with a predetermined luminance difference ΔBV0. If the difference exceeds the predetermined value, it is determined that an empty portion has been detected.

The processing when steps S40 and S41 are branched to y is the same as steps S34 and S35. Further, similarly to step S36, in step S42, the luminance difference between the upper row and the middle row at the time of vertical composition,
The larger difference between the luminances of the upper and lower stages is stored as ΔBVT. In step S43, ΔBVY obtained in step S36 and BVV obtained in step S42
T is determined, and when the difference is equal to or more than the predetermined luminance ΔBV1, it is determined that the possibility of the horizontal composition is high again, and step S6
Branch to zero. This is a countermeasure for preventing a brightness difference of ΔBV0 or more from appearing in steps S34 and S35 when a dark landscape is set as a background. However, when the luminance difference is not equal to or more than the predetermined level, steps S44 to S44 are executed.
At 47, the priority is given to the areas as shown in FIG.

Returning to FIG. 14, step S55 is a step of selecting a point when distance measurement is actually performed by correlation calculation or interpolation calculation, and selects a point indicating a value of 5 or more from P1 to P5. In step S56, the distance is measured from the one with the highest priority. At this time, the distance measurement sequence may be stopped midway without performing the interpolation calculation to reduce the time lag for those which are not considered to be the main subject distance at the stage of the correlation calculation. In step S57,
The closest distance is selected from the obtained distances and used as the focusing distance. Here, instead of the closest distance, a distance close to a predetermined distance may be selected, or an average value may be used as the focusing distance.

As described above, in the present embodiment, since the sky area is omitted by the distribution of the luminance change in the screen, the vertical / horizontal detecting means as shown in FIG. 9 becomes unnecessary. As described above, in this embodiment, the focus detection priority area is first narrowed down, and in that area, a finer point is measured to increase the information on the main subject.
High-precision focusing is possible.

In this embodiment, the screen is simply displayed as 9
Although the area is divided into two areas, the number of areas may be switched. Hereinafter, this will be described.

In FIGS. 17A and 17C, the size of the main subject occupying the screen is different, so that the area of the area and the number of points to be focused on naturally change. Therefore, here, as shown in steps S301 to S319 in FIG. 18, the area to be measured and the number of ranging points are switched according to the luminance information in the screen. Again, Figure 1
2. Using the data obtained by adding the luminance information in the direction of the predetermined axis as described with reference to FIG. 13, the division points and the number of divisions of the screen are determined from the positions xH and yH of the inflection points. Also, since the area and shape of the area change depending on the division method, the number of ranging points is increased or decreased accordingly.

In the flow of FIG. 18, first, luminance data is added in the x direction, and the change in the y direction is monitored to determine the number of divisions in the y direction and the number of distance measurement points. Next, the luminance data is added in the y direction, the change in the x direction is monitored, and the number of divisions in the x direction and the number of distance measurement points are determined. According to such a procedure, the scene shown in FIG. 17A is divided into areas and the distance measuring points are arranged as shown in FIG.
In such a scene, distance measurement is performed by area division and distance measurement point arrangement as shown in FIG.

An area having a high priority is selected from the obtained areas by the method shown in FIGS. 14 and 15 to obtain distance measurement data, and the main subject position and distance are obtained from the result. The distance can be eliminated, and high-speed, high-precision focusing can be achieved. In the scene of FIG. 17B, the same effect can be obtained by measuring only 9 points in the scene of FIG. 17D and 10 points in the scene of FIG.

Next, an auto focus technique called super combination AF will be described with reference to FIG.

FIG. 19A is a block diagram showing a main configuration for measuring the distance of an object 520 by this type of AF.

Subject 52 incident from two light receiving lenses
Light from 0 is transmitted to two area sensors 502a and 502b.
Incident on. These area sensors 502a and 502b
The object image is received and subjected to photoelectric conversion. The output is A
The digital value of each pixel is input to a camera control microcomputer (CPU) 501 after A / D conversion by the / D conversion unit 502c.

The area sensors 502a, 502
A stationary light removal circuit 502d is connected to b. When the stationary light removal circuit 502d is operating, a DC light signal that constantly enters from the subject 520 is removed, and the pulse light from the light source 505a is removed. Only as an output signal. Therefore, the steady light removing circuit 5
02d is activated, and the CPU 501
When the electronic flash 505a is driven by controlling the electronic flash 05, the reflected signal light from the subject 520 forms an image as shown in FIG. 19B on the area sensor 502a. The black part indicates the part where light is incident.

In the CPU 501, software 501a for discriminating the image pattern on the area sensor is incorporated, and if it is determined that the image is a human figure, it can be regarded as a main subject.

FIG. 20 is a flowchart showing details of the distance measurement procedure of the super combination AF. First, in steps S101 to S102, prior to the distance measurement, the light source 50 is used.
A light beam is applied to the subject 520 according to 5a, and only the pattern of the reflected signal light is extracted as shown in FIG. 19B, and it is determined whether or not the pattern is a main subject based on the shape of the person or the like. If it is determined that the subject is the main subject, the extracted pattern is set as a priority area in step S10.
3 branches to Y, and the process proceeds to step S104. Step S
At 104, it is determined whether the optical signal forming the pattern is weak or whether there is a sufficient contrast, and a so-called active method (a signal light is projected from the camera side and the reflected signal light is reflected as a distance measuring method). ) Or a passive type (a type in which distance measurement is performed based on an image signal of a subject). That is, when the contrast of the image signal is weak, the process branches to step S110, the distance measuring light is irradiated again by the auxiliary light irradiation, and the active AF is performed for the priority area previously obtained based on the reflected signal light. 3 per area
The process is performed point by point (step S121).

If it is determined in step S104 that the reflected signal light is weak, it is determined that the passive AF is more suitable, and the flow branches to step S105. Also in this case, three points are measured for each area by the passive method using the image signal of the priority area which has already been obtained. Also,
If the main subject cannot be found in step S103, the flow branches to step S120 to perform active distance measurement with emphasis on the central portion of the screen where the existence probability of the subject is high.

The CPU 501 controls the audio signal generator 507 according to the distance measurement method or whether the main subject can be determined.
By selecting and controlling the output of the audio signal generated by the
A distance measurement with a sense of security can be realized while promoting the features of F. It should be noted that the name of “super combination AF” is derived from the fact that the main subject is detected by using two methods instead of simply combining the active method and the passive method in a hybrid manner.

FIG. 21 is a flowchart for explaining the details of the priority determination process in step S102 in FIG. Steps S80 and S81 are steps for comparing the amounts of reflected signal light in order to detect the length and width of the screen composition. Here, FIGS. 22 (a) and 22 (b)
In the case of FIG. 22A, since the reflected light from the block is smaller than that in the case of FIG. 22B, the step S82 is easily branched to y. At this time, since it is considered that the main subject is unlikely to exist in the upper area of the screen, the weight of the position is set to 1 (step S
83). In addition, since it is considered that the main subject is likely to exist in the area, the weight is set to 4 (step S
84). In addition, the position weighting of the other areas is set to 2 assuming that the existence probability is intermediate (step S85).
And

On the other hand, in a composition as shown in FIG.
Is considered to be above, step S82 is branched to step S95, and the weight of the position is set to 1. The weight of the center of the screen is 8 (step S96), and the other weights are 2 (step S97).

The vertical and horizontal figures are often portraits, so the center weight is increased. In step S86, the area of the reflected light amount equal to or larger than the predetermined light amount is removed. This is a composition as shown in FIG. 23A, and the bin 601 and the table 602 in front of the object are preferentially measured for distance to the person 603. This is to prevent the camera from becoming out of focus. Next, in steps S87 to S90, weighting is performed based on the light amount in descending order of the light amount, and the obtained position and the weighting of the light amount are multiplied in step S91.
Are calculated. In step S92, the three high-priority areas obtained in this manner are selected and set as distance measurement candidate areas. In the example of FIG.
This is a portion surrounded by a thick line as shown in FIG. If the three areas obtained in this way are focused on three points and the other points are not measured, the time lag can be shortened,
Moreover, highly accurate focusing can be performed.

As described above, the correlation operation and the interpolation operation in the passive distance measurement tend to cause a time lag due to a complicated operation and require a large program capacity and a large memory capacity. By narrowing the points to a small area, it is possible to sufficiently cope with the above.

From the distance measurement effects of the three areas obtained in this way, a distance where the main subject is most likely to be selected is selected for focusing. At this time, first, only the correlation calculation is performed for three points in each area. Then, a rough distance distribution as shown in FIG. 23C is obtained, and after the focusing area is determined,
If the interpolation calculation is performed with emphasis on only one point, the effect can be further enhanced. If the distribution as shown in FIG. 23 (c) is obtained, the part where the distance changes sequentially from below can be determined to be a table and the person can be focused.

According to the above-described embodiment, the main subject position can be narrowed down by reducing the distance measuring points, and high-speed focusing can be performed. Simply reducing the distance measurement points would likely result in defocus if the main subject is located at a position where distance measurement was not performed. Since the scene is determined based on the distribution of the height, the main subject can be correctly focused.

The invention having the following configuration is extracted from the specific embodiment described above.

(Supplementary Note) (1) Luminance distribution measuring means for measuring the luminance distribution of the entire area in the screen, distance measuring area setting means for setting a distance measuring area in accordance with the measured luminance distribution, and this distance measuring A distance measuring means for performing a distance measuring operation for a distance measuring point in the area; and a means for evaluating a plurality of distance measuring results to specify an area including a main subject, and A distance measuring device for a camera, wherein the distance of the camera is determined.

(2) The distance measuring apparatus according to (1), wherein the luminance distribution measuring means measures a luminance distribution based on a distance measuring light emitted from a camera.

(3) A ranging device for a camera capable of ranging over the entire screen, wherein a first ranging region is set according to the luminance distribution in the screen, and based on the result of ranging in the first ranging region. A second ranging area including a main subject is extracted, a ranging operation is performed on the second ranging area, and a focusing position is determined.

[0101]

According to the present invention, a distance measuring apparatus for a camera capable of quickly and efficiently finding a distance measuring point while being inexpensively constructed using a simple arithmetic control circuit such as a normal one-chip microcomputer. Is provided.

[Brief description of the drawings]

FIG. 1 is a diagram showing a configuration of a distance measuring apparatus to which an embodiment of the present invention is applied.

FIG. 2 is a diagram mainly showing a detailed configuration of an integration unit in FIG. 1;

FIG. 3 is a timing chart when performing integral control by turning on a switch 7c.

FIG. 4 is a diagram showing a configuration in which a sensor array is added to measure a distance in an area perpendicular to the base line length direction.

FIG. 5 is a diagram showing a configuration for adding an area sensor to measure the entire distance in a screen.

FIG. 6 is a flowchart for explaining a distance measuring procedure according to the embodiment.

FIG. 7 is a front part of a flowchart for explaining details of a procedure for determining a priority in FIG. 6;

FIG. 8 is a latter part of a flowchart for explaining details of a procedure for determining a priority in FIG. 6;

FIG. 9 is a diagram for describing vertical / horizontal detection or upward detection of a camera.

FIG. 10 is a diagram for explaining a method of determining a position of a specific block on a divided screen.

11A is a diagram illustrating a specific distance measurement candidate area selected by priority determination, FIG. 11B is a diagram illustrating a distance distribution, and FIG. 11C is a diagram illustrating an example of a subject; is there.

FIG. 12 is a diagram for explaining an example of a method of selecting an optimum ranging area from an area sensor.

FIG. 13 is a diagram for explaining another example of a method of selecting an optimum ranging area from an area sensor.

FIG. 14 is a front part of a flowchart for describing a procedure for selecting a ranging point to be prioritized;

FIG. 15 is the latter part of the flowchart for explaining the procedure when selecting a ranging point to be prioritized.

FIG. 16 is a diagram illustrating an example of a screen having a space removed from a distance measurement priority area and an area in the center of the screen that is a priority area.

FIG. 17 is a diagram for explaining a procedure in a case where distance measurement is performed by switching the number of areas.

FIG. 18 is a flowchart for explaining a procedure for performing distance measurement by switching an area to be measured or the number of distance measurement points in accordance with luminance information in a screen.

FIG. 19 is a diagram for describing an autofocus technique of the super combination AF.

FIG. 20 is a flowchart illustrating details of a distance measurement procedure of the super combination AF.

FIG. 21 is a flowchart illustrating details of a priority determination process in step S102 of FIG. 20;

FIG. 22 is a diagram illustrating an example of two screens used for comparing the amounts of reflected signals to detect the length and width of the screen composition.

FIG. 23 is a diagram for explaining a method of removing an area having a reflected light amount equal to or larger than a predetermined light amount.

FIG. 24 is a diagram showing the contents of “Equation 1”.

FIG. 25 is a diagram showing the contents of “Equation 2”.

FIG. 26 is a diagram showing the contents of “Equation 3”.

FIG. 27 is a diagram illustrating the content of “Equation 4”.

FIG. 28 is a diagram showing a basic configuration of a conventional distance measuring device.

FIG. 29 is a diagram for explaining a conventional distance measuring procedure.

FIG. 30 is a diagram for explaining the calculation principle of distance measurement.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 CPU 1a A / D converter 1b Integral control part 1c Correlation calculating part 2a, 2b Sensor array 3a, 3b Light receiving lens 4 Focusing part 5 Subject 6 Maximum integrated value detection circuit 7 Integrating part 8 Selecting part A1-A4 Integrating amplifier

 ──────────────────────────────────────────────────続 き Continuing on the front page F term (reference) 2F065 AA06 BB05 DD06 FF01 FF05 FF09 JJ03 JJ05 JJ26 QQ03 QQ13 QQ14 QQ25 QQ27 QQ38 QQ42 QQ47 RR02 SS13 SS15 UU05 2F112 AC03 BA05 CA02 FA03 FA07 FA01 BA35 BA04 2H BA17 BA39 BB07 BB24 CB20 CC07 DA22

Claims (5)

[Claims] A first division unit that divides the screen into M points; a luminance measurement unit that measures a luminance distribution of each of the M points; and a luminance measurement result obtained by the luminance measurement unit. Selecting means for selecting N points among the M points; second dividing means for further dividing each of the N points selected by the selecting means in determining a main subject position; A distance measuring device for a camera, comprising: 2. The distance measuring apparatus according to claim 1, wherein said luminance measuring means measures a luminance distribution when auxiliary light is projected from a camera. 3. The method according to claim 1, wherein each of the N points has a plurality of distance measurement points, and the position of the main subject is determined based on the plurality of distance measurement results. The distance measuring device as described. 4. A camera for determining a position of a main subject in a screen, a sensor array for determining a luminance distribution in the screen, and dividing the screen into M points based on an output of the sensor array. Determining means for performing a distance measurement operation on a limited number of N points among these M points and determining a position of a main subject based on a plurality of pieces of distance information obtained by the distance measurement operation. Camera ranging device. 5. The distance measuring apparatus according to claim 4, further comprising switching means for switching the number of points for which a distance measuring operation is performed among the M divided points.