JP2010008507A - Camera - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To perform accurate AF by following a subject moving at high speed. <P>SOLUTION: A camera has: a thin film mirror 201 which divides a photographic luminous flux from a photographic optical system 101 which images a luminous flux from a subject, and guides the resulting luminous fluxes to an imaging element 202 and a range finding block 215 including a focus detection part 206 which detects a focal position by a phase difference detection system; a mobile body detection part 203 which detects a mobile body based on a signal from the imaging element 202; and a lens control part 102 which adjusts the focus position of the photographic optical system based on the focal position from the focus detection part 206. The camera further has: a control part 208 which controls the timing of the focus detection part 206 and the mobile body detection part 203; and a prediction arithmetic part 209 which predicts a motion of the subject based on focus detection information detected by the focus detection part 206 and mobile body information detected by the mobile body detection part 203. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to a camera, and more particularly to an autofocus (hereinafter abbreviated as AF) technique in a digital camera.

  Conventional AF techniques in digital cameras are roughly classified into a contrast AF method and a TTL (Through the Lens) phase difference AF method.

  In general, compact digital cameras often use the contrast AF method. The high-frequency component of the subject is detected while moving the focus adjustment lens (focus lens) in the direction of the optical axis, and the peak value of the contrast is calculated. By doing so, focus detection is performed. The contrast AF method has a feature that the focus detection speed is not high, but high-precision focus control is possible. A technique for detecting the face of a subject by using the output of an image sensor and focusing on the detected face is known.

  On the other hand, the phase difference AF method is often used for AF of a lens interchangeable digital single-lens reflex camera, and has a feature that focus detection can be performed at high speed. In this phase difference AF method, an AF sensor for detecting a shift in the focal position is provided in the camera body, and focus detection of the focus lens is performed based on a focus shift amount (defocus amount) detected by the AF sensor. However, since the distance measuring points (element arrangement of the AF sensor) must be discretely arranged, it is difficult to follow the subject in the vertical and horizontal directions across the distance measuring points.

  In view of such a problem, there is a need for a high-speed AF technique that can follow the vertical and horizontal directions in the screen.

As a prior example of such a technique, there is Patent Document 1. In this method, face detection is performed using the output of the image sensor, grouping is performed according to the size and position of the face for each photographing, and an optimum focus detection area is designated based on this.
JP 2007-286255 A

  In Patent Document 1, a focusing operation is performed using only distance measurement information of a distance measurement point corresponding to an area in which a face is detected using the output of an image sensor. However, for a fast-moving subject, the subject is not always in the same place during distance measurement and shooting, and an error occurs between the defocus amount at the time of distance measurement and the defocus amount at the time of shooting. There is a problem. Even when the subject is not a person's face, there is a problem in tracking because face detection information cannot be used.

  The present invention has been made in view of the above points, and an object of the present invention is to provide a camera capable of AF tracking with high accuracy with respect to a subject that moves at high speed in the vertical and horizontal directions within a shooting screen.

One aspect of the camera of the present invention is:
A group of photographic lenses that image the luminous flux from the subject;
Photoelectric conversion means for converting imaging light from the photographing lens group into an electrical signal;
A focus detection means for detecting a focus position of the photographing lens group by a phase difference detection method;
Lens control means for adjusting the focus position of the photographing lens group based on the focus position detected by the focus detection means;
A light splitting means for splitting a photographing light flux into a plurality of parts and directing the light flux to the photoelectric conversion means and the focus detection means;
Moving body detection means for detecting a moving body based on the signal generated by the photoelectric conversion means;
Control means for controlling the timing of the focus detection means and the moving object detection means;
Based on the focus detection information detected by the focus detection means and the mobile body information detected by the mobile body detection means, a prediction calculation means for predicting the motion of the subject,
It is characterized by comprising.

  According to the present invention, the photographing light flux is always guided by the light splitting means to the photoelectric conversion means and the phase difference detection type focus detection means, and subject motion prediction based on the detected focus detection information and the detected subject position. By performing the above, it is possible to provide a camera capable of AF tracking with high accuracy with respect to a subject that moves at high speed in the vertical and horizontal directions within the shooting screen.

  The best mode for carrying out the present invention will be described below with reference to the drawings.

  FIG. 1 is a diagram showing a configuration of a camera according to an embodiment of the present invention.

  This camera includes a lens body 10 and a camera body 20.

  The lens body 10 is provided as an interchangeable lens configured to be detachable from the camera body 20, and includes a photographing optical system 101, a lens control unit 102, a communication unit 103, a lens side contact 104, and the like.

  Here, the photographing optical system 101 is a photographing lens group that forms an image of a light beam from a subject. The lens control unit 102 is a lens control unit that controls each unit in the lens main body 10, and adjusts the focus position of the photographing optical system 101 according to a command from the camera main body 20 or a lens position (not shown). A function of measuring the lens position by the detection mechanism and transmitting the lens pulse to the camera body 20 by the communication unit 103 is provided. The lens-side contact 104 comes into contact with a corresponding contact of the camera body 20 in a state where the lens body 10 is mounted on the camera body 20 and is electrically connected, and the lens control unit 102 is connected by the communication unit 103. Can communicate with the camera body 20.

  The camera body 20 includes a thin film mirror 201, an image sensor 202, a moving body detection unit 203, a memory 204, an AF sensor 205, a focus detection unit 206, a memory 207, a control unit 208, a prediction calculation unit 209, a memory 210, and a main CPU 211. , Communication unit 212, camera side contact 213, and the like.

  Here, the thin film mirror 201 is a light splitting unit that splits a photographing light beam from the photographing optical system 101 into a plurality of light beams and guides the light beams to the image sensor 202 and the AF sensor 205, for example, a pellicle mirror or the like is used. The

  The image sensor 202 is a photoelectric conversion unit that converts the imaging light from the photographing optical system 101 divided by the thin film mirror 201 into an electric signal. The moving body detection unit 203 is a moving body detection unit that detects a moving body based on the signal generated by the image sensor 202 and stores the detected moving body information in the memory 204. The imaging element 202, the moving object detection unit 203, and the memory 204 constitute a moving object detection block 214.

  The AF sensor 205 has a plurality of separator diaphragm masks and separator lenses that divide the light flux from the photographing optical system 101 divided by the thin film mirror 201 into a plurality of pieces, and a plurality of pieces of image light that is divided into the plurality of pieces, respectively. An AF sensor module including a line sensor and the like. As shown in FIG. 2, the AF points 205A of the AF sensor are discretely arranged. The focus detection unit 206 is a focus detection unit that detects the focus position of the photographing optical system 101 by the phase difference AF method based on a plurality of signals generated by the AF sensor 205, and stores the detected focus detection information in a memory. Store in 207. The AF sensor 205, the focus detection unit 206, and the memory 207 constitute a distance measuring block 215.

  As will be described later, the control unit 208 functions as a control unit that controls the operation timing of the moving object detection block 214 and the distance measurement block 215. The prediction calculation unit 209 is a prediction calculation unit that performs subject motion prediction based on the focus detection information detected by the distance measurement block 215 and the moving body information detected by the moving body detection block 214. Is stored in the memory 210.

  The main CPU 211 controls each part in the camera body 20. The prediction calculation unit 209 operates under the control of the main CPU 211, calculates a predicted lens position based on the motion of the subject predicted and stored in the memory 210, and transmits the predicted lens position to the lens control unit 102 through the communication unit 212. Then, control is performed so that the lens is driven to the predicted lens position.

  The camera-side contact 213 comes into contact with the lens-side contact 104 of the lens body 10 in a state where the lens body 10 is mounted on the camera body 20 and is electrically connected. The prediction calculation unit 209 and the main CPU 211 can communicate with the lens control unit 102 in the lens body 10.

  FIG. 3 is a diagram showing a flowchart of the control operation of the distance measuring block 215 by the control unit 208. When the main CPU 211 receives a distance measurement start request corresponding to the first stage operation of a two-stage release button (not shown) from the main CPU 211 via the prediction calculation section 209, the control section 208 starts the operation of this flowchart.

  That is, first, the main CPU 211 gives a distance measurement end request in accordance with the release of the first operation of the release button or the like through the prediction calculation unit 209 or a shooting request by the second operation of the release button. It is determined whether or not it has been made (step S11). Here, if a distance measurement end request or a photographing request is not given, it is further determined whether or not a distance measurement timing has come (step S12). As the distance measurement timing, as described later, the distance measurement timing may be in accordance with a synchronization pulse given from the main CPU 211 via the prediction calculation unit 209, or may be one that measures a predetermined period with a self-timer. If the timing is not reached, the process returns to step S11.

  Thus, when the distance measurement timing comes, accumulation control (exposure) is performed for all distance measurement points 205A of the AF sensor 205 (step S13). Thereafter, the data of each ranging point of the AF sensor 205 is read out (step S14) and given to the focus detection unit 206, and the ranging calculation by the phase difference method, that is, the defocus amount calculation is performed on each ranging point (step S15). ). Then, the focus detection unit 206 stores the obtained distance measurement data (range measurement value, distance measurement interval, distance measurement result) in the memory 207 (step S16), and the process returns to step S11.

  In this way, distance measurement data is acquired and stored in the memory 207 at each distance measurement timing.

  If a distance measurement end request or a photographing request is given from the main CPU 211 via the prediction calculation unit 209 (step S11), the operation of this flowchart ends.

  FIG. 4 is a diagram showing a flowchart of the control operation of the moving object detection block 214 by the control unit 208. When the prediction calculation unit 209 selects a ranging point from the plurality of ranging points 205A measured by the focus detection unit 206 and gives a moving body detection request, the control unit 208 in the flowchart of FIG. Start operation.

  That is, first, a moving body detection end request corresponding to the release of the first operation of the release button or the like from the main CPU 211 via the prediction calculation unit 209 or a shooting request by the second operation of the release button is issued. It is determined whether or not it has been given (step S31). Here, if the moving object detection end request or the imaging request is not given, it is further determined whether or not the moving object detection timing has come (step S32). As described later, the moving body detection timing may correspond to a synchronization pulse provided from the main CPU 211 via the prediction calculation unit 209, or may measure a predetermined period using a self-timer. If it is not the moving object detection timing, the process returns to step S31.

  Thus, when it is time to detect the moving body, the image sensor 202 is exposed (imager exposure) (step S33). Thereafter, the imager data is read from the image sensor 202 (step S34), and is given to the moving body detection unit 203 to perform the moving body detection (step S35). Then, the mobile body detection unit 203 stores the obtained mobile body detection data (X position, Y position, mobile body detection result) in the memory 204 (step S36), and the process returns to step S31.

  In this way, mobile body detection data (mobile body information) is acquired and stored in the memory 204 at each mobile body detection timing.

  Then, if a moving body detection end request or a photographing request is given from the main CPU 211 via the prediction calculation unit 209 (step S31), the operation of this flowchart ends.

  Note that there are various known techniques for the moving object detection method performed by the moving object detection unit 203 in step S35 as described below.

(A) Moving object detection using difference image
(1) Background subtraction method: A moving object is detected by taking a difference between a background image in which no moving object exists and an image in which a moving object exists. It does not hold if there is no background-only image.

    (2) Interframe difference method: Three images are used. The first and second difference images and the second and third difference images are generated, and the moving object detection of the second image is performed by performing a logical product process (AND) of the generated two images. . Although an image of only the background may not exist, an object (such as a tree) that moves minutely in the background may be detected.

    (3) Statistical background difference method: Use Bayes' theorem. There must be a background-only image.

(B) Moving object detection using optical flow
(1) Block matching: Search for a point with the smallest difference from the target region (template) using template matching. The robustness such as the rotational movement and enlargement / reduction of the object is high, but the processing time becomes long when the moving amount of the moving body is large between images.

    (2) Gradient method: This is established under the preconditions that “the moving amount of the moving body is small between images” and “the movement in the object changes smoothly”. If all the areas are not searched, the processing speed is high, but if the constraint condition is not satisfied, the flow accuracy deteriorates.

  Considering the above moving object detection method, since the image captured by the digital camera is hand-held, there is an influence due to background image blurring and camera shake of the photographer. The gradient method is suitable. However, since there is a difference in accuracy and processing time, the gradient method with a fast processing time for a high-speed moving object, the gradient method for a low-speed moving object, or the like may be used depending on the shooting situation. Details of position detection by block matching and gradient method are described below.

First, block matching will be described.
Assuming that the pixel value of the reference region is I (x, y) and the pixel value of the next frame image is T (x, y), the similarity SSD (Sum of Squared) between the two images is given by the following equation (1). Find the difference.

  Here, the size of the reference template is M * N, the pixel value at the template position (i, j) is T (i, j), and the pixel value of the target image superimposed on the template is I (i, j). And

  The value of Rssd becomes 0 when it completely matches the template, but in reality, it hardly becomes 0 due to changes in noise or subject brightness. It can be said that the position where the minimum Rssd within the search area of the target image is the position where the template (target subject) has moved in the next frame image. With this position as the latest position, the X position and the Y position are stored. Since the calculation speed, detection accuracy (area), and the like vary depending on the template size M * N, the size of the template and the size of the search area for the target image in the camera mode depend on the speed priority and accuracy priority (detection area). May be switched.

Next, the gradient method will be described.
When the pixel value at the target pixel (x, y) is I (x, y), it is assumed that the pixel has moved (Δx, Δy) after the time Δt. Since it can be expected that the pixel value of this pixel does not change, the following equation (2) is established.

I (x, y) = I (x + Δx, y + Δy, t + Δt) (2)
Next, assuming that the pixel value is changing smoothly, if the right side of the above equation (2) is Taylor expanded and the second and subsequent terms are ignored as small, the following equation (3) is obtained.

When both sides of the above equation (3) are divided and arranged by Δt, the following equation (4) is obtained.

Since the movement amount is small in a very short time, the following equation (5) is obtained.

Here, the movement amount (optical flow) of the pixel is represented by the following expression (6).

  Then, assuming that the gradient of the pixel value is Ix, Iy and the time differentiation of the pixel value is It, the following equation (7) is obtained.

Ix · u + Iy · v + It = 0 (7)
This is an optical flow constraint condition formula.

  In this equation (7), there are two unknowns, u and v, and a solution cannot be obtained as it is.

  For this reason, the same optical flow is obtained from a plurality of pixels in the vicinity on the assumption that the movement is smooth in the vicinity of the target pixel. As a result, a plurality of optical flows are obtained from the pixels in the vicinity of the target pixel, that is, a plurality of constraint conditions are obtained, and the optimum (u, v) is obtained. If (u, v) is obtained in this way, the latest X position and Y position are calculated and stored using this as follows.

X position = I (x) + u · t
Y position = I (y) + v · t
5A and 5B are diagrams illustrating a series of flowcharts of the prediction calculation operation of the prediction calculation unit 209. FIG. The prediction calculation unit 209 starts the operation of this flowchart when a prediction calculation request is given from the main CPU 211 in response to a first-stage operation of a release button (not shown).

  That is, first, it is determined whether or not a prediction calculation end request corresponding to the release of the first operation of the release button or a shooting request by the second operation of the release button has been given from the main CPU 211. (Step S41). Here, if the prediction calculation end request or the imaging request is not given, it is further determined whether or not the mobile body follow-up flag configured inside or using a part of the memory 210 is turned on (step S42). .

  This moving body follow-up flag is off in the initial state. Therefore, at first, it is determined that the moving body follow-up flag is off. In this case, the distance measurement value acquired by the distance measurement operation in the distance measurement block 215 and stored in the memory 207 is read out through the focus detection unit 206, and AF reliability is determined based on the distance measurement value. A distance measuring point to be used for prediction calculation is selected from a plurality of distance measuring points 205A as shown in FIG. 2 by a known distance measuring point selection method such as a high distance measuring point or the closest distance measuring point (step S43). ).

  Thereafter, the moving object detection data detected by the moving object detection operation in the moving object detection block 214 and stored in the memory 204 is read out via the moving object detection unit 203, and based on the moving object detection data, the above described It is determined whether or not a moving body area exists at the selected distance measuring point (step S44).

  Here, if there is no moving object region at the selected distance measuring point, it is further determined whether or not the distance measuring point selected in step S43 is the same as the previously selected distance measuring point ( Step S45). Note that since the previous selected distance measuring point itself does not exist, it is naturally determined that they are not the same. As described above, when the current selected distance measuring point is different from the previous distance measuring point, the predicted lens position is calculated based on the distance value (defocus amount) of the distance measuring point selected in step S43. (Step S46), the process returns to step S41. The predicted lens position calculated in this way is transmitted by the communication unit 212 to the communication unit 103 of the lens body 10 via the camera side contact 213 and the lens side contact 104, and is transmitted from the communication unit 103 to the lens control unit 102. The photographing optical system 101 is driven to the predicted lens position.

  If it is determined in step S45 that the current selected distance measurement point and the previous distance measurement point are the same, a distance measurement prediction function calculation is performed (step S47). Then, the predicted lens position is calculated from the distance measurement prediction function of the distance measurement point selected in step S43 (step S48), and the process returns to step S41.

  FIG. 6A is a diagram showing an operation flowchart of the distance measurement prediction function calculation in step S47, which is performed for all distance measurement points 205A.

  That is, first, the distance value related data used for the distance prediction function calculation is determined and stored in the memory 210 (step S471). This is to determine and store α data retroactively from the latest among the data in which the distance measurement result is OK (AF possible). Α is set in consideration of prediction accuracy and calculation speed. Next, a secondary prediction function is calculated from the distance measurement value, distance measurement interval, and lens position using the exposure center-of-gravity time of the latest distance measurement as a reference (step S472), and the coefficient (Ad / Bd / Cd) is stored in the memory 210 (step S473).

  This distance measurement prediction function calculation will be described with reference to FIG. In FIG. 7, a black circle indicates the lens position at the time of exposure of the AF sensor 205 at the time of distance measurement (at the time of exposure center of gravity (1/2 of the exposure time T)), and Def [area] [i-3] described in the vicinity thereof. ]. ldpls, Def [area] [i-2]. ldpls, Def [area] [i-1]. ldpls, Def [area] [i]. ldpls is the value of these lens positions. Here, [area] is a variable representing the distance measuring point 205A, and [i-3], [i-2], [i-1], and [i] are time series variables. Further, g_base_time_af indicates the exposure center-of-gravity time of the latest distance measurement, and nowtime is the current time. As described above with reference to FIG. 3, the distance measurement is performed by sequentially performing exposure, reading, and distance value calculation (defocus amount calculation), and distance measurement prediction function calculation is performed thereafter. The lens driving is started after the first moving object prediction function calculation. Dt (n) is the time for the n + 1th distance measurement.

A white circle is a lens position of the ideal image plane, and L (n) is a value of the lens position. The lens positions of these ideal image planes are obtained as follows.
L0 = Def [area] [i-3] .ldpls + Def [area] [i-3] .Def * Defocus lens position conversion factor
L1 = Def [area] [i-2] .ldpls + Def [area] [i-2] .Def * Defocus lens position conversion factor
L2 = Def [area] [i-1] .ldpls + Def [area] [i-1] .Def * Defocus lens position conversion factor
L3 = Def [area] [i-0] .ldpls + Def [area] [i-0] .Def * Defocus lens position conversion coefficient Here, Def [area] [i-0]. Def is a distance measurement value (defocus amount) acquired by distance measurement. The defocus lens position conversion coefficient is a coefficient that varies depending on the zoom state of the photographing optical system 101, for example, a defocus amount of 30 μm is set to five lens pulses. The defocus lens position conversion coefficient can be acquired in advance from the lens control unit 102 by communication.

  If α = 4 in step S471, the lens position Def [area] [i-3]. ldpls to Def [area] [i-0]. ldpls, ranging value Def [area] [i-3]. Def-Def [area] [i-0]. Def and exposure center-of-gravity times T0 to T3 are stored in the memory 210. In step S472, the lens positions L0 to L3 of the ideal image plane are obtained as described above based on the stored data, and the lens positions of these ideal image planes are obtained. A prediction function is calculated from the L0 to L3 and the exposure center-of-gravity times T0 to T3 of each distance measurement using a quadratic approximation expression by the least square method or Lagrange interpolation. In step S473, the coefficient (second order / first order / 0th order) of the quadratic approximate expression is stored in the memory 210 as the coefficient (Ad / Bd / Cd) of the distance measurement prediction function.

Thus, in step S48, from the coefficient (Ad / Bd / Cd) stored in the memory 210, the predicted lens position Lt that is the target position at the target time Tt is
Lt = Ad * Tt ² + Bd * Tt ¹ + Cd
Therefore, the predicted lens position L4 at the target time T4 in this case is
L4 = Ad * T4 ² + Bd * T4 ¹ + Cd
Can be calculated. The predicted lens position calculated in this way is transmitted by the communication unit 212 to the communication unit 103 of the lens body 10 via the camera side contact 213 and the lens side contact 104, and is transmitted from the communication unit 103 to the lens control unit 102. The photographing optical system 101 is driven to the predicted lens position.

  On the other hand, if it is determined in step S44 that a moving body region exists at the selected distance measuring point, the moving body detection flag is turned on (step S49). Then, the predicted lens position is calculated based on the distance measurement value (defocus amount) of the distance measurement point selected in step S43 (step S50), and the process returns to step S41. The calculated predicted lens position is transmitted to the communication unit 103 of the lens body 10 via the camera side contact 213 and the lens side contact 104 by the communication unit 212, and is transmitted from the communication unit 103 to the lens control unit 102. The photographing optical system 101 is driven to the predicted lens position.

  If it is determined in step S42 that the moving body follow-up flag is on, the same distance measurement prediction function calculation as in step S47 is performed (step S51). However, in the distance measurement prediction function calculation in step S51, if the distance measurement data has not been updated, no processing is performed. After the distance measurement prediction function calculation, further, a mobile object detection prediction function calculation is performed (step S52), and the target mobile object position (X position Xf_dest, Y position) at the target time is calculated from the calculated mobile object detection prediction function. Yf_dest) is calculated (step S53).

  FIG. 6B is a diagram showing an operation flowchart of the mobile object detection prediction function calculation in step S52.

  That is, first, the mobile body detection related data used for the mobile body detection prediction function calculation is determined and stored in the memory 210 (step S521). This is to determine and store β pieces of data detected from the moving object from the latest to the past. Note that β is set in consideration of prediction accuracy and calculation speed. Next, a secondary prediction function is calculated from the X position, the Y position, and the exposure interval with reference to the exposure center-of-gravity time of the latest moving body detection (step S522). Here, the prediction function calculation for detecting the moving body is performed for a total of two patterns of X position / Y position. And the coefficient (X position: Am_x / Bm_x / Cm_x, Y position: Am_y / Bm_y / Cm_y) of the calculated mobile body detection prediction function is stored in the memory 210 (step S523).

  When there are a plurality of moving object detection results, that is, when a plurality of moving objects are detected in the screen by one exposure, the moving object detection prediction function is calculated for each. Further, the prediction calculation is performed by regarding the moving body having the closest X position / Y position detected during the front-back exposure as the same moving body.

  This moving object detection prediction function calculation will be described with reference to FIG. In FIG. 8, the white circle X (n) indicates the X position of the moving body when the image sensor 202 is exposed (at the time of exposure center of gravity), and the white triangle Y (n) similarly indicates the Y position. Yes. Here, the X position and the Y position of the moving body are xy coordinates (x, y) of the center of the moving body region, as shown in FIG. The origin of the xy coordinates is the upper left of the screen when the camera body 20 is held in the horizontal position, as shown in FIG. The moving body detection timings T0 to T3 in FIG. 9B correspond to those in FIG. Note that the first moving object detection (during (T0) exposure) sets the pixel area (template) of the image sensor output at the position corresponding to the distance measurement point selection by the phase difference method.

  As described above with reference to FIG. 4, the moving body detection is performed by sequentially performing exposure, reading, and moving body detection calculation, and the moving body detection prediction function calculation is performed thereafter. Note that Ft (n) in FIG. 8 is the exposure centroid time in the (n + 1) th moving object detection.

  In the above step S521, if β = 4, the X position X (0) to X (3) and Y position Y (0) to Y (3) of the moving body detection and the exposure gravity center time T0 to T3 are stored in the memory 210. In step S522, a prediction function is calculated from the stored data using a quadratic approximation formula for each of the X position and the Y position by the least square method or Lagrange interpolation. In step S523, the coefficients (second order / first order / 0th order) of the quadratic approximation equations are used as the coefficients (X position: Am_x / Bm_x / Cm_x, Y position: Am_y / Bm_y / Cm_y) and stored in the memory 210.

Thus, in step S53, the predicted X position Xm_dest of the moving body at the current target time Tt is calculated from the coefficients (X position: Am_x / Bm_x / Cm_x, Y position: Am_y / Bm_y / Cm_y) stored in the memory 210. And the predicted Y position Ym_dest are respectively
Xm_dest = Am_x * Tt ² + Bm_x * Tt ¹ + Cm_x
Ym_dest = Am_y * Tt ² + Bm_y * Tt ¹ + Cm_y
Can be calculated. Thus, for example, if the current target time is T4, the predicted X position Xm_dest (indicated by a black circle in FIG. 8) and the predicted Y position Ym_dest (indicated by a black triangle in FIG. 8) of the moving body are respectively
Xm_dest = Am_x * T4 ² + Bm_x * T4 ¹ + Cm_x
Ym_dest = Am_y * T4 ² + Bm_y * T4 ¹ + Cm_y
Can be calculated.

  If the current target moving object position is calculated from the moving object detection prediction function as described above, then the calculated X position and Y position of the target moving object, and whether the movement amount is equal to or less than the threshold value. Is determined (step S54). Here, when the threshold value is exceeded, it is considered that the following moving body has moved at a speed higher than expected or the moving body detection is erroneously detected. Therefore, it is preferable not to use the calculation result. Therefore, the moving body follow-up flag is turned off (step S55), and a distance measuring point is reselected by a known distance measuring point selection method such as a distance measuring point with high AF reliability or the closest distance measuring point ( Step S56). Then, a predicted lens position is calculated based on the distance measurement value (defocus amount) of the reselected distance measurement point (step S57), and the process returns to step S41. The predicted lens position calculated in this way is transmitted by the communication unit 212 to the communication unit 103 of the lens body 10 via the camera side contact 213 and the lens side contact 104, and is transmitted from the communication unit 103 to the lens control unit 102. The photographing optical system 101 is driven to the predicted lens position.

  On the other hand, if it is determined in step S54 that the calculated X and Y positions of the target moving body and the amount of movement are equal to or less than the threshold value, this corresponds to the current detected moving body position. It is determined whether or not the range-finding value of the range-finding point is reliable (step S58). If it is determined that the distance measurement value at the distance measurement point is not reliable, the process proceeds to step S55.

  If it is determined that there is also reliability, the distance measuring point 205A corresponding to the target moving body position (following moving body position) calculated in step S53 is reselected (step S59), and the moving body following prediction function is selected. Calculation is performed (step S60). Then, a predicted lens position is calculated from the calculated moving body follow-up prediction function (step S61), and the process returns to step S41.

  FIG. 10 is a diagram showing an operation flowchart of the mobile object follow-up prediction function calculation in step S60, and this operation is performed for the reselected distance measuring point.

  That is, first, distance measurement value-related data used for the mobile object follow-up prediction function calculation is determined and stored in the memory 210 (step S601). This is to determine and store α data retroactively from the latest among the data in which the distance measurement result is OK (AF possible). In this case, there is a possibility of straddling the distance measuring point 205A for tracking the moving body. Next, a secondary prediction function is calculated from the distance value, distance measurement interval, and lens position using the exposure center-of-gravity time of the latest distance measurement as a reference (step S602), and the coefficient (Add) of the calculated mobile object tracking prediction function is calculated. / Bdd / Cdd) is stored in the memory 210 (step S603).

  Since this mobile object follow-up prediction function calculation is the same operation as the distance measurement prediction function calculation, description thereof will be omitted.

Thus, in step S61, the predicted lens position Lt, which is the target position at the target time Tt, is calculated from the coefficient (Add / Bdd / Cdd) stored in the memory 210.
Lt = Add * Tt ² + Bdd * Tt ¹ + Cdd
Calculated by Therefore, for example, if the target time is T4, the predicted lens position L4 is
L4 = Add * T4 ² + Bdd * T4 ¹ + Cdd
Can be calculated. The predicted lens position calculated in this way is transmitted by the communication unit 212 to the communication unit 103 of the lens body 10 via the camera side contact 213 and the lens side contact 104, and is transmitted from the communication unit 103 to the lens control unit 102. The photographing optical system 101 is driven to the predicted lens position.

  Here, for example, if each distance measuring point 205A is represented by XY coordinates as shown in FIG. 11A, the distance measuring point 205A circled in FIG. If t is t, the time and coordinates are combined and written as (t, 6, 3). In this case, the relationship between the distance measurement position and the moving object detection position in the moving object follow-up prediction is as shown in FIG. In FIG. 11B, a broken-line rectangle indicates a moving object detection position, a black square distance measurement point indicates a distance measurement position corresponding to the movement object detection position, and a double-line rectangle is predicted by a moving object detection position. Respectively.

  In the case of the example of FIG. 11B, the current target time, for example, the predicted moving body detection position at T4, is obtained in step S53 in accordance with the moving body detection prediction function calculation in step S52, and the above step S59. Thus, the distance measuring point 205A at the coordinates (7, 3) is reselected. In step S60, the distance measurement value ((T1, 2, 1), (T2, 3, 1) and (T3, 5, 2)) of the distance measurement point that coincides with the moving object detection position is used to move. A body following prediction function calculation is performed, and the distance measurement value is predicted in step S61, and the predicted lens position L4 is calculated.

  As described above, the ranging block 215 and the moving object detection block 214 operate at synchronous or asynchronous timing.

  FIG. 12 shows a distance prediction function calculation, a mobile object detection prediction function calculation, and a mobile object tracking prediction when the operations are performed in synchronization with the exposure timing of the distance measurement block 215 and the exposure timing of the mobile object detection block 214, respectively. It is a figure which shows the timing of a function calculation.

  As shown in the figure, the exposure of the distance measurement operation and the exposure of the moving object detection are started by the prediction calculation unit 209 starting both exposure operations by the control unit 208 in accordance with the synchronization pulse from the main CPU 211. The Dt (n) and Ft (n) indicate the respective exposure centroid times. Further, Def (n) indicates distance measurement data, and F (n) indicates the calculated moving object detection prediction function.

  In this case, since the time required for ranging and the time required for moving object detection are different, the start time Bt (n) of the mobile object follow-up prediction function calculation by the prediction calculation unit 209 is the longer required time. Along with the end of a certain mobile object detection, it is after the end of the mobile object detection prediction function calculation.

  In this mobile object follow-up prediction function calculation, when the first mobile object follow-up prediction function Dest (0) is calculated, since α pieces of data are not yet prepared, the calculation using only the distance measurement data D (0) is performed. It becomes. When calculating the next moving body follow-up prediction function Dest (1), since α data are not yet prepared, the distance measurement data D (1), D (0), and the distance-exposure center-of-gravity time Dt (1) ), Dt (0), moving body detection prediction functions F (1), F (0), exposure center-of-gravity times Ft (1), Ft (0), and target time Tt (2) for moving body detection. It becomes. Then, when calculating the m + 1th moving object follow-up prediction function Dest (m), since α data are already prepared, the distance measurement data D (m) to D (m−α), the distance measurement data Exposure center of gravity time Dt (m) to Dt (m-α), moving body detection prediction function F (m) to F (m-α), exposure center of gravity time Ft (m) to Ft (m−α) for moving body detection , And the target time Tt (m + 1). The target time Tt (m + 1) is the timing of the next synchronization pulse.

  Then, as shown in FIG. 11, when a shooting request is made, the shooting process is started. However, in this case, as shown in the series of flowcharts of the prediction calculation operation in FIG. 5A and FIG. 5B, the calculation of the predicted lens position is performed and the lens is driven to the predicted position. As shown in FIG. 11, the m + 2th moving object follow-up prediction function Dest (m + 1) is calculated.

  On the other hand, FIG. 13 shows a distance measurement prediction function calculation, a mobile object detection prediction function calculation, and a mobile object tracking when the exposure timing of the distance measurement block 215 and the exposure timing of the mobile object detection block 214 are operated asynchronously. It is a figure which shows the timing of prediction function calculation.

  In this case, the exposure of the distance measuring block 215 is periodically performed at a predetermined first interval, and the exposure of the moving object detection block 214 is periodically performed at a predetermined second interval. The moving body tracking prediction function calculation is performed at the end of the distance measurement prediction function calculation.

  Also, the number of distance measurement prediction function calculations does not match the number of mobile object detection prediction function calculations. For example, when calculating the m + 1-th moving object follow-up prediction function Dest (m), distance measurement data D (m) to D (m−α), distance exposure center-of-gravity time Dt (m) to Dt (m) −α), moving object detection prediction functions F (k) to F (k−α), exposure object gravity center times Ft (k) to Ft (k−α), and target time Tt (m + 1) for moving object detection were used. It becomes an operation. The target time Tt (m + 1) is the exposure timing of the next ranging block 215.

  Also in this case, the imaging process is started when an imaging request is made. However, the calculation of the m + 2 mobile object tracking prediction function Dest (m + 1) is performed in the same manner as in the case of the synchronization shown in FIG. .

  As described above, the distance measurement block 215 and the moving object detection block 214 may be operated in synchronization or asynchronously.

  In the case of non-synchronization, the distance measurement block 215 is not periodically exposed, but the position of the moving object is predicted by the moving object detection prediction function calculation, and the moving object moves to the distance measuring point 205A. It is possible to obtain highly reliable distance measurement data by measuring the timing.

  That is, as shown in FIG. 14, the exposure of the moving object detection block 214 and the moving object detection prediction function calculation of the prediction calculation unit 209 are always performed periodically at the maximum speed. Then, as a result of the calculation of the moving object detection prediction function, the distance measuring timing control signal (synchronization pulse) is sent by the control unit 208 at the timing when the moving object moves to a position corresponding to the distance measuring point 205A according to the moving amount and moving speed of the moving object. The distance measurement is performed by giving the distance measurement block 215.

  When such distance measurement timing control is performed, the target time Tt (m + 1) used in the mobile object follow-up prediction function calculation is obtained from the movement amount / movement speed of the mobile object as a result of the mobile object detection prediction function calculation. This is the timing of the next synchronization pulse of the distance measurement timing control signal.

  As described above, according to the present embodiment, the photographing light flux is always guided by the thin film mirror 201 to the moving object detection block 214 and the distance measurement block 215, and the detected object position is detected by the prediction calculation unit 209. By predicting the movement of the subject based on the focus detection information, it is possible to follow the detected moving body even for subjects moving in the vertical and horizontal directions on the screen at high speed, and AF tracking can be performed with high accuracy for subjects moving at high speed. Camera can be provided.

  In addition, by performing focus detection by the distance measuring block 215 and moving body detection by the moving body detection block 214 while driving the lens, it is possible to provide a camera that enables high-speed and high-precision AF.

  Although the present invention has been described above based on the embodiments, the present invention is not limited to the above-described embodiments, and various modifications and applications are naturally possible within the scope of the gist of the present invention.

  For example, the lens body 10 is described as an example of a camera that can replace the camera body 20, but the lens body 10 may be a camera in which the lens body 10 is configured integrally with the camera body 20 and cannot be replaced.

  Further, the arrangement of the distance measuring points 205A of the AF sensor 205 is not limited to the arrangement shown in FIG. 2, and the number of distance measuring points 205A is not limited thereto.

FIG. 1 is a diagram showing a configuration of a camera according to an embodiment of the present invention. FIG. 2 is a diagram illustrating an example of the arrangement of the AF points of the AF sensor. FIG. 3 is a diagram illustrating a flowchart of the control operation of the ranging block by the control unit. FIG. 4 is a flowchart of the control operation of the moving object detection block by the control unit. FIG. 5A is a diagram illustrating a first part of a series of flowcharts of the prediction calculation operation of the prediction calculation unit. FIG. 5B is a diagram illustrating a second part of a series of flowcharts of the prediction calculation operation of the prediction calculation unit. FIG. 6A is a diagram illustrating an operation flowchart of the distance measurement prediction function calculation, and FIG. 6B is a diagram illustrating an operation flowchart of the moving object detection prediction function calculation. FIG. 7 is a diagram for explaining the concept of the distance measurement prediction function calculation. FIG. 8 is a diagram for explaining the concept of the mobile object detection prediction function calculation. FIG. 9A is a diagram for explaining the X position and the Y position of the moving body. FIG. 9B shows the movement of the moving body and examples of the X position and the Y position of the moving body. FIG. FIG. 10 is a diagram illustrating an operation flowchart of the mobile body follow-up prediction function calculation. FIG. 11A is a diagram for explaining the coordinates of the distance measurement point of the AF sensor, and FIG. 11B is a diagram for explaining the relationship between the distance measurement position and the mobile object detection position in the mobile object tracking prediction. FIG. FIG. 12 is a diagram illustrating timings of the distance measurement prediction function calculation, the mobile object detection prediction function calculation, and the mobile object tracking prediction function calculation when distance measurement and mobile object detection are performed in synchronization. FIG. 13 is a diagram showing timings of the distance measurement prediction function calculation, the mobile object detection prediction function calculation, and the mobile object tracking prediction function calculation when the distance measurement and the mobile object detection are performed asynchronously. FIG. 14 is a diagram illustrating timings of the distance measurement prediction function calculation, the mobile object detection prediction function calculation, and the mobile object tracking prediction function calculation when the distance measurement is performed at the timing when the mobile object moves to the distance measurement point.

Explanation of symbols

    DESCRIPTION OF SYMBOLS 10 ... Lens main body, 20 ... Camera main body, 101 ... Imaging optical system, 102 ... Lens control part, 103 ... Communication part, 104 ... Lens side contact, 201 ... Thin film mirror, 202 ... Imaging element, 203 ... Moving body detection part, 204, 207, 210 ... memory, 205 ... AF sensor, 205A ... distance measuring point, 206 ... focus detection unit, 208 ... control unit, 209 ... prediction calculation unit, 211 ... main CPU, 212 ... communication unit, 213 ... camera side Contact point 214... Moving object detection block 215.

Claims (8)

A group of photographic lenses that image the luminous flux from the subject;
Photoelectric conversion means for converting imaging light from the photographing lens group into an electrical signal;
A focus detection means for detecting a focus position of the photographing lens group by a phase difference detection method;
Lens control means for adjusting the focus position of the photographing lens group based on the focus position detected by the focus detection means;
A light splitting means for splitting a photographing light flux into a plurality of parts and directing the light flux to the photoelectric conversion means and the focus detection means;
Moving body detection means for detecting a moving body based on the signal generated by the photoelectric conversion means;
Control means for controlling the timing of the focus detection means and the moving object detection means;
Based on the focus detection information detected by the focus detection means and the mobile body information detected by the mobile body detection means, a prediction calculation means for predicting the motion of the subject,
A camera comprising:   The camera according to claim 1, wherein the focus detection unit and the moving body detection unit each operate at an asynchronous timing set by the control unit.   2. The camera according to claim 1, wherein the focus detection unit and the moving body detection unit operate at synchronized timings set by the control unit.   The camera according to claim 1, wherein the prediction calculation unit controls distance measurement timing of the focus detection unit based on moving body information detected by the moving body detection unit.   2. The prediction calculation unit according to claim 1, wherein when the information of the moving body detection unit is low in reliability, the prediction calculation unit performs motion prediction of the subject using only the focus detection information detected by the focus detection unit. The listed camera.   The camera according to claim 1, wherein the focus detection unit performs focus detection during lens driving by the lens control unit.   The camera according to claim 1, wherein the moving body detection unit performs moving body detection during lens driving by the lens control unit.   The camera according to claim 1, wherein the light splitting unit is a pellicle mirror.

Images