JPH05145822A - Moving body tracking device - Google Patents

Abstract

PURPOSE:To use an idle area of an addition video signal memory for tracking for a memory of a frequency component for focus detection in order to suppress the increase in the memory in which the focus detection frequency signal is stored to increase the AF accuracy of the tracking object. CONSTITUTION:A photoelectric conversion element 12 converts a light distribution from an image pickup optical system 11 into a distribution of an electric signal and a frequency detection circuit 13 extracts a specific frequency from a picture element signal and a projection signal of the photoelectric conversion element 12. Then a focus detection circuit 16 outputs a drive signal to the drive circuit 14 based on a signal from the frequency detection circuit 13 and the information of the image pickup optical system 11 to detect the focus of the image pickup optical system 11. Furthermore, the focus detection circuit 16 reads a signal of a specific area of the photoelectric conversion element 12 based on the specific area designation signal from the tracking circuit 15 or the area designation circuit 18 and the specific area or the like is displayed on a display circuit 17.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a moving body tracking device,
In particular, the present invention relates to a moving object tracking device that improves the AF accuracy of a tracking subject without increasing the memory.

[0002]

2. Description of the Related Art In recent years, various systems have been proposed in which a moving body is tracked based on a video signal.

For example, in Japanese Patent Application No. 02-104637,
A system is disclosed in which a video signal is projected in the vertical and horizontal directions to perform focus detection based on a differential value of a tracking signal while tracking a moving object. 16 (a) and 16 (b),
This outline is shown, and for the sake of simplicity, a point light source will be described. In the figure, XSUM indicates a signal projected in the vertical direction, and YSUM indicates a signal projected in the horizontal direction.
The MOS sensor is used to easily output XSUM and YSUM. 16 (a) and 16 (b)
Is for detecting the movement of the image at the time of focusing with XSUM and YSUM.

The NHK technique No. 1 is a system for detecting the in-focus point by paying attention to the change in the frequency component using a video signal.
It is disclosed in "Automatic Focus Adjustment of Television Camera by Mountain Climbing Servo System", which was published by Ishida et al.

[0005]

FIG. 17 shows the voltage values of lens defocus and frequency components. In the figure, the sharpness is high at the time of focusing, that is, the high frequency component is large, and conversely, the sharpness is low at the time of non-focusing,
This indicates that the edge portion is blurred, that is, the high frequency component becomes small. In this way, focusing on the change in the specific frequency component, the photographing optical system is driven to the peak position of the frequency component. That is, as shown in FIG. 18, such a focus detection system has an A / D converter 1,
Memory 2, tracking processing circuit 3, frequency detection circuit 4, A / D
It has a converter 5, a memory 6 and a focus detection circuit 7. Tracking A / from the same input video signal
The signal is A / D converted by the D converter 1 and stored in the memory 2. At the same time, the frequency value of the frequency detection circuit 4 is set to A
The data is converted by the / D converter 5 and stored in the memory 6. However, if the memory 6 storing the frequency value is accessed based on the result of the tracking processing and the focus detection is performed, the scale of the memory is increased.

Regarding the reading of the video signal at the time of tracking the moving object, the output in the tracking block is X-rayed by the MOS sensor.
There is a method of adding currents in the Y direction and the Y direction for reading. Also,
There is also a method of once converting a video signal into a digital value and performing addition by digital signal processing.

However, in these methods, since the focus detection is performed by using the projection signal, the detection accuracy is not high.

By the way, when moving object tracking and in-focus point detection are performed by the same sensor, it is necessary to detect a moving object position from a video signal and then perform frequency detection for in-focus point detection.
If the moving object detection and the frequency detection are performed by time division, it is difficult to detect the in-focus point while tracking the moving object. That is,
If the signals for frequency detection and moving object tracking are extracted from the same region from the video signal, the tracking object may be out of the focus area, and accurate focus detection cannot be performed.

FIGS. 19A, 19B and 19C show movements of time, a tracking object and a focus area. It is assumed that the fields (a), (b) and (c) in the figure are shifted by one field. In the figure (a), at the time of setting (t = t ₀ ), the tracking object (dotted line) and the focus (solid line) are in the same area. FIG. 6B shows that the tracking object is moving (t = t ₁ ), so that if tracking and focus detection are performed at the same time, the tracking object and the focus area are misaligned. Further, FIG. 7C shows that the tracking object and the focus area coincide with each other when the tracking object moves only in the vertical direction of the screen (t = t ₃ ). Therefore, the accuracy of focus detection decreases in the process from FIG.

In order to improve the accuracy of in-focus point detection,
If it is attempted to simultaneously hold a frequency signal corresponding to a video signal, the amount of memory to be stored will become huge and problems such as cost increase will occur.

It is an object of the present invention to provide a moving body tracking device capable of simultaneously performing moving body tracking and in-focus point detection with the same sensor with a simple structure without increasing the memory and increasing the cost. ..

[0012]

That is, the present invention specifies an image pickup optical system, a photoelectric conversion element for converting an object image through the image pickup optical system into an image signal, and an output signal of the photoelectric conversion element. Frequency detection means for extracting frequency components, focus area setting means for setting a focus area, addition means for calculating an addition value for each row of the photoelectric conversion elements in a tracking area wider than the focus area, and the focus area Having a capacity capable of storing the added value of all rows in the tracking area, and storing the added value for a region for performing a correlation calculation. And the second storage means for storing the specific frequency component in the vacant area, and the added value stored in each of the first and second storage means. A tracking means for performing a correlation calculation to perform tracking, and a focus detection means for performing focus detection on the focus area tracked by the tracking means based on the specific frequency component stored in the second storage means. It is characterized by

[0013]

In the moving body tracking device of the present invention, the photoelectric conversion element converts the subject image into an image signal via the image pickup optical system,
A specific frequency component is extracted from the output signal of the photoelectric conversion element by the frequency detecting means. Further, the focus area is set by the focus area setting means, and the addition value is calculated by the addition means for each row of the photoelectric conversion elements in the tracking area wider than the focus area. Then, the added value of the row including the focus area is stored in the first storage means for each row, and on the other hand, in the second storage means having a capacity capable of storing the added value of all the rows in the tracking area. Stores the added value in the area where the correlation calculation is performed, and also stores the specific frequency component in the empty area. In the tracking means, correlation calculation is performed based on the added values stored in the first and second storage means, and the specific area stored in the second storage means for the focus area tracked by the tracking means. Focus detection is performed by the focus detection means based on the component.

[0014]

Embodiments of the present invention will be described below with reference to the drawings.

FIG. 1 is a block diagram showing the concept of an embodiment of a moving object tracking device according to the present invention. In the figure, a photographing optical system 11 guides the speed of light from a subject to a photoelectric conversion element 12, and the photoelectric conversion element 12 converts the light distribution from the photographing optical system 11 into an electric signal distribution. The frequency detection circuit 13 extracts a specific frequency from the pixel signal and the projection signal of the photoelectric conversion element 12. Then, the focus detection circuit 16 drives the drive circuit based on the signal from the frequency detection circuit 13 and the information (MTF characteristic, lens position, aperture, focal length etc) of the photographing optical system 11 obtained through the drive circuit 14. A drive signal is output to 14, the focusing point of the photographing optical system 11 is detected, and a tracking circuit is also used.
15, or read the signal of the specific area of the photoelectric conversion element 12 based on the specific area designating signal from the area designating circuit 18,
Further, the display circuit 17 is caused to display a specific area, a focus mark and the like.

The drive circuit 14 outputs the information of the photographing optical system 11 to the focusing point detection circuit 16 and drives the photographing optical system 11 based on the drive signal from the focusing point detection circuit 16. Further, the tracking circuit 15 detects the position of the moving body based on the projection signal from the photoelectric conversion element 12, and outputs a specific area designation signal to the focusing point detection circuit 16 and the display circuit 17. Further, the display circuit 17 is for displaying the specific area from the tracking circuit 15 and the area designating circuit 18 and the focus signal from the focus detecting circuit 16, and the area designating circuit 18 displays the moving body position and the focus area from the outside. This is the circuit to set.

In the moving body tracking device having such a structure,
As shown in FIG. 2, the focus detection circuit 16 detects the tracking area.
19 is taken larger than the focus area 20, and the specific frequency signal of the frequency detection circuit 13 is sampled at a specific cycle in the entire tracking area 19. Then, the sampled signal is A / D converted, and the A / D converted signals of the area corresponding to the focus area 20 are added. This added signal has the frequency components shown in FIG. The tracking area 19 and the focus area 20 move within the screen as the moving body moves.

Here, for the sake of simplicity of explanation, the data of the specific frequency of the one-dimensional tracking region 19 is shown in FIGS. 3 (a) and 3 (b).
Shown in. In all areas of the tracking area 19, a specific frequency component is provided as a block smaller than the focus area 20 together with the position information of that block. As a result, FIG.
Even if the tracking object moves from (a) to (b) in the figure, after the position of the tracking object is detected, the specific frequency component of the area of the focus area corresponding to the detected position can be added. Furthermore, the moving object detection operation and the signal reading are performed almost at the same time, and a part of the memory used for moving object detection is temporarily used, that is, in the initial stage, a memory area in which no data is input is used, After the detection correlation is performed, the low correlation portion is repeatedly used in the memory area.
Then, the specific frequency component for each block is stored in the memory.

In this way, it is possible to track a moving object, perform in-focus detection using a specific frequency component of the focus area that has always been tracked, and achieve high speed and high accuracy with almost no increase in memory. become.

Next, details of the circuit shown in FIG. 1 will be described.

The photoelectric conversion element 12 is composed of a solid-state image pickup device such as a CCD and reads out a video signal after the integration of the solid-state image pickup device is completed. Further, the frequency detection circuit 13 is composed of a plurality of bandpass filters (BPF) and has a plurality of BPFs.
Are switched depending on the conditions of the image pickup optical system 11 and used to output a signal to the focus detection circuit 16.

The drive circuit 14 outputs various characteristics of the photographic optical system 11 to the focusing point detection circuit 16 and receives the drive signal from the focusing point detection circuit 16 to drive the photographic optical system.
Further, the tracking circuit 15 A / D-converts the signal in the tracking area 19 and projects the projection signals XSUM and YSUM as shown in FIGS. 16 (a) and 16 (b) by digital addition calculation processing.
To detect the position of the moving body and detect the position signal by the focus detection circuit 16
Is output to the display circuit 17. Further, the area designating circuit 18 for setting the focus area 20 is also used as a joystick or a fast release, or in Japanese Patent Application No. 2-235074.
The method disclosed in No.

The focusing point detection circuit 16 is a frequency detection circuit.
The signal of 13 is A / D converted and the digital values in the focus area are added. While driving the drive circuit 14, the peak position of the added value, that is, the focus position of the imaging optical system 11 is detected. The BPF switching control is performed by optical information such as the diaphragm of the photographing optical system 11, the focal length, the MTF characteristic, and the position of the photographing optical system 11. The display circuit 17 displays this information in the finder based on the information of the focus area 20 from the tracking circuit 15 and the area designating circuit 18. Furthermore, the focus state is displayed by the focus state signal from the focus detection circuit 16.

Next, FIG. 4 shows a configuration of a portion for detecting a specific frequency in the focus detection circuit 16.

The focus detection circuit 16 is a frequency detection circuit.
A / D conversion circuit 161 for A / D converting 13 signals, and
A memory circuit 162 for storing the data of the D conversion circuit 161;
The focus area 20 for focus detection and the tracking area 19 are set based on the position signal of the tracking moving body of the tracking circuit 15 or the signal from the area designating circuit 18, and only the frequency component of the focus area 20 including the moving body position is added. Operation processing circuit 163
Composed of.

Memory configuration of memory circuit 162 and tracking area 1
The relationship between 9 and the focus area 20 is as shown in FIG. In the figure, the tracking subject is captured in the focus area 20 at t = t ₀ , and the tracking subject is tracked at t = t ₁ . Let t = t _{1 be} the time to capture the image data next to t = t ₀ . The memory circuit 162 is a memory H
a, a memory Va, a memory Hb, and a memory Vb. In this case, the number of pixels in the tracking area 19 is Nth * Nt.
v, and the number of pixels in the focus area 20 is Nfh * Nfv, the memory Ha and the memory Va for storing the projection signals onto the V and H of the focus area 20 at t = t ₀ are N respectively.
The number of memories is fh and Nfv. Further, a memory Hb for storing projection signals on V and H of the tracking area 19 at t = t ₁ ,
The memory Vb has Nth and Ntv memory numbers, respectively.

Here, the signal of one pixel in the tracking area 19 in one screen is A (h, j) (where h is a parameter in the H direction and j is a parameter in the V direction). The operation of focus detection will be described with reference to the flowchart.

First, the subroutine program DF is started. Next, in step S1, the area setting circuit
At 18, the moving object position setting and the frequency component amount memory F (0) in the focus area 20, the focusing direction determination memory SS, the correlation value memory Sk0, and the subscripts i, j, and h are initialized (F (0) = 0, SS = 0, Sk0 = 99, i = j =
h = 1). In step S2, the focus area
20 is set based on step S1 described above. Thus, in step S3, a region centering on the focus area 20 and larger than the focus area 20 is set as the tracking region 19.

In step S4, the tracking data (V and H projection signals) in the focus area 20 are stored in the memories Va and Ha for the V and H directions, respectively. Then, in step S5, the specific frequency in the focus area 20 is detected, and F
It is stored in (0). Next, in step S6, the photographing optical system 11 is driven in the forward direction, and in step S7, the driving direction memory of the photographing optical system 11 in step S6 is set to FS = 0.

Here, the flag indicating the first correlation is set to WE.
= 0 (step S8). WE is related to which area of the memory Vb is empty. Then
It is determined whether or not WE = 0 (step S9).
In step S9, if the flag WE indicating the memory state is WE = 1, the process proceeds to step S10, and the frequency addition value of the focus area block is input to the newly inputtable memory area P or Q. Proceed to S12. On the other hand, when WE = 0, the process proceeds to step S11, the specific frequency components in the tracking area 19 are sampled at a constant interval, and A / D converted for each block smaller than the focus area 20 after A / D conversion. Memory Vb in which the specific frequency is added and new addition data is not stored yet
The data of the tracking correlation is stored in (lower address in the V direction) before mixing occurs.

Next, at step S12, the data read out and added (projected) in the V direction are stored in the memory Hb (h). Here, h is a loop in which the number of pixels in one line in the H direction is 1 to Nth. Then, in step S13, the data similarly added (projected) in the H direction is stored in the memory Vb (j). Then, in step S14, the 1H read end determination (h = Nth) is made. In this step S14, 1H
If the reading has not been completed, the process proceeds to step S15, the number of pixels h in the H direction is incremented, and then step S12.
Return to.

When 1H reading is completed in step S14, the number of read lines is determined (j ≧ Nfv) in step S16. If the Nfv line has not been read in step S16,
In step S17, j is incremented and h is set to 1. Then, the process returns to step S11. On the other hand, step S
If the Nfv line is read at 16, i is initialized (i = 1) at step S18. Then, in step S19, the V-direction memories Va (i) and Vb (j
-Nfv + i) is substituted into D0 (i) and D1 (i).

Then, in step S20, i = Nf
If v is determined and i = Nfv is not satisfied, the process proceeds to step S21, i is incremented, and then the process returns to step S19. If i = Nfv in step S20, the position detection subroutine program D in step S22.
The position in the V direction is detected by S. Then, step S
At 23, the flag WE indicating the first correlation is set to WE = 1.

At step S24, the correlation values Sk1 and Sk0
Are compared, and if Sk0> Sk1, step S
Proceeding to 25, Sk1 is substituted for Sk0. This indicates that the new data has a higher correlation and is used as the next criterion. Then, in step S26, as a memory into which a value obtained by adding the specific frequency to each block smaller than the focus area 20 in the tracking area 19 is substituted is set to Vb.
The memory area P in which the frequency component for each block corresponding to (j-Nfv + 1) is stored is cleared. On the other hand, if Sk0> Sk1 is not satisfied in step S24, it means that the new data has a lower correlation, and the specific frequency is added to each block smaller than the focus area 20 in the tracking area 19. As the memory into which the value is substituted, the memory area Q in which the frequency component of each block corresponding to Vb (j) is stored is cleared.

Next, in step S28, the tracking area 19
The read status of is detected (j = Ntv). If all the data has not been read in step S28, the process proceeds to step S29, j is incremented, and step S9
Return to. Then, in step S29, if the flag WE indicating the memory state is WE = 1, the process proceeds to step S10, and the frequency addition value of the focus area block is input to the newly enabled memory area P or Q. It On the other hand, if all the data in the tracking area 19 is read in step S28, step S30
To m is initialized (m = 0), and then
In step S31, i is initialized (i = 1).
Then, in step S32, Ha (i), Hb (m +
i) is substituted into D0 (i) and D1 (i). However, m
= 0 to Nth-Nfh.

Next, in step S33, it is determined whether i = Nfh. In this step S33, i = Nfh
If not, the process proceeds to step S34, i is incremented, and then the process returns to step S32. If i = Nfh in step S33, the position in the H direction is detected by the position detection subroutine program DS in step S35.

Next, in step S36, it is detected whether the correlation in all the H directions is completed (m = Nth-Hfh).
In step S36, if the correlation in the H direction is not yet completed (m = Nth−Nfh is not satisfied), the process proceeds to step S37, where m is incremented in the H direction.
Return to step S31. On the other hand, in step S36, H
When all the correlations in the directions have ended (m = Nth−Nf
In h), the position of the tracking subject is detected from the correlation values in both V and H directions, and the focus area 20 is reset (step S38). Then, in step S38, the focus area
The tracking area 19 is set such that 20 is the center (step S39), and based on the specific frequency information stored for each block, the data of the block corresponding to the focus area 20 reset in the above step S38. Is added, F
It is substituted into (1) (step S40).

In step S41, the specific frequency components F (0) and F (1) in the tracked focus area 20 are compared. In this step S41, F (1)> F
If it is (0), it means that the vehicle is climbing the curve shown in FIG.
The data of (1) is input to F (0). Then, in step S43, the focus state determination memory SS state is determined (SS = 2). Here, if SS = 2, the process proceeds to step S44, and the photographing optical system 11 is driven in the forward direction. Then, in step S45, the focus direction determination memory S
S is set to SS = 1, and then the process returns to step S8. On the other hand, if SS = 2 in step S43, the process proceeds to step S47 described later.

By the way, in the step S41, F
If (1) ≦ F (0), the process proceeds to step S46, and the determination of the focusing direction determination memory SS is performed. In this step S46, if SS = 1 or SS = 2 is not satisfied, it indicates that the curve goes down the curve shown in FIG. 17, and in step S47 the photographing optical system 11 is driven in the negative direction. After that, in step S48, the driving direction memory FS is set to FS = 1, and in step S49, the focus direction determination memory SS is set to SS.
After substituting = 2, the process returns to step S8.

In step S46, if the focusing direction judgment memory SS is SS = 1 or SS = 2, it indicates that the peak position of the curve shown in FIG. 17 has been exceeded, and the flow advances to step S50 to take an image. The optical system 11 is redriven to the in-focus position based on the value of the drive direction memory FS. That is, when FS = 1, the photographing optical system 11 is driven in the positive direction, and when FS = 0, the photographing optical system 11 is driven in the negative direction. Then, step S
At 51, the specific frequency component F (0) is input to F (2). Then, in step S52, the data obtained by adding the predetermined value ε to F (2) is substituted into F (0). Thereafter, in step S53, the interrupt determination is made, and if there is no interrupt (second release, etc.), the process returns to step S8,
When an interrupt occurs, the subroutine program DF is exited.

Next, the tracking correlation subroutine program DS will be described with reference to FIG.

The subroutine program DS is started, and in step A1, correlation calculation is performed at D0 (i) and D1 (i). Next, in step A2, the correlation value Sk1 is detected and stored. And step A2
When is completed, the process returns to the main routine.

FIG. 9 shows a state of signal processing for one line of the video signal in the tracking area 20. FIG. 3A shows a video signal itself, which is A / D converted for each pixel to create V and H projection signals. For j line, V
b (j) has a value proportional to this area. Further, FIG. 11B shows an output obtained by passing the signal shown in FIG. The BPF output is A / D-converted for each pixel, and the signals within the fixed intervals are added to form a block signal. The block signal is stored in the area not used for tracking correlation in the memory Vb. This sequence is sequentially performed for each line.

FIG. 10 is a diagram showing how to use the memory Vb in steps S11 and S26 of FIG. Since the video signal is obtained by sequentially scanning from the photoelectric conversion element 12, the signal added in the H direction is first input to the memory Vb, and the specific frequency component extracted for each block from the signal scanned in the H direction. Is input to the memory area that does not contain the signal added in the H direction. The data added in the H direction is subjected to the correlation calculation for each Nfv line, and only the data in the vicinity of the place where the correlation is high is stored. Further, as for the data of the specific frequency component obtained for each block, only the data in the vicinity of the high correlation are stored. The memory area in which the specific frequency corresponding to the line having a low correlation is stored is used as a memory area for new data.

In the initial state of FIG. 10A, the added value in the H direction is sequentially stored from Vb (1), and the frequency signal for each block is stored backward from the end address of the memory Vb. It FIG. 10B shows the S line in the case of the reading process in the middle (when the memory in which the addition signal of the read line is stored approaches the memory area storing the frequency component of each block when the S line is read). The added signal in the H direction is stored in Vb (s). Further, the frequency component (memory K) stored so far is moved to an area (memory J) where the correlation calculation has already been completed, and S
The frequency component for each block of the line is stored in the vacant address below it.

Further, the state of data input to the memory Vb is shown in FIGS. 11 (a) to 11 (f) and 12 (a) to 12.
It shows in (c). The 1H of the tracking area 19 shown in FIG. 5 is divided into n, and the frequency component corresponding to this n division is stored. The memory used for frequency storage is (Nfv + 1)
* For n addresses. FIG. 11A shows the case where j = Nfv in step S16 of the flowchart of FIG.
1 (b) and (c) are j = Nfv + 1 and j =
The time of Nfv + 2 is shown. When the tracking object exists near the center, the correlation is low, so the focus area block memory P appears in (b) and (c) of FIG. 11. That is, in the case of j = Nfv + 1, the number of n memories storing the frequency component of j = 1 becomes P, and the frequency component of j = Nfv + 2 is stored in the memory P.

FIG. 11D shows the time when the correlation becomes the lowest, and by this time, the immediately preceding unnecessary memory is treated as the memory P. In FIGS. 11E and 11F, since the correlation decreases again, the newly fetched frequency data is treated as the memory Q. As the correlation further progresses (reading progresses in the V direction), the memory for frequency storage and the memory for pixel addition intersect. Figure 1
2 (a), when the difference between the memory area for frequency storage and the address of the pixel addition memory is within a predetermined range, the memory area shifts as shown in FIG. 12 (b). To be done. In the same figure (b), of the addition data stored in the pixel addition memory, the pixel addition data with the highest correlation is moved to the upper memory, and after the memories of the other addition data are cleared, The frequency storage memory is moved to the memory. Further, in the next line, the processing is performed with the memory area moved (shown in (c) of FIG. 12).

Next, with reference to FIG. 13, the manner of correlation detection in the subroutine program DS of FIG. 8 will be described. H0 (i) and H1 (i) are shown in FIG. H
The shift δ1 between the images of 0 (i) and H1 (i) is determined as Ss = Σ | H0 (i) -H2 (is) | while fixing H0 (i) and shifting H1 (i). This is shown in FIG. 13 (b). The shift amount when the value of Ss is the smallest is detected.

FIG. 14 shows a display state of the display circuit 17. In the figure, an LED 172 for displaying the focus, front focus, and rear focus, and focus areas 173 and 174 (corresponding to the focus area 20 in FIG. 2) are displayed in the viewfinder 171. The position of the focus area 173 is the initializing position, and the position when the focus area is set during the tracking operation or at the arbitrary position is the focus area 174. During tracking operation or when the focus area 174 set at an arbitrary position is displayed,
The 173 display disappears, leaving only the 174 display.

Further, FIG. 15 shows a display system of the focus area and the like in FIG. Near the primary or secondary image plane,
An image is guided to the finder optical system 22 using the scattering type liquid crystal 21. The scattering type liquid crystal 21 has a focus area 1 according to the control signal.
The 73 or 174 part is colored. .

According to the above-described embodiment, since tracking and focusing point detection are performed by the same sensor, a signal without parallax can be obtained, and since the memory is also used for moving object tracking, cost and mounting are also advantageous. is there.

Further, since an object being tracked is always caught during tracking of a moving object and a specific frequency component is added by using the same video signal as the tracking, a focus detection device having a tracking function with high speed and high accuracy is provided.

In the embodiment, a CCD is used as the sensor, but a non-destructive and randomly accessible sensor, for example, a SIT (Static Induction Transistor) type solid-state image sensor, a CMD (Charge Modulation Device) type solid-state image sensor, or the like is used. By changing the integration time of the tracking operation and the in-focus point detection and reading the respective signals nondestructively,
The D conversion may be performed in time division.

Further, A / D conversion may be performed in a time division manner by using a delay line.

[0055]

As described above, according to the present invention, it is possible to perform moving object tracking and in-focus point detection with the same sensor, and the storage capacity is hardly increased, which is advantageous in terms of cost and mounting. Moreover, since the focus area is always detected in the tracking area after detecting the tracking area when tracking the moving body,
Focus detection can be performed accurately.

[Brief description of drawings]

FIG. 1 is a block diagram showing the concept in an embodiment of a moving object tracking device of the present invention.

FIG. 2 is a diagram showing a tracking area and a focus area.

FIG. 3 is a diagram showing data of specific frequencies in a one-dimensional tracking area.

FIG. 4 is a block diagram showing a configuration of a part for detecting a specific frequency of a focus detection circuit.

5 is a diagram showing a relationship between a memory configuration of the memory circuit of FIG. 4, a tracking area, and a focus area.

FIG. 6 is the first half of a flowchart for explaining the operation of focus detection.

FIG. 7 is the second half of the flowchart for explaining the operation of focus detection.

FIG. 8 is a diagram showing a subroutine program DS for tracking correlation.

FIG. 9 is a diagram showing a state of signal processing for one line of a video signal in a tracking area.

10 is a diagram showing how to use the memory Vb in step S11 and step S26 of FIG.

FIG. 11 is a diagram showing how data in a memory Vb is input.

FIG. 12 is a diagram showing how data in a memory Vb is input.

FIG. 13 is a diagram for explaining how correlation detection is performed by the subroutine program DS of FIG.

FIG. 14 is a diagram showing a display example of a display circuit.

15 is a diagram showing a display system of the focus area and the like in FIG.

FIG. 16 is a diagram illustrating an example in which a conventional focus detection system performs focus detection based on a differential value of a tracking signal while performing moving object tracking.

FIG. 17 is a diagram showing lens defocus and voltage values of frequency components.

FIG. 18 is a block diagram showing an example of a conventional focus detection system.

FIG. 19 is a diagram showing time, movement of a tracking object, and movement of a focus area.

[Explanation of symbols]

11 ... Shooting optical system, 12 ... Photoelectric conversion element, 13 ... Frequency detection circuit, 14 ... Drive circuit, 15 ... Tracking circuit, 16 ... Focus detection circuit, 17 ... Display circuit, 18 ... Area designation circuit, 19 ... Tracking area , 20 ... Focus area, 161 ... A / D conversion circuit, 1
62 ... Memory circuit, 163 ... Arithmetic processing circuit.

Claims (1)

[Claims] 1. An image pickup optical system, a photoelectric conversion element for converting a subject image through the image pickup optical system into an image signal, and frequency detection means for extracting a specific frequency component from an output signal of the photoelectric conversion element, A focus area setting means for setting the focus area, an addition means for calculating an addition value for each row of the photoelectric conversion elements in a tracking area wider than the focus area, and an addition value for a row including the focus area in each row. A first storage unit that stores the added value for each row, and a capacity that can store the added value of all rows in the tracking area. Second storage means for storing the specific frequency component, and tracking for performing correlation calculation based on the added values stored in the first and second storage means, respectively. Means a moving object tracking apparatus characterized by comprising the focus detection means for performing focus detection based on the specific frequency component stored in said second storage means for focus area that is tracked by the tracking means.