JP2621285B2 - Focus control circuit - Google Patents

Description

Description: BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a contrast detection type focus control circuit used for auto focus of a video camera, for example.

SUMMARY OF THE INVENTION According to the present invention, there is provided a focus control circuit for detecting a spectral component in a video signal and performing focus control, wherein direction determining means for determining whether the movement of a subject is far or near is provided. By determining the moving direction of the focus lens from the result of the means, the focus lens can be moved in the far direction when the subject moves in the far direction, and the focus lens can be moved in the near direction when the subject moves in the near direction. egoism,
Further, the swing caused by the movement of the focus lens can be suppressed as much as possible.

[Related Art] The video camera's autofocus method includes a distance measurement method and a focus detection method when classified based on the principle of a focusing method. The distance measuring method measures the distance to a subject and controls the position of a lens in accordance with the distance. The focus detection method detects focus on an imaging surface and controls the position of a lens at a position where focus is achieved.

One of the focus detection methods is a contrast detection method. The contrast detection method performs focus control using the fact that the edge of a captured image becomes clearer as the lens approaches a focus position. The sharpening of the edge of the captured image corresponds to an increase in the high frequency component in the video signal.

A conventional example of a contrast detection type autofocus circuit is shown in FIG.

In FIG. 5, a focus lens 20 is displaced by a motor 21 and light from a subject is projected to an image sensor 22 via the focus lens 20. The image sensor 22 converts light from a subject into a video signal, and the video signal is passed through a signal processing system circuit 23 and an amplifier 24 to a band-pass filter (BP).
F) Output to 25. The band-pass filter 25 passes, for example, a component having a frequency of 200 KHz to 2 MHz.
The output of 25 is supplied to the detector 26. The detector 26 performs double-wave rectification on the input signal, and the output of the detector 26 is supplied to the A / D converter 27. The A / D converter 27 A / D converts only the signal corresponding to the focus area based on the focus area setting signal sent from the microcomputer 29, and supplies this digital signal to the integration circuit. The integration circuit 28 integrates the digital signal of the A / D converter 27, and the output of the integration circuit 28 is supplied to the microcomputer 29 as a focus evaluation value. And
The microcomputer 29 searches for the position of the focus lens 20 at which the output of the integration circuit 28 is maximized by, for example, a hill-climbing control algorithm as shown in FIG. That is, the microcomputer
A motor drive command is given from 29 to the motor drive circuit 30, and the position of the focus lens 20 is controlled so that the focus evaluation value output from the integration circuit 28 is maximized.

[Problem to be Solved by the Invention] However, if the subject moves in a direction away from the camera (hereinafter, referred to as “far direction”) or in a direction approaching (hereinafter, referred to as “near direction”) after focusing, the hill climbing control is performed again. The focusing operation is performed again by the algorithm. In this focusing operation, since the moving direction of the subject is unknown, the focus lens is temporarily moved in the near direction (or far direction), and the focus evaluation values are compared. Or, if the evaluation value is a small value, it is controlled to move in the opposite far (or near) direction. Therefore, when the moving direction of the subject and the moving direction of the focus lens do not coincide with each other, there is a drawback that the followability of the screen is poor, and that since the focus lens moves a lot, the screen swings violently and is unsightly.

SUMMARY OF THE INVENTION It is an object of the present invention to provide a focus control circuit that has good followability by recognizing the moving direction of a subject and that minimizes screen swing.

[Means for Solving the Problems] To achieve the above object, a focus control circuit according to the present invention comprises: an imaging unit for converting an optical signal from a subject into a video signal; A focus control circuit for detecting a spectral component amount and calculating a focus position, wherein a detection area is provided in the image frame, and a predetermined time interval in the detection area is set. A vector calculating means for comparing data based on the two separated video signals to calculate a movement vector of the subject in the focus area; and moving the subject in a far or near direction based on the movement vector calculated by the vector calculation means. Direction discriminating means for discriminating the direction of movement of the focus lens based on the result of the direction discriminating means. The direction is controlled.

[Operation] The data in the current detection area by the vector calculation means is compared with the data before it, a movement vector of the subject in the focus area is calculated, and the direction of the movement of the subject is determined by the direction determination means from the calculation result. Then, the focus lens is moved in the far direction if it is in the far direction, and the focus lens is moved in the near direction if it is in the near direction.

Embodiment An embodiment of the present invention will be described below with reference to the drawings.

FIGS. 1 to 4D show an embodiment of the present invention. In this embodiment, a case where the focus control circuit of the present invention is applied to a video camera is shown.

In FIG. 1, the position of a focus lens 2 is changed by a motor 1, and an image of a subject a is formed on an image sensor 3 via the focus lens 2. The image sensor 3 converts light from the subject a into a video signal, and the output of the image sensor 3 is supplied to a signal processing circuit 4. The signal processing system circuit 4 creates a luminance signal by adding, for example, each of R, G, and B signals at a predetermined ratio, and sends this signal to the band-pass filter 6 and the A / D converter 7 via the amplifier 5. Can be The band-pass filter 6 serving as a spectrum detecting means is a filter that passes a component having a frequency of, for example, 200 KHz to 2 MHz. The band-pass filter 6 extracts all the spectral components other than the DC component and the noise component in the luminance signal. This bandpass filter 6
Is supplied to a detector 8, which rectifies the output of the band-pass filter 6 in both waves. Double-wave rectified detector 8
Is supplied to an A / D converter 9 and digitized by a sampling clock, and this digital signal is output to an integrating circuit 10. For example, a color subcarrier having a frequency fsc is used as the sampling clock. The integrator 10 integrates data as a digital signal in the focus area A set by the focus area control circuit 15, and the integrated data is sent to the digital controller 13. The digital controller 13 is a focus lens 2 in which the output of the integration circuit 10 is maximized by a hill-climbing control algorithm.
A motor drive command is issued to the motor drive circuit 16 in order to search for the position. When a motor drive command is given, the motor drive circuit 16 outputs a motor drive signal to the motor 1 so that the motor 1 is driven. The rotation of the motor 1 is picked up by a frequency generator (FG) 17 and this frequency generator (F)
G) The output of 17 is sent to the digital controller 13 for feedback control.

On the other hand, the A / D converter 7 digitizes the luminance signal from the amplifier 5 by sampling lock and sends the digitized luminance signal to the first memory 11. First
The data captured in the memory 11 is transferred to the second memory 12, and the storage capacity of the first memory 11 and the second memory 12 has a capacity capable of storing the data of the detection area B. The first memory 11 and the second memory 12 are controlled by the digital controller 13 to write and read the data in the detection area B into the first memory 11 at intervals of several frames. The data captured in the memory 11 is transferred to the second memory 12. By repeating this sequentially, the previous data B (x, y) is stored in the second memory 12 in the first memory.
The following data B '(x', y ') is fetched into the memory 11 of FIG. As shown in FIG. 2, the detection area B includes the focus area A within the image frame C and is set to be larger than the focus area A. First memory 11 and second memory
The output of the memory 12 is output to the vector operation circuit 14 in synchronization with the data of the corresponding address. The vector operation circuit 14 calculates the movement vector S of the subject a based on the data sent from the first memory 11 and the second memory 12.
Is calculated. There are various methods for calculating the movement vector S. Here, the movement vector S is obtained by a method using a cross-correlation function.

An image having a value of f ₁ (x, y) at time t ₁ is calculated at time t _{1
It is} assumed that f ₂ (x, y) ≒ f ₁ (x−ξ ₂ , y−η ₂ ) in ₂ . Then, when the cross-correlation function of φ (ξ, η) = ∫f ₁ (x, y) f ₂ (X + ξ, y + η) dxdy is obtained, φ (ξ, η) is given by ξ = ξ ₂ , η = η ₂ Take the maximum value. Therefore, it is possible to obtain the movement amount (ξ ₂ , η ₂ ) or the speed from this. Hereinafter, a specific description will be given.

Now, as shown in FIG. 3, in the image having a value of g (x, y) at time t = 0, the moving object moves at a speed (α, β) in the image, and at time t = τ. g (x-ατ, y-β
τ).

 However, it is assumed that τ is sufficiently small.

Here, when g (x-ατ, y-βτ) is Taylor-expanded and only the first-order term is adopted, Becomes Here, if the difference between the images at the time of t = τ and t = 0 is expressed as d (x, y) = g (x−ατ, y−βτ) −g (x, y), the expression (1) becomes Becomes

this is, That is, V · grad g (x, y) = − d (x, y) / τ where • is a scalar product, and i and j can be written as unit vectors in the x and y directions, respectively.

(V + ΔV) [grad g (x, y) + Δg] = − d (x, y) / τ. This is arranged and described as -V · grad g (x, y) = − d (x, y) / τ + e (x, y).

 e (x, y) is the error.

Here, ∂f / ∂x = a, ∂f / ∂y = b, −d (x, y) / τ
= D, and the value of each point is distinguished by a suffix i, then a _i α + b _i β = d _i = e _i . As a method for obtaining α and β, the sum of squares of the error Σe _i
If you take the method of minimizing ² , Is the condition required. From this, as the velocity vector V, Get. Then, the movement vector S can be calculated from V · τ.

The vector operation circuit 14 sends the data of the movement vector S to a direction discriminating circuit 18 which is a direction discriminating means. The direction determination circuit 18 determines whether the movement of the subject a is “far” or “near” based on the movement vector S of each part of the subject a. As shown in FIG. 4 (a), when the subject a moves in the far direction, the subject a in the detection area B becomes small and the movement vectors (S ₁ , S ₂ ,. The direction of ()) is directed toward the center. When all the x components of the movement vector (S ₁ , S ₂ ,...) Of each cylinder location are added, the characteristic diagram becomes as shown in FIG. 4 (b).

Conversely, as shown in FIG. 4C, when the subject a moves in the near direction, the subject a in the detection area B becomes large, and the movement vector (S ₁ , S ₂ ,...) The direction is radial. Then, the movement vector (S ₁ , S ₂ ,
..) are all added, the characteristic diagram becomes as shown in FIG. 4 (d). The coordinates are set as shown in FIGS. 4 (a) and 4 (c), and FIG.
When comparing FIG. 4 (d), in the case of FIG. 4 (b), as the value of x increases, the integral value shifts from + → 0 → −, and in the case of FIG. It is displaced from 0 to + and the relationship is reversed.

That is, when the subject a moves in the far direction, the characteristic diagram of the movement vector S becomes as shown in FIG.
When the subject a moves in the near direction, the movement vector S
FIG. 4 (d) shows a characteristic diagram of the above-mentioned case, and it is determined whether the movement of the subject a is in the far direction or the near direction. This determination result is output to the digital controller 13.

When the direction determination circuit 18 outputs a signal indicating the far direction, the digital controller 13
When the direction discriminating circuit 18 outputs a signal indicating the near direction, a motor driving command is issued to the motor drive circuit 16 to move the focus lens 2 in the near direction. Issue a drive command.

Then, the digital controller 13 searches for the position of the focus lens 2 at which the output of the integrating circuit 10 becomes maximum by the hill-climbing control algorithm as described above.

 Hereinafter, the operation of the above configuration will be described.

As shown in FIG. 2, when the target subject a is captured in the focus area A of the image frame C, the image sensor 3 converts the light from the subject a into a video signal. This video signal is converted into predetermined data via a band pass filter 6, a detector 8 and an A / D converter 9, and
Are integrated by the integration circuit 10. The integrated data is sent to the digital controller 13, and the digital controller 13 controls the position of the focus lens 2 by using the hill-climbing control algorithm so that the integrated value becomes maximum.

Now, when the subject a moves away from the camera (far direction), the video signal data in this state is loaded into the first memory 11. At the same time, the video signal data (several frames before) taken in the first memory 11 is transferred to the second memory 12, and each data of the first memory 11 and the second memory 12 is transferred to the vector operation circuit. Sent to 14.
Then, as described above, the subject a
Is calculated, and the direction discriminating circuit 18 integrates the x component of the motion vector S based on the calculation result, and it is determined from the characteristic that the movement of the subject a is in the far direction. The digital controller 13 issues a motor drive command to the motor drive circuit 16 to move the focus lens 2 in the far direction. Then, the focus evaluation value always sent from the integration circuit 10 gradually shows a large value, and the position of the focus lens 2 at which the focus evaluation value becomes the maximum is immediately searched.

In this embodiment, two memories 11 and 12 are used to simultaneously obtain data based on video signals separated by a predetermined time interval in the detection area B. However, one memory may be used. May be obtained. In this embodiment, the movement of the subject a is calculated using the video signal. However, the movement of the subject a can be calculated using the output of the detector 8.

[Effects of the Invention] As described above, according to the present invention, in a focus control circuit that performs focus control by detecting a spectral component in a video signal, direction discrimination for discriminating whether the movement of a subject is far or near. Since the means is provided, when the subject moves in the far direction (or the near direction), the focus lens can be first moved in the far direction (or the near direction), so that the screen can follow the screen well. There is an effect that the swing is suppressed as much as possible.

[Brief description of the drawings]

FIGS. 1 to 4D show an embodiment of the present invention.
FIG. 2 is a block diagram of a focus control circuit, FIG. 2 is a diagram showing a screen of an image frame, FIG. 3 is a diagram for explaining calculation of a movement vector, and FIG. FIG. 4 (b) is a diagram showing a movement vector, FIG. 4 (b) is a characteristic diagram when the x component of the movement vector is added, and FIG. 4 (c) is a diagram showing a movement vector when the subject moves in the near direction.
FIG. 4D is a characteristic diagram when the x component of the movement vector is added, FIG. 5 is a block diagram of a conventional focus control circuit, and FIG. 6 is a characteristic diagram of the integral data. is there. 3... Imaging element (imaging means), 6... Band-pass filter (spectrum detecting means), 14... Vector computing circuit (vector computing means), 18... Direction discriminating circuit (direction discriminating means), a. , A ... Focus area, B ...
... Detection area, C ... Image frame.

Claims (1)

(57) [Claims] 1. An image pickup means for converting an optical signal from a subject into a video signal, and a spectrum detection means for detecting a spectrum component in the video signal in a focus area of an image frame, wherein a spectrum component amount is detected. A focus control circuit that calculates a focus position by providing a detection area in the image frame; comparing data based on two video signals separated by a predetermined time interval in the detection area; And a direction determining means for determining whether the movement of the subject is far or near based on the movement vector calculated by the vector calculating means. A focus control circuit, wherein a movement direction of a focus lens is controlled.