JPH05145826A - Automatic focusing device - Google Patents

Abstract

PURPOSE:To enable stable focusing with high accuracy by using a mean level of a signal obtained through M-times sampling of noise for an N period width, thereby performing focusing control. CONSTITUTION:When N-periods of noise X1 is set within M periods of a sampling pulse, that is, within MTs, the relation represented by equation I is obtained, where Ts is a period of the sampling pulse and Tn is a period of the noise X1. In this case, the equation I represents M times sampling of the noise X1 by N periods. Thus, letting the sampling pulse width ts, then when the ts is set according to equation II or set twice the value of the equation II, one period of the noise X1 by M sets of sampling is just sampled without missing. Thus, the periodic component of the periodic noise X1 is completely eliminated by averaging the M sets of sampling values obtained by performing sampling to satisfy the equations I, II.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a camera auto-forward.
More particularly, the present invention relates to an autofocusing device for a video camera that controls a focusing mechanism so that a high frequency component of a video signal is maximized.

[0002]

2. Description of the Related Art A conventional autofocusing method for a video camera is disclosed in Japanese Patent Application Laid-Open No. 58-188966, and a high level in a video signal taken out from an image pickup device through a bandpass filter (BPF). The position of the focusing lens is controlled so that the frequency component is maximized. In the case of fluorescent lighting, the intensity of the video signal is proportional to the brightness of the fluorescent lighting. In the case of lighting at the commercial frequency, the fluorescent lamp discharges at a frequency twice as high as the commercial frequency, so that noise that is twice as high as the commercial frequency is superimposed on the video signal. No flicker removal is discussed in the prior art.

[0003]

Since the fluorescent lamp emits light by discharge, the intensity of light emitted from the fluorescent lamp has an irregular waveform even though the light emission repetition frequency is the commercial frequency. As a result, the conventional apparatus suffers from the problem that the accuracy of autofocusing becomes unsatisfactory and the operation becomes stable because it is affected by the fluctuation of the emission intensity of the fluorescent lamp. SUMMARY OF THE INVENTION It is an object of the present invention to provide a stable and highly accurate autofocusing device by solving the above problems.

[0004]

In order to solve the above-mentioned problems, the focusing lens position is obtained by the output obtained by extracting the high frequency component of the video signal output from the image sensor and sampling the intensity signal for each cycle. In an autofocusing apparatus for a camera, the average value of a predetermined number of consecutive samples of the sampling output is compared, and the focusing mechanism of the camera is driven according to the comparison result.

Further, the period of periodic noise superimposed on the video signal is Tn, and the period of the sampling pulse is Ts.
As an example, the noise with N period width is sampled M times by using integers M and N that satisfy the relation MTs = NTn. Further, the pulse width ts of the sampling pulse is set to the noise period Tn.
1 / M, or an integral multiple thereof.
Further, the sampling pulse is synchronized with the vertical synchronizing frequency of the image sensor.

Further, the average value of the samples is calculated by storing and adding a predetermined number of consecutive sampling outputs, and the difference signal between the average values obtained by storing the predetermined number of the average values is predetermined. A zero difference signal is generated when the difference is less than a value, and the maximum value of the average value is detected and stored when the polarity of the difference signal is inverted, and the difference between the stored average values or the maximum value and the average value is stored. The focus lens position is controlled by the difference signal between the values. Further, when the focusing lens position is controlled by the difference signal between the maximum value and the average value, the control speed is controlled by the difference signal.

[0007]

In the present invention, the focus state of the camera is controlled by the average value of the high frequency components of the video signal. In addition, with M and N as integers, noise of N cycle width is sampled M times, and M
The average value of the sampling values is used as the high frequency component average value of the video signal. Further, by setting the sampling pulse width in the sampling to 1 / M of the noise cycle or an integral multiple thereof, noise of the N cycle width is sampled without omission. Further, the vertical synchronizing frequency of the video signal in the camera is used as the sampling frequency source.

Further, the M number of sampling values are stored and added and averaged to calculate the average value of the high frequency components.
Focusing is controlled by using the difference signal between the average values, and the focusing control is stopped when the difference signal falls below a predetermined value. Further, since the polarity of the difference signal is inverted when the focusing control is over-executed, the maximum value of the average value at this time is stored, and thereafter, the average value is within a predetermined range from the maximum value. Focusing control is performed so as to fit. Further, in the focusing control for bringing the average value close to the maximum value, the control speed is lowered.

[0009]

1 is a block diagram of an embodiment of an autofocusing circuit according to the present invention. In FIG. 1, light is incident on a sensor 6 via a front lens 1 of a zoom lens, a variable power lens group 2, and a focusing lens 3 and converted into a video signal. The focusing lens 3 is moved in the optical axis direction by the AF motor 5 such as a step motor via the gear 4 to adjust the focus. The video signal output of the sensor 6 is amplified by the preamplifier 7 and then N is output by the camera circuit 8.
Simultaneously with conversion to a TSC standard video signal, a high-frequency component is extracted by a band pass filter (BPF) 9, the level is optimized by an attenuator 10, the level is detected by a detection circuit 11, and averaged by an integration circuit 12. It is converted into a focus voltage which is a value.

The detection circuit 11 detects only the video signal in the central portion of the one-field screen. Therefore, the synchronizing signal generating circuit 13 generates a gate pulse synchronized with the horizontal and vertical synchronizing signals necessary for cutting out only the central portion of the one field screen from the output of the attenuator 10 and sends it to the detecting circuit 11. , Control its input. The integrating circuit 12 integrates the video signal at the center of the one-field screen and is reset for each screen. The gate pulse generation circuit 14 supplies the reset pulse to the integration circuit 12.

Next, the output of the integration circuit 12 is converted into a digital signal VF by the AD conversion circuit 15 and applied to the signal processing circuit 17. The signal processing circuit 17 generates a driving signal for the AF motor 5 according to the VF, and the driving circuit 1
The rotation of the AF motor 5 is controlled via 9. AF
When the motor 5 is stopped, the signal processing circuit 17 does not output because the VF does not change. Therefore, the starting voltage is applied from the restarting circuit 18 only for a predetermined time at the time of starting or simultaneously with restarting. The gain detection circuit 16 monitors the above VF and performs automatic gain adjustment for controlling the attenuation value of the attenuator 10. Also, according to this gain adjustment, V
Since F changes in level, the gain detection circuit 16 informs the signal processing circuit 17 of this fact.

As shown in FIG. 2, the AF value is a forward value.
It increases when the focusing lens 3 moves in the focusing direction, and decreases when it moves away from the focusing point. In addition, a fluorescent lamp or other power frequency noise is superimposed on the AF signal,
The effect is greatest near the in-focus point where the AF signal changes slowly. A first object of the present invention is to eliminate the power supply frequency noise and improve focusing accuracy.
The second object is to stop the focusing lens by accurately detecting the in-focus point by the focusing signal in which the power supply frequency noise is eliminated. Further, the focusing signal (maximum value) of the in-focus point is stored, and when the in-focus point is exceeded, the focusing lens is returned toward the maximum value.

FIG. 3 is a waveform diagram for explaining the method of the present invention for canceling the power supply frequency noise. Noise X1 is superimposed on the output of the integrating circuit 12. When the noise X1 is generated by the discharge of the fluorescent lamp, the repetition frequency is 10
0 Hz, the video signal has the above intensity of 100 Hz
Amplitude-modulated by the flicker of the fluorescent lamp, and this is detected by the detection circuit 11, and has a waveform as shown in FIG.

As described above, the central portion of the screen in the detected waveform is cut out and integrated by the integrating circuit 12. In FIG. 3, the shaded portion of the output of the integrating circuit is the portion sampled corresponding to the central portion of the screen. Generally, since the period Tn of the noise X1 and the sampling period Ts are different, the area of the integrated portion as shown in the figure changes depending on the sampling position, and the video signal strength fluctuates apparently, so that the focusing position is not fixed. There will be no.

According to the present invention, an average value of the sampled video signals is obtained to reduce the fluctuation and stabilize the stable signal.
Make it possible to obtain the custody movement. For this reason, first, the condition that the average value does not fluctuate due to the noise X1 is obtained.
Now, assuming that the period of the sampling pulse is Ts and the period of the noise X1 is Tn, and N periods of the noise X1 are exactly accommodated during the M periods of the sampling pulse, that is, MTs, the relation MTs = NTn (1) is obtained. Be done.

Expression (1) means that the noise X1 for N cycles is sampled M times. Therefore, if the sampling pulse width is ts, then ts = Tn / M (2) or if it is set to an integer multiple of this value, it is possible to sample exactly one cycle of the noise X1 without omission by M sampling. From the above, the same as the formula (1) (2)
By averaging the M sample values obtained by sampling so as to satisfy the above condition, the periodic component of the periodic noise X1 can be completely removed.

For example, under fluorescent lamp illumination with a commercial frequency of 50 Hz, the repetition frequency of the noise X1 is 100 Hz, and therefore Tn = 1/100 sec. Further, when the vertical synchronizing frequency (field frequency, 60 Hz) of the NTSC standard built into the camera is used as the sampling frequency, Ts = 1/60 sec.
Becomes Ts / Tn = 5/3 M = 3, N = 5 (3), and it is sufficient to sample the noise X1 for 5 cycles three times from the equation (1). Further, according to the equation (2), ts = 1/60 sec (4) or an integral multiple of this value.

By satisfying the above equations (3) and (4), the periodic component in the noise X1 can be completely removed regardless of the waveform. Further, when the sampling pulse width ts deviates from the above condition, although the noise removal is slightly incomplete, a sufficiently practical effect can still be obtained.

The process of averaging will be described with reference to FIG. 2. When a _{1 to} a ₇ , b ₁ b ₂ etc. are the sampling points, the present invention uses, for example, a ₇ ,
The average value of a ₆ , a ₅ is calculated, then the average value of a ₆ , a ₅ , a ₄ is calculated, and the average value is calculated in the same manner. As a result, it is possible to obtain a video signal sample that is not affected by the fluctuation due to the noise X1 every sampling period Ts.

The signal processing circuit 17 of FIG. 1 generates the above-mentioned video signal sample and outputs a focusing signal. Figure 4
Is a block diagram of a main part of the signal processing circuit 17. The signal processing circuit 17 and the A / D converter 15,
Of course, the function of the gain detection circuit 16 or the like may be performed by the microcomputer. In FIG. 4, the output of the A / D converter 15 is the VF in the memory 20.
The data are sequentially stored in the memories M1 to M3 and are sequentially pushed out. The data stored in each VF memory is a shown in FIG.
It is an integrated signal of the high frequency component intensity of the video signal of the central portion of each screen corresponding to _{1 to} a ₇ , b ₁ b _2, etc.

The data of the VF memories M1 to M3 is added by the adder 21 every time one data is pushed from the A / D converter 15.
Is added and averaged by the divider 22. Therefore, the divider 22 sequentially outputs the average value of the video signal data for 3 screens (3 fields). The divider 22 may be omitted. The output of the divider 22 is sequentially stored in the memories MO1 and MO2 in the average value memory 23,
The increment detection circuit 24 determines the magnitude relationship. If the in-focus direction and the out-of-focus direction are judged by the above-mentioned increment detection result,
The moving direction of the focusing means can be determined without being affected by flicker. For example, the data value of the memory MO2 is MO1.
If the value is larger than the data value of, the focusing lens 3 is moving toward the focal point, and in the opposite case, the focusing lens 3 is away from the focal point.
Therefore, since the position at which the magnitude relation is reversed is the in-focus point, the increment detection circuit 24 also detects the inversion of the magnitude relation and sends a control signal to the maximum value memory 25 to maximize the average value data at the in-focus point. Store as a value.

The AF motor 5 for driving the focusing lens 3 is driven by a voltage proportional to the difference between the data of the memories MO1 and MO2. Usually, this difference value becomes small near the in-focus point, so that the focusing lens 3 moves relatively slowly and stops at the in-focus point because it goes too far. When the focus point is exceeded too much, the moving direction of the AF motor 5 is reversed so that the data of the average value memory 23 is within the range of a predetermined width in the data stored in the maximum value memory 25. Stop where you stand. Mean value memory 23
Can be detected by the discrepancy between the data in the maximum value memory 25 and the data in the maximum value memory 25, and the increment detection circuit 24 also outputs a drive signal of the AF motor 5 for correcting this discrepancy.

In the above operation, the maximum value memory 25
It is important to associate the average value data stored by the focus point with the focus data as accurately as possible. However, since the average value data is the average of the two past VF data and the current VF data, when the magnitude relationship between the two average value data is reversed, the AF motor has already been combined. The focus may have been overshot, and the inertia of the AF motor system also causes overshoot. Therefore, there are many cases where the true focusing data exists in the data slightly past the time when the inversion is detected. In FIG. 4, MO2 in the average value memory 23 is past MO1 rather than MO1.
Is stored in the maximum value memory 25. To what extent the past data is the maximum value depends on the characteristics of the system. Therefore, if necessary, a memory such as MO3 or MO4 is added next to the above MO2, and one of these is set to the maximum value. It is stored in the memory 25. Further, in FIG. 4, the data of the three VF memories are averaged, but according to the equation (3), the above V
The number of F memories can be increased.

FIG. 5 is a flowchart for explaining the signal processing procedure in FIGS. 1 and 4 in more detail. At step 31, an AF motor starting voltage at the time of starting or at the time of restart is generated and applied for a predetermined time, and then,
At step 32, only the video signal at the center of each field screen is detected, integrated and A / D converted, and at step 33 A
The a / D conversion output is monitored and the attenuator gain is controlled so as to optimize the VF signal level.

Next, at step 34, the VF signal is checked and if there is no abnormality, at step 35 three consecutive VFs are detected.
Find the average value of the signal. Next, in step 36, the difference ΔVF between the average values is calculated. If ΔVF is within a predetermined minute width, it is determined that ΔVF≈0, the AF motor is stopped in step 37, and step 32 is executed. Return. ΔV
If F is larger than the above-mentioned minute width, the process proceeds to step 38 to drive the AF motor, and if ΔVF> 0. If it is determined that the focal point has not been reached yet, the operation proceeds to step 38 and the driving of the AF motor is continued. If the polarity of ΔVF is reversed, it is determined that the focal point has passed and V
The maximum value of the F signal average value is stored, and AF is performed in step 40.
After reversing the moving direction of the motor, the process proceeds to step 38, the AF motor is driven, and the process returns to step 32.

[0025]

According to the present invention, a video signal monitor type video camera capable of performing stable and accurate focusing by removing the influence of the flicker of the brightness of the commercial frequency under the fluorescent lamp illumination, and an electronic type. It is possible to provide an autofocusing device for a still camera or the like. Further, according to the present invention, since the video signal including the noise is sampled at the vertical synchronizing frequency, the circuit can be economically constructed. Further, since it is possible to detect and correct the overshooting of the autofocusing, it is possible to improve the focusing accuracy, prevent hunting at the same time, and perform focusing.
You can increase the casting speed.

[Brief description of drawings]

FIG. 1 is a block diagram of an embodiment of an autofocusing system according to the present invention.

FIG. 2 is a characteristic diagram showing a relationship between a position of a focusing lens and a video signal level.

FIG. 3 is a waveform diagram illustrating the effect of video signal averaging in the present invention.

FIG. 4 is a block diagram of a focusing signal generation circuit according to the present invention.

FIG. 5 is a flowchart of focusing control according to the present invention.

[Explanation of symbols]

 3 Focusing Lens 5 Motor for Auto Focusing (AF Motor) 6 Sensor (Imaging Element) 9 Band Pass Filter (BPF) 10 Attenuator Circuit 11 Detection Circuit 12 Integration Circuit 13 Sync Signal Generation Circuit 15 AD Converter 16 Gain detection circuit 17 Signal processing circuit 18 Restart circuit 19 AF motor drive circuit 20 Memory 21 Adder 22 Divider 23 Average value memory 24 Increment detection circuit 25 Maximum value memory

─────────────────────────────────────────────────── ─── Continuation of the front page (51) Int.Cl. ⁵ Identification code Office reference number FI technical display location G03B 13/32

Claims (7)

[Claims] 1. An autofocus for a camera in which a focusing lens position is controlled by an output obtained by extracting a high frequency component of a video signal output from an image sensor and sampling the intensity signal at each cycle. In the Casing device,
The in-focus state is provided with an average value calculating means for calculating an average value of a predetermined number of consecutive samples of the sampling output, and an in-focus state determining means for comparing the increase and decrease of the average value to determine an in-focus state. An autofocusing device characterized in that the focusing mechanism of the camera is driven by the output of the judging means. 2. The N cycles according to claim 1, wherein the cycle of the periodic noise superimposed on the video signal is Tn, the cycle of the sampling pulse is Ts, and integers M and N satisfying the relationship MTs = NTn are used. An autofocusing device characterized in that the noise of the width is sampled M times. 3. The autofocusing device according to claim 2, wherein the pulse width ts of the sampling pulse is set to 1 / M of the noise period Tn or an integral multiple thereof. 4. The method according to any one of claims 1 to 3,
An autofocusing device, wherein the pulse rate of the sampling pulse is the vertical synchronizing frequency of the image sensor. 5. The method according to any one of claims 1 to 4,
The autofocusing device according to claim 1, wherein the average value calculating means includes a memory for storing a predetermined number of consecutive sampling outputs and a means for adding the stored values of the memories. 6. The method according to any one of claims 1 to 5,
The in-focus state determining means detects a difference between an average value memory storing a predetermined number of continuous outputs of the average value calculating means and an average value stored in the average value memory and outputs a difference signal. Means, means for generating a zero difference signal when the difference between the average values is less than or equal to a predetermined value, means for detecting and storing the maximum value of the average value when the polarity of the difference signal is inverted, and the average A means for detecting a difference between the maximum value of the average value and the average value stored in the average value memory is provided, and by a difference signal between the average values or a difference signal between the maximum value of the average value and the average value. An autofocusing device characterized in that the position of the focusing lens is controlled. 7. A means for varying the control speed of the focusing lens according to claim 6, when the focusing lens position is controlled by a difference signal between the maximum value of the average value and the average value. An autofocusing device characterized in that