JP3382969B2 - Automatic focusing method - Google Patents

Description

Description: BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an automatic focusing system, and more particularly to an automatic focusing system capable of performing high-precision automatic focusing on any object. [0002] In recent years, automatic focusing (AF) based on a video signal obtained from an imager without using a special sensor,
Automatic exposure adjustment (AE), automatic white balance (AW
Techniques for performing B) and the like by digital processing have become common. The automatic focusing method digitally integrates a high-frequency component of a video signal in a focusing target area, and uses an obtained value as an evaluation value indicating a degree of focusing.
This is a method for driving and controlling the F motor. The extraction of the high frequency component of the video signal is performed using a band pass filter. As described above, the conventional automatic focusing control is performed based on an integrated value of a high frequency component of a video signal obtained from an imager. However, in such an automatic focusing method, since a high-frequency component of a video signal in a horizontal pixel direction is extracted, a high-level high-frequency component is obtained when a subject is a vertical line, but an oblique line is obtained. Sometimes, the level of the high-frequency component decreases, and in the case of a horizontal line, no high-frequency component is obtained, so that accurate focusing control cannot be performed. For example, when the subject is a vertical line A shown in FIG. 14, the video signal obtained from the imager has a sharp fall and a rise as shown by A1 in FIG. Also becomes a high level like A2. However, when the subject is oblique as shown in FIG.
The level change of the imager output does not become steep as shown by B1 in FIG. 15B, but changes gently. Therefore, the signal level after passing through the band-pass filter decreases, and the focusing accuracy decreases. . In the case of a horizontal line subject as shown in FIG. 14C, the imager output has only a DC component.
As shown in (C), no output appears in the bump pass filter. [0004] As described above, the conventional automatic focusing method uses the following method when the subject is close to an oblique line or a horizontal line.
High-precision focusing control is not possible. For horizontal line subjects,
This problem can be solved by forming a band-pass filter for extracting high-frequency components with a vertical filter, obtaining a steep level change like D1 in FIG. 14, and obtaining a high level like D2 as a band-pass filter output. , In that case 1H
The need for the delay means not only increases the circuit configuration scale, but also causes a problem that it cannot cope with a diagonally shaded subject. In addition, although a technique for rotating the entire optical system by an optimum angle with respect to oblique or horizontal lines has been proposed (Japanese Patent Publication No. 58-708), the mechanism for rotating the optical system is complicated.
There is also a problem that the size is increased. SUMMARY OF THE INVENTION It is an object of the present invention to provide an automatic focusing method which enables highly accurate automatic focusing of any object including a horizontal line and an oblique line. In order to solve the above-mentioned problems, an automatic focusing method according to the present invention uses a memory for storing an original image corresponding to an area to be focused, and stores the original image in a predetermined format. Address conversion means for rotating by a predetermined angle around the position of, a reading means for reading the original image by the address signal converted by the address conversion means, and a video signal read by the reading means. Interpolating means for performing an interpolation process on the image signal, a filter for extracting a high-frequency component of the video signal interpolated by the interpolating means, and an evaluating means for evaluating a focus state based on an output of the filter. Is done. According to the present invention, the address of the original image is converted.
The original image is read out by rotating it by a predetermined angle.
The rotation angle can be set freely,
Optimal focus signal can be obtained even for subjects with angles
it can. Also, interpolate the rotated and read signal
Therefore, high-precision AF information can be obtained,
The recording means can record the interpolated video signal.
it can. Next, an embodiment of the present invention will be described in detail with reference to the drawings. FIG. 1 is a block diagram showing one embodiment of the automatic focusing system according to the present invention. In this embodiment,
A tilted image signal of a subject image is electrically rotated to an optimal angle (vertical direction), so that high-frequency component extraction can always be performed with high accuracy. For example, the original images shown in FIG. 2A are rotated by an appropriate angle to obtain images (B) to (F) at desired angles. In (B), the original image (A) is rotated to obtain an original image of the subject image having an angle surrounded by a thick frame. At this time, although the shaded portion is a portion that is not in the original image, no problem arises because the subject image exists substantially in the center. In addition, the original image portion that protrudes from the thick frame is deleted. In FIG. 1, a subject image formed on an imager 2 such as a CCD via an optical system 1 is converted into an electric signal and input to an imaging process circuit 3. The imaging process circuit 3 performs well-known signal processing on the signal from the imager 2, and outputs a video signal to a recording system via the terminal 4 a of the switch 4. The luminance Y signal from the imaging process circuit 3 is converted into a digital signal by the A / D converter 5 and written into the field memory 6. Various synchronization signals are generated from the SSG circuit 10, and the trigger circuit 9 is driven to control the operation of the imager 2, to control the operation timing of the imaging process circuit 3, and to control the write timing of the field memory 6 by the write control circuit 1.
Control via 1. From the field memory 6, a video signal rotated by a predetermined angle as described later is read from the read control circuit 15 under the control of the address signal Add, and after an interpolation process is performed by the interpolation coefficient K, The signal is converted into an analog signal by the / A converter 8 and output to the recording system via the terminal 4b of the changeover switch 4. The signal interpolated by the interpolation processing circuit 7 is
The high-frequency component is extracted by the band-pass filter 13, and the integrated value obtained by the integrating circuit 14 is supplied to the microcomputer 12 as an evaluation value indicating the degree of focusing described above. The microcomputer 12 determines the AF motor driver 1 based on the evaluation value.
7 is driven to perform a focusing operation. The microcomputer 12 sends a control signal for rotating the subject image from the imager 2 by a predetermined angle to the read control circuit 15. The principle of address conversion for controlling the rotation of the subject image will be described with reference to FIG. FIG.
The address positional relationship when a thick line image is obtained by oblique scanning by rotating the original image indicated by the thin line by θ. In the figure, white circles indicate real pixels stored in the memory, and black circles indicate virtual pixels read from the memory. Each address position P
(00), P (10), P (20), P (01), P
(11), P (21), P (02), P (12), P
(22) Corresponding pixel data is written in the field memory 6, and using the pixel data at these address positions, the corresponding address position Q ( 10), Q (20), Q
(01), Q (11), Q (21),... Are obtained and sent to the field memory 6 as address signals Add. For example, the address position Q (1
The virtual pixel addresses 0), Q (20), Q (01), and Q (11) are obtained from the relationship shown in the drawing as follows. Q (10): x P (00) + cos θ y P (00) + sin θ Q (20): x P (00) +2 cos θ = P (10) +2 cos θ−1 y P (00) +2 sin θ = P (10 ) +2 sin θ Q (01): x P (00) −sin θ y P (00) + cos θ Q (11): x P (00) −sin θ + cos θ = P (01) −sin θ + cos θ y P (00) + cos θ + sin θ = P (01) + cos θ + sin θ−1 FIG. 4 shows an example of a circuit for generating the X address. In the XST register 151X, a pixel address to be read first, 0 in this example, is set.
XW = cos θ shown in FIG. 3 is generated from the register 152X, and XW shown in FIG.
0 = −sin θ has been generated. The output of the adder 154X is delayed by one clock (one pixel) by the delay unit 156X. The adder 154X adds cos θ from the XW register 152X and the output from the delay unit 156X.
The output of the delay unit 156X is added to the output (0 in this example) from the XST register 151X in the adder 158X. The delay unit 157X outputs the output of the adder 155X to 1
Delay by H. The adder 155X is connected to the X0 register 15
-Sin θ from 3X and the output from the delay unit 157X are added. Adder 159X adds the output of delay unit 157X and the output of adder 158X, and outputs the result as an X address signal. FIG. 5 shows an example of a circuit for generating a Y address signal similar to that of FIG. YST register 151Y
Is set to 0, and from the YW register 152Y, FIG.
Is generated, and Y0 register 153 is generated.
From Y, Y0 = cos θ shown in FIG. 3 is generated. The output of the adder 154Y is delayed by one clock (one pixel) by the delay unit 156Y. Adder 154Y adds sin θ from YW register 152Y and the output from delay unit 156Y. The output of the delay unit 156Y is YST
The output from the register 151Y (0 in this example) is added to the output by the adder 158Y. Delay unit 157Y delays the output of adder 155Y by 1H. Adder 155Y
Is cos θ from the Y0 register 153Y and the delay unit 1
The output from 57Y is added. Adder 159Y adds the output of delay unit 157Y and the output of adder 158Y, and outputs the result as a Y address signal. FIG. 6 is a diagram showing the principle of address conversion shown in FIG. 3 and the aspect ratio (768 pixels, 2 pixels) of 3 to 4 shown in FIG.
An address conversion diagram in the case where the present invention is applied to (40 lines) and rotated by 30 degrees is shown. In this case, as shown in FIG. 7, one pixel has a size of 2.4 to 1 in length and width. At this time, XST = 0 XW = 0.866 X
0 = −2.4 × 0.5 YST = 0 YW = 0.5 / 2.4 Y0 = 0.8
66, and as is apparent from the figure, the general formula representing the X address Xmn and the Y address Ymn in the number of pixels m and the number of lines n is as follows. Xmn = XST + m.XW + n.X0 Ymn = YST + m.YW + n.Y0 For example, the address (coordinate) of the 0th line (n = 0) is
(XY) = (0,0), (0.866, 0.208),
(1.732, 0.417), 1st line (n = 1)
Then, (XY) = (− 1.2, 0.866), (−0.
334, 1.074), (0.532, 1.28), ...
It becomes. Here, the integer part of each address is the address Add.
It is clear from the figure that the minority part indicates the interpolation coefficient K. Referring to FIG. 1, the present embodiment can cope with an electronic zoom operation when the zoom motor driver 16 is driven in response to the operation of the zoom switch 18.
FIG. 8 shows actual pixel addresses P (00), P (10), P (20), P (30), P indicated by white circles during the electronic zoom operation.
(01), P (11), P (21), P (31), and virtual pixel addresses Q (00), Q (10), Q after zooming
(20), Q (30), Q (40), Q (01), Q
(11), Q (21), Q (31), Q (41) are shown. At this time, XW and Y0 are constant values, and X0 and Y0
W is 0. The interpolation process for extracting higher-precision high-frequency components using the band-pass filter in FIG. 1 is preferably, for example, a four-point weighting method as shown in FIG. When X1 and X2 are determined as shown in the figure, the address position Q to be read from the memory is determined by four surrounding points P (11),
Using the weighted average of P (21), P (12), and P (22), it is determined by the following equation. Q = (1-Ky) X1 + Ky.X2 X1 = (1-Kx) P (11) + KxP (21) X2 = (1-Kx) P (12) + KxP (22) Therefore, Q = (1-Kx) ( 1−Ky) P (11) + Kx (1−Ky) P (21) + Ky (1−Kx) P (12) + Kx · Ky · P (22) (1) The operation of equation (1) is one cycle Within 4 pixel address P
(11), P (21), P (12), and P (22) can be realized by reading them simultaneously. Simultaneous reading of the four pixels can be performed using a memory configuration as shown in FIG. 10, for example. In the example shown in FIG. 10, even-numbered columns,
Even row dedicated memory (A), odd column, even row dedicated memory (B), even column, odd row dedicated memory (C) and odd column
Four independent memories of an odd-row dedicated memory (D) are provided. FIG. 11 shows an address generation circuit for reading data from a memory in order to perform an operation by the above-mentioned four-point weighted averaging circuit. An address, a column address for even-numbered memory, a row address for odd-numbered memory, and a row address for even-numbered memory are generated. Column address 0
The bit is output as a select signal HSEL, and is added to 1 to 9 bits by an adder 251. 1 to
9 bits become the column address for the odd-numbered column memory, and the adder 2
The output of 51 becomes the column address for the even-numbered column memory. Similarly, the 0 bit of the row address is output as the select signal VSEL, and is added by the adder 252 to 1 to 7 bits. Bits 1 to 7 become the row address for the odd-numbered memory, and the output of the adder 252 becomes the row address for the even-numbered memory. FIG. 12 shows an example of a circuit for performing the four-point weighted averaging operation shown in equation (1) using the read data read from the memory. In FIG. 12, selectors 253 and 254 are provided with select signal HS obtained in FIG.
When EL is “H”, the “H” terminal is selected, and when “L”, the “L” terminal is selected. Similarly, the corresponding terminal of the selector 261 is selected by the select signal VSEL. The selector 253 receives the input data of the even-numbered column and the even-numbered row, and the input data of the odd-numbered column and the even-numbered row.
Even-column odd-row read data and odd-column odd-row read data are input. The two outputs from the selector 253 are respectively subjected to coefficients (1-K
x), Kx. The outputs of the multipliers 255 and 256 are added by the adder 257 and output to the two input terminals (L, H) of the selector 261. On the other hand, two outputs from the selector 254 are multiplied by coefficients (1−Kx) and Kx by multipliers 258 and 259, respectively. Multiplier 258
And 259 are added by the adder 260 and output to the other two input terminals (L, H) of the selector 261. The two outputs from the selector 261 are the above X1 and X2, and the coefficients (1-K)
x), Kx. The outputs of the multipliers 262 and 263 are added by an adder 264 to obtain interpolated data Q. In the examples of FIGS. 11 and 12, the selection signal is required because, as shown in FIG. 13, addresses of four points to be selected are generated in accordance with four patterns # 1 to # 4. In this example, an example of pattern # 2 is shown. As described above, according to the automatic focusing method of the present invention, the address of the original image is converted to
Since the image is read by rotating it by a predetermined angle, the rotation angle
It can be set freely, and what angle
An optimum focus signal can be obtained for the object.
In addition, since the signal read by rotation is interpolated,
Accurate AF information can be obtained and recording means
Thus, the interpolated video signal can be recorded.

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a block diagram showing one embodiment of an automatic focusing system according to the present invention. FIG. 2 is a diagram for explaining the principle of the automatic focusing method according to the present invention. FIG. 3 is an address generation principle diagram showing an image rotation principle in the embodiment of the present invention. FIG. 4 is a circuit diagram for generating an X address based on the principle diagram shown in FIG. 3; FIG. 5 is a circuit diagram for generating a Y address based on the principle diagram shown in FIG. 3; FIG. 6 is a diagram showing an address generation principle when the principle shown in FIG. 3 is applied to actual image rotation. FIG. 7 is a basic image configuration diagram of the principle diagram shown in FIG. 6; FIG. 8 is a diagram illustrating a principle of an electronic zoom operation according to the embodiment of the present invention. FIG. 9 is a principle diagram for performing an interpolation process by a four-point weighted average calculation in the interpolation processing circuit 7 in the embodiment of the present invention. FIG. 10 is a memory configuration diagram used for performing the interpolation processing shown in FIG. 9; 11 is a circuit diagram illustrating an example of a memory read address generation circuit used in the interpolation processing illustrated in FIG. 9; FIG. 12 is a circuit diagram illustrating an example of the interpolation processing illustrated in FIG. 9; 13 is a diagram showing an example of an even / odd combination of four points selected in the interpolation processing shown in FIG. 9; FIG. 14 is a diagram showing vertical, horizontal, and oblique subject images. FIG. 15 is a diagram showing a change in a high-frequency component obtained from a band-pass filter in the conventional automatic focusing method for each subject shown in FIG. 14; [Description of Signs] 1 Optical system 2 Imager 3 Imaging process circuit 4 Changeover switch 5 A / D converter 6 Field memory 7 Interpolation processing circuit 8 D / A
Converter 9 Trigger circuit 10 SSG
Circuit 11 Write control circuit 12 Microcomputer 13 Band pass filter 14 Integration circuit 15 Read control circuit 16 Zoom motor driver 17 AF motor driver 18 Zoom switch

Claims (1)

(57) [Claim 1] A memory for storing an original image corresponding to a focusing target area, and an address for rotating the original image by a predetermined angle around a predetermined position. Conversion means; reading means for reading the original image based on the address signal converted by the address conversion means; interpolation means for performing an interpolation process on the video signal read by the reading means; interpolation by the interpolation means An automatic focusing method, comprising: a filter for extracting a high-frequency component of the obtained video signal; and evaluation means for evaluating a focus state based on an output of the filter.