JPH0353676A - Focus detector - Google Patents

Abstract

PURPOSE:To detect a focal point independently of a pattern of an object by extracting only a video signal of a picture element having an angle with respect to a vertical and a horizontal direction from video signals as a focal point detection signal. CONSTITUTION:An object image formed onto an image pickup element 2 by an image forming optical system 1 is photoelectric-converted into a video signal. Then only a signal obtained from a picture element of a direction having an angle with respect to a vertical and a horizontal direction is used as a focal point detection signal and a signal obtained from other picture elements is used as an output video signal at the detection of a focal point thereby branching the video signal outputted from the image pickup element 2 respectively. The focal point detection signal is sent to a focal point detection block 3, from which a focal point evaluation signal is outputted to a control block 6. Since a picture element having an angle with respect to a vertical and a horizontal direction is used as the focal point detection signal in this way, the focal point is detected independently of the pattern of the object.

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention is a focus detection device, more specifically, a focus detection device, which detects a focused point by extracting a specific frequency component in a video signal output from an image sensor irradiated with object light. The present invention relates to a focus detection device.

[Prior Art] Conventionally, a focus detection device that drives a part of an imaging system, extracts a specific frequency component from a video signal of an image sensor, and detects a focused point is described, for example, in NHK Technical Report, Vol. 17, No. 1. It is disclosed in "Automatic focus adjustment of television camera using mountain-climbing servo system" published by Ishida et al. This "automatic focus adjustment of TV cameras using the hill-climbing servo method" detects the definition of the screen based on the Takagi component in the video signal, performs hill-climbing control to maximize the definition, and automatically adjusts the focus of the TV camera. This paper describes the principles and circuit design conditions of a device that performs optical focus adjustment. Based on measurements of the response of an experimental circuit to a two-dimensional object and simulation experiments using an analog computer, we found that the shape of the hill-climbing curve, which changes depending on the lens aperture and focal length, determines the accuracy of focusing using the hill-climbing servo system. Accuracy is improved by increasing the gain of the control system and decreasing the time constant of the time delay element, and the time constants of the detection and hold circuits in the control system are reduced to 0. It is disclosed that if Ol seconds is selected, a focusing accuracy of 5% can be obtained for a fourfold change in the height and slope of the climbing curve, respectively. Furthermore, it is disclosed that by using adaptive control that detects the shape of the hill-climbing curve and controls the gain of the control system, it is possible to expand the control range and improve accuracy.

However, with the above-mentioned "TV camera automatic focusing using the mountain-climbing servo method: A adjustment," which detects the focused point by evaluating the power of specific frequency components in the video signal, the focused point may not be detected at all depending on the pattern of the subject. .

For example, in the current NTS, C color video signal recording system, signals are read out in the horizontal direction, so when trying to focus on a subject with horizontal stripes, the illuminance distribution is inconsistent with the contrast of the subject. The focus becomes clear and the contrast deteriorates, resulting in poor focusing accuracy. Therefore, as a solution to this problem, horizontal lines can be stored in multiple delay lines and adjacent vertical pixels can be compared and evaluated, or all signals can be stored once in memory and digitally processed. Japanese Patent Laid-Open No. 53-97826 discloses means for detecting focus without focusing on the subject.

[Problems to be Solved by the Invention] However, by accumulating horizontal lines into multiple delay lines and comparing and evaluating adjacent vertical pixels, or by accumulating all signals once in memory and digitally processing them, it is possible to The proposal described in Japanese Patent Application Laid-Open No. 53-97826, which performs focus detection regardless of the pattern, requires multiple delay lines or dedicated memory, making it expensive and requiring a processing circuit. It also had many disadvantages, such as being complicated.

SUMMARY OF THE INVENTION An object of the present invention is to solve the above-mentioned problems and to detect the pattern of an object using a low-cost and simple processing circuit when extracting a specific frequency component from a video signal of an image sensor and detecting a focused point. An object of the present invention is to provide a focus detection device that enables in-focus point detection without being influenced by.

[Means and effects for solving the problems] The focus detection device of the present invention includes an imaging optical system that forms an image of a subject, and an imaging element that photoelectrically converts the image formed by this imaging optical system and outputs a video signal. and a video signal extracting means for extracting only a video signal having a certain angle vertically and horizontally from among the video signals;
The present invention is characterized by comprising a detection means for detecting the focal state of the imaging optical system based on the specific frequency component level of the extracted video signal.

[Example〕 The present invention will be explained below with reference to illustrated embodiments.

First, before describing embodiments of the present invention, the basic principle of the present invention will be explained with reference to FIGS. 2 to 4.

In general, the optical image of a subject produced by an optical system is
The sharpness changes depending on the incorrect match, and when the sharpness is high in the correct match, it shows a clear illuminance distribution according to the contrast of the subject, but when the incorrect match is in a so-called blurred state and the sharpness is low, The illuminance distribution becomes unclear compared to the contrast of the subject, and the contrast deteriorates. Therefore, the image signal obtained by scanning the focus area set approximately in the center of the viewfinder screen and photoelectrically converting it changes depending on the sharpness, and when the sharpness is high, the level changes large and sharply. However, the lower the sharpness, the smaller the amount of fluctuation in the level, and the more gradual the change.

Figure 2 (A) is a diagram showing the frequency characteristics of the video signal obtained by scanning the focus area and photoelectrically converting it. In contrast, the video signal obtained from the out-of-focus subject image contains relatively few high frequency components, as shown by the dotted line fI2 in the figure. Therefore, when the power of a specific frequency (usually a high frequency component) in a video signal is plotted against the defocus position of the focusing lens, as shown in Figure 2 (B), when the sharpness is high in the correct position, A comparatively sharp characteristic diagram as shown by the solid line p3 is obtained, but if the sharpness is low, the diagram becomes a diagram that changes broadly as shown by the dotted line g4. In other words, automatic focus adjustment is possible by detecting when the output level is at its maximum or when the level fluctuation is at its steepest.

In this case, if you try to focus on a horizontally striped object while reading out the video signal in the horizontal direction as explained in the conventional example, the specific frequency power distribution curve like the dotted line fI4 in FIG. 2(B) will appear. Since the in-focus point is determined, the focusing accuracy becomes poor.

Therefore, in the present invention, each pixel in the focus area is scanned in a staggered manner, that is, at a certain angle in the horizontal and vertical directions, as shown in FIG. It is said that By extracting a specific frequency component from this in-focus point detection signal to obtain a focus evaluation value, the in-focus point can be detected with high precision even if the subject image has a horizontal striped pattern. can be done. On the other hand, the pixels not shaded in FIG. 3 are horizontally scanned, and the video signals output from these pixels are used as output video signals.

FIG. 4 is a block system diagram illustrating the basic principle of the focus detection device of the present invention. In the figure, a subject image focused on an image sensor 2 by an imaging optical system 1 is photoelectrically converted into a video signal by the image sensor 2.

When detecting a focused point, only signals obtained from pixels in directions having a certain angle with the vertical and horizontal directions are used as focused point detection signals, and signals obtained from other pixels are used as output video signals to capture images. The video signals output from the elements 2 are each branched.

The output video signal is sent to the video signal output block 5, and outputted from the video signal output block 5 to an external system. On the other hand, the in-focus point detection signal is transmitted to the in-focus point detection block 3.
The in-focus point detection block 3 outputs the in-focus point evaluation signal to the control block 6. This control block 6 includes a monitor block 4, a focused point detection block 3. Based on information from the imaging optical system 1, the imaging optical system 1, the image sensor 2, the focused point detection block 3, the monitor block 4, and the video signal output block 5 are controlled, respectively.

In such a sequence, by separating the video signal output from the image sensor 2 into the in-focus point detection signal and the output video signal, the influence of the video signal on the external system during in-focus point detection can be reduced, and Since the in-focus point detection signal is a signal from a pixel having a certain vertical and horizontal angle as described above, in-focus point detection is possible without being influenced by the pattern of the subject.

In other words, the focus detection device of the present invention detects the focus area with a monitor, extracts the video signal of the focus area with a specific directionality, and uses the power of the extracted signal in a specific frequency band as an evaluation value to determine the focused point. It is designed to detect.

Next, a focus detection device showing one embodiment of the present invention will be described. FIG. 5 is a block system diagram of the above embodiment. In the figure, a photographing lens 101 and a lens drive circuit 103 connect the coupling optical system 1 shown in FIG. , the focus evaluation circuit 104 detects the in-focus point detection block 3, and the CPU
107 constitutes the control block 6, respectively. The CPU 107 controls the lens drive circuit 10 based on information such as the aperture, subject distance, and lens focal length from the photographic lens 101 and signals from the focus evaluation circuit 104.
The photographing lens 101 is driven through the lens 3.

FIG. 6 is a block diagram showing the structure of the monitor circuit 105. In the figure, the monitor circuit 105 is composed of an amplifier 151 for signal amplification and a latch circuit 152 that latches with a timing signal output from the CPU 107.

FIG. 7 is a block diagram showing the structure of the focus evaluation circuit 104. In the figure, the focus evaluation circuit 104 includes a bandpass filter (hereinafter abbreviated as B.P.F.) 1
61. Detection circuit 162. It has a plurality of blocks each including an integrating circuit 163 and an A/D converting circuit 164.

FIG. 8 is a block system diagram showing the configuration of one line of the CCD 102. In the figure, the CCD 102 includes a light receiving element 114, a signal CCD 115 for signal transfer, and an electrode (hereinafter abbreviated as transfer electrode) 119 for applying a transfer clock for transferring the signal CCD 115.
A reset CCD 113 for signal reset, an electrode (hereinafter abbreviated as reset electrode) 118 for applying a reset transfer clock for transferring the signal of this reset CCD 113, and a signal of the signal CCD 115 that is horizontally transferred as an output video signal. A video signal line CCD116 that transfers the signal of the signal CCD115 in the horizontal direction as a focused point detection signal, an AF signal line CCD117 that transfers the signal in the horizontal direction as a focused point detection signal, and a video signal line CCD116 and an AF signal line CCD117 that transfer the signal of the signal CCD115.
An electrode (hereinafter referred to as
(abbreviated as switching electrode) 120, and reset C C D
113 signal to monitor line 111 and reset line]
Switches 121 and 122 that branch into 12 and 12, respectively;
A monitor address circuit 123 that controls on/off of these switches 121 and 122, and a video signal line CCD 11
．． Output circuit 124 that cuts noise and amplifies the signal of No. 6
, an output circuit 125 that noise-cuts and amplifies the signal of the AF signal line CCD 117 and outputs the signal, and an AF address circuit 1-26 that controls the AF switching clock.

In the focus detection device of this embodiment configured in this way, subject light passes through the photographic lens 101 (see FIG. 5) and is
When the image is formed on the light receiving element 114 of DI○2, the same element 1
14, the signal is photoelectrically converted and transferred onto the signal CCD 115.

At times other than focus detection, the signal transferred onto the signal CCD 115 is transferred to the video signal line CCD 116 as an output video signal by the control signal of the AF address circuit 1.26. Furthermore, the same C C D 1. 1 6 in the horizontal direction and is transmitted through the output circuit 124 to the signal output circuit 106 (
The output is shown in Figure 5 (see Figure 5).

After being processed into a video signal by the output circuit 106, it is output to the outside by a control signal applied from the CPU 107 (see FIG. 5).

On the other hand, during focus detection, a signal extracted with a specific direction from the video signal transferred onto the signal CCD 115 is transferred to the AF signal line CCD 117 as a focused point detection signal by the control signal of the AF address circuit 126. be done. Then, it is further transferred in the horizontal direction on the same CCDI 17, and is transmitted through the output circuit 125 to the focus evaluation circuit 104 (
(see Figure 5).

The focus evaluation circuit 104 extracts only a specific frequency component from the manually input in-focus point detection signal. P. F161(
(see FIG. 7), detected by a detection circuit 162, and integrated by an integration circuit 163. And A/D conversion circuit 16
4, it is converted into a digital signal and output to the CPU 107.

Here, the in-focus point detection signal transferred to the AF signal line CCD 117 will be explained with reference to FIGS. 1(A) and 1(B). In FIG. 1(A), the row numbers and column numbers in the light receiving element 114 are i--m, . . .-1, respectively. 1. , 2.
3,...mJ""nt......-1.
i-1 in an area composed of 1, 2, 3,...n pixels. 2. 3 j-1.2, . After the signals of each pixel in the i-1 row of the light receiving element 1]4 are shifted to the signal CCD 115, in response to the transfer clock on the transfer electrode 119, as shown in FIG. 1(B), Video signal line C C D 1 1
．． Transferred onto 6.

Then, in FIG. 1(A), the ci. j) - Only the charges accumulated on the pixels (1.3) and (1.7) are shifted from the video signal line CCD116 to the AF signal line CCD117 in response to the AF switching clock on the switching electrode 120, The charges accumulated in the remaining pixels are transferred to the video signal line CCD116.
is transferred in the horizontal direction and output to the signal output circuit 106 as an output video signal.

Similarly, row number i-2 has j-2.4, 6, 8, and row number i-3 has j-1.5. Only the charges accumulated in each pixel of the AF signal line CCD
I17, and the same CCD1 at a certain frequency which will be described later.
17 and is outputted to the focus evaluation circuit 104 as an in-focus point detection signal via the output circuit 125. The frequency of the transfer clock pulse that transfers the in-focus point detection signal on the AF signal line CCD117 is determined by the CPU.
107, the frequency when in focus is set to be the same as or lower than when out of focus. Next, the reason for this will be explained.

That is, when the object is out of focus, the low frequency component of the object image passing through the photographic lens is larger than the high frequency component, but when the object is in focus, the high frequency component also becomes larger. Therefore, it is easier to detect the focus state using a signal on the high frequency side when the object is in focus than when it is out of focus. In this case, bandpass filters may be prepared for each band, but the AF signal line C
If the transfer clock of the CD 117 is set to a lower frequency when in focus than when out of focus, the high frequency components of the subject image are shifted to the low frequency side, so that they can be detected with the same bandpass filter.

In this way, after scanning all the pixels in each row of i1m SJm and reading out the accumulated charges, there is still A
If it is in F, the focus area i on the light receiving element 114 is 1. 2. 3 j-1.2, ・・・・・・・・・The pixel from which the in-focus point detection signal is extracted from 8 is, for example, a pixel notched with a diagonal line downward to the right and shifted to the right by one pixel. In other words, the charges accumulated in each pixel are j-4.8 in row i-1, j-1.3, 5, 7 in row i-2, and j-2.6 in row i-3. AF signal line CCD11
Shift to 7. In other words, the AF signal line CCD117
The signal to be shifted is shifted by one pixel every frame. In other words, the in-focus point is being detected. The output video signal is composed of the remaining signal after extracting the in-focus point detection signal, but since the in-focus point detection signal is a signal from a different pixel for each frame, the output video signal is sent to an external system even when the in-focus point is detected. There is practically no possibility that signal quality will deteriorate. After AF is completed, the signals accumulated in all pixels are transferred to the video signal line CC.
Shift only to D116, AF signal line CCD117
There is no output at all.

Further, the integration time of the subject light irradiated onto the light receiving element 114 of the CCD 102 is determined by the reset CCD 113 and the switch 12.
1. Monitor circuit 1 using 122 and monitor line 111
The decision is made by sending a signal to 05.

The process of this CCD integration time control will be explained with reference to the flowchart of FIG.

First, when this flow starts in step S201, the process proceeds to step S202, in which the charge accumulated in each pixel of the light receiving element 114 is shifted to the reset CCD 113 at 1-0, and the irradiation of the subject light to the light receiving element 114 is stopped. Start (t-0)L, Proceed to step S203. In this step S203, the switch 122 formed of, for example, MOS is turned on and the switch 121 is turned off, that is, the reset COD 113 is connected to the reset line 112, and the process proceeds to step S204. In this step S204, all signal charges on the reset COD 113 are transferred at high speed to the reset line 112, and the process proceeds to step S205. Now, the time t calculated in advance by the AE sensor (not shown)
1, in this step S205, the charge of the light receiving element 114 is shifted to the reset CCD 113 after the above-mentioned time t1 has elapsed starting from 1-0 in the above step S202. That is, the irradiation of the subject light onto the light receiving element 114 starts at t-0 in step S202 and ends at 1-1 in step S205, thereby emitting a photocharge corresponding to the amount of irradiation light during the integral time t1. is shifted from the light receiving element 114 to the reset CCDll3, and the process proceeds to steps S206 and S207. This step S
206, S207, the above step 5203. S20
The same process as in step 4 is performed and the process advances to step S208.

In step S3208, it is determined whether the high-speed transfer has ended, and if it has not ended, the process advances to step S209, and it is determined whether the area corresponding to the focus area has been reached. When the area corresponding to the focus area is reached, the process proceeds to step S210, where the switch 121 is turned on and the switch 122 is turned off, that is, the reset CCD1
13 from the reset line 112 to the monitor line 111
After changing the connection, the process returns to step S207.

That is, focus area i-1. 2. 3 j-1.2. Since steps 8206 to S209 are repeatedly executed in the area excluding . . Also, in the focus area,
Since steps 5207 to S210 are repeatedly executed, photocharges corresponding to the amount of irradiation light during the integration time t1 are detected by the monitor circuit 105 via the monitor line lit. In other words, the monitor level, which is the integral value of the subject light irradiated to all pixels within the focus area during the integration time t1, has been detected.

If the transfer is completed in step S208, step S2
Proceeding to step 11, it is determined whether the monitor level is OK. If the monitor level is insufficient, proceed to step S212,
A time t set in advance by calculation using an AE sensor (not shown)
After replacing t2l with a new time t2l, the process returns to step S202 and the above operation is repeated again. Further, if the monitor level in step S211 is OK, it means that the automatic exposure setting has been completed, so step S202. S203. S
Step S213 similar to step S204. S214.5215
After executing, the process advances to step S216. That is, in these steps 8213 to S216, the above step S211
After rapidly discharging the photocharges accumulated on the light receiving element 114 in the automatic exposure setting routine up to the reset line 112, the integration time t1. Integration time t predicted from t2
3, the light receiving element 114 is irradiated with object light to perform main exposure. Then, the signal charges accumulated on the light receiving element 114 are shifted to the signal CCDI 15, and are used as the in-focus point detection signal and the output video signal on the AF signal line CCD 117 and the video signal line CCD 116, respectively.
It is designed to be transferred to

In the description of the above embodiments, a means for detecting a focal point using a CCD without being affected by the pattern of the subject was described.
Oxiclde Ses1conduator),
AM 1 (Agpllflded MOS Intel
ligent Isager). S I T (St
atie Induct1onTrans1stor
), etc., and it goes without saying that the same effect can be obtained by cutting out a part for the AF signal in the readout part of the video signal.

[Effects of the Invention] As described above, according to the present invention, by extracting only the video signals of pixels having a certain angle vertically and horizontally from the video signal, it is possible to detect a focused point regardless of the pattern of the subject. This makes it possible to achieve excellent effects such as not affecting the video signal itself.

[Brief explanation of drawings]

FIG. 1(A) is a layout diagram of a light receiving element of a CCD in a focus detection device showing an embodiment of the present invention, and FIG. 1(B) is a layout diagram of a video signal line CCD and an AF signal line CCD, respectively. Figure M42 (^) is the frequency characteristic of the video signal, Figure 2 (B) is a characteristic diagram showing the distribution of the power of a specific frequency in the video signal with respect to the lens defocus position, and Figure 3 is the
4 is a block system diagram showing the basic principle of the present invention; FIG. 5 is a block system diagram of a focus detection device showing an embodiment of the present invention; Figure 6 is a block system diagram showing details of the monitor circuit in Figure 5 above; Figure 7 is a block system diagram showing details of the focus evaluation circuit in Figure 5 above; FIG. 9 is a block diagram showing the configuration of one line of CCD in FIG. 9. FIG. 9 is a flowchart of CCDvt minute time control in an embodiment of the present invention. 1... Imaging optical system 2... Imaging element (imaging element and video signal extraction means)

Claims (1)

[Claims] (1) An imaging optical system that forms a subject image, an imaging device that photoelectrically converts the image formed by this imaging optical system and outputs a video signal, and the video signal has a certain vertical and horizontal angle. It is characterized by comprising: a video signal extracting means for extracting only the video signal; and a detecting means for detecting the focal state of the imaging optical system based on the specific frequency component level of the extracted video signal. Focus detection device.