JPH08334683A - Automatic in-focus device - Google Patents

Abstract

PURPOSE: To accurately detect the in-focus position and to improve reliability, even if an object is one having high luminance and immediately saturated to a white level, such as flames. CONSTITUTION: A contrast signal is detected by means 2 and 3, based on a video signal inputted from a camera and the contrast signal per unit-region in the region where the contrast signal is high between far and near points is obtained with the detected contrast signal by means 1 and 4. This automatic focusing device is provided with a means 5 for making levels at the far and near points except an in-focus point relatively lower than the level at the in- focus point by using not the contrast signal for making its level high at the far and near points except the in-focus point, but the contrast signal per unit- region, when the object is one having the high luminance and immediately saturated to the white level, such as the lames, to make the peak point of the contrast signal per unit-region the in-focus point.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an automatic focusing device.

[0002]

2. Description of the Related Art Automatic focusing in which a contrast signal (high frequency component signal or high frequency component amount signal) is detected from a video signal input from a camera (television camera) and automatic focusing is performed based on the contrast signal. The device is known.

FIG. 7 shows a schematic configuration of this automatic focusing device. In the figure, reference numeral 7 indicates a subject, and the subject 7 is, for example, a CCD (C
harge Coupled Device) Imaged by the camera 8.

The subject imaged by the camera 8 is output as a video signal from the camera 8, and the contrast signal is detected by the high frequency component extracting means 10 based on this video signal. The range from which the contrast signal is extracted, that is, the range of the subject 7 to be focused (focus area) is arbitrarily set in, for example, a television image.

The extraction of the contrast signal by the high frequency component extracting means 10 is performed by differentiating the video signal in the scanning line direction and adding the peak value of this differential waveform or the differential value in the focus area. Which of these is used depends on the condition of the subject and needs, but in general, the peak value (high frequency component) of the differential waveform when the contrast is high, and the value obtained by adding the differential value in the focus area when the contrast is low. (Amount of high frequency component) is used respectively.

The optical system 9 is moved in a direction facing the subject 7 by a motor (not shown). That is, the optical system 9 is moved from the far point (right side in the drawing) to the near point (left side in the drawing) with respect to the subject 7. This optical system 9 is the subject 7
On the other hand, the contrast signal extracted when moving from the far point to the near point is shown in the graph portion of FIG. 7. As is clear from this, when the optical system 9 is located at the focus position, the contrast signal is As the peak value of the differential waveform or the value obtained by adding the differential value in the focus area, the contrast signal has a peak (maximum value). That is, the focus position detection means 11 moves the optical system 9 so that the optical system 9 is located at this peak value, and when this optical system 9 is located at the peak value position, focus is achieved. It has become.

[0007]

By the way, the subject 7
Is a flame of a candle as shown in FIG. 8 (a), the flame has high brightness and is immediately saturated with a white level, and a blurred image is shown in FIG. 8 (b). As shown, it grows saturated.

9 (a) and 9 (b) show differential waveform images corresponding to the images shown in FIGS. 8 (a) and 8 (b). There is no change in that the size of is also expanding. Therefore, the extracted contrast signal is, as shown in FIG.
On the far point side and the near point side, the blur increases as the blur expands and becomes larger than the peak value at the in-focus point.

That is, the conventional method, that is, the peak value of the differential waveform or the value obtained by adding the differential value in the focus area is obtained as a contrast signal from the far point to the near point,
In the method in which the peak value in these contrast signals is used as the in-focus position, when the subject 7 is saturated to a white level immediately with high brightness such as flame, for example, the far point side other than the in-focus position is used. There is a problem in that the point and the point on the near point side are mistakenly recognized as the in-focus position and the reliability is lowered.

Therefore, the present invention provides an automatic focusing device in which the focusing position is accurately detected and the reliability is improved even when the subject is saturated to a white level immediately with high brightness such as flame. The purpose is to

[0011]

In order to achieve the above object, an automatic focusing apparatus according to the present invention detects a contrast signal from a video signal input from a camera, and detects the contrast signal from a far point to a near point. In the automatic focusing device that is designed to perform automatic focusing based on the peak value between the points, the contrast signal per unit area of the area of the contrast signal large from the far point to the near point is calculated, and The peak point of the contrast signal of was set as the focal point.

In order to achieve the above object, the automatic focusing apparatus according to the second aspect detects a high frequency component signal from the video signal from the far point to the near point input from the camera, and the high frequency component of this video signal is detected. In an automatic focusing device configured to detect a focus position based on a component, a generation area detection unit that obtains a generation area of a high frequency component from the video signal, and a high frequency component in the generation area from the video signal High frequency component detecting means, high frequency component amount detecting means for detecting the amount of high frequency components in the generation area using the output from the high frequency component detecting means, and the output from the high frequency component amount detecting means for detecting the generation area. The high-frequency component amount correcting means for detecting the high-frequency component amount per unit area by dividing by the output from the means, and the peak value of the output from the far point to the near point of this high-frequency component amount correcting means A judging means for focus, with a.

Further, in order to achieve the above-mentioned object, the automatic focusing device according to claim 3 is characterized in that, in addition to the above-mentioned claim 2, the high-frequency component detecting means excludes high-frequency components in the saturation region. .

[0014]

According to the automatic focusing device of the first aspect, the contrast signal is detected based on the video signal input from the camera, and the contrast signal from the far point to the near point is used by using the contrast signal. The contrast signal per unit area of the signal-large area is obtained, and the peak point of the contrast signal per unit area is set as the focal point. Here, assuming that the subject has a high brightness like a flame and is saturated to an immediately white level, the rising edge of the video signal is steep at the in-focus point and is blurred and gradual at the far point and the near point other than the in-focus point. Therefore, as shown in FIG. 10, the contrast signal is large at the in-focus point and small at the periphery thereof, becomes larger than the in-focus point at the far point and the near point other than the in-focus point, and the area where the contrast signal changes is Since the signal sharply changes, it is small at the in-focus point, and becomes very large at the far point and near point other than the in-focus point because the signal change becomes gentle. Therefore, the level of the contrast signal per unit area becomes low level at the far point and near point other than the in-focus point and peaks at the in-focus point.
Since the peak point of the contrast signal per unit area is the focus point, the focus position can be accurately detected.

Further, according to the automatic focusing apparatus of the second aspect, the generation area detecting means can obtain the generation area of the high frequency component from the video signal, but the subject has a high brightness such as a flame and an immediately white level. If it is saturated, the rising edge of the video signal will be steep at the in-focus point and will become blurred and gradual at far points and near points other than the in-focus point. It becomes very large at far points and near points other than. Further, the high frequency component detecting means detects the high frequency component in the generation area from the video signal, and the high frequency component amount detecting means uses the output from the high frequency component detecting means to detect the amount of the high frequency component in the generation area. However, the amount of high-frequency components is large at the in-focus point and small at the periphery, and far points other than the in-focus point,
It becomes larger than the focal point at the near point. Then, when the high frequency component amount correcting means divides the output from the high frequency component amount detecting means by the output from the generation area detecting means to detect the high frequency component amount per unit area, the level thereof is other than the in-focus point. It becomes a low level at the far point and the near point of, and peaks at the in-focus point. Therefore, since the peak value of the output from the far point to the near point of the high frequency component amount correcting means is set as the in-focus point by the determining means, the in-focus position can be accurately detected.

Further, according to the automatic focusing device of the third aspect, the high frequency component detecting means excludes the high frequency component in the saturation region, and the detection accuracy of the high frequency component is further improved.

[0017]

Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram showing a configuration of an automatic focusing device according to an embodiment of the present invention. A portion surrounded by reference numeral 12 shown in FIG. 7 of the prior art, that is, a high frequency component extracting means 10 and a focus position detecting means. It corresponds to 11.

The automatic focusing apparatus according to the present embodiment comprises a generation area detecting means 1 for obtaining a generation area of a high frequency component in accordance with a video signal from a camera 8 (the same as that of the prior art), and within the generation area from the video signal. And a high frequency component detecting means 2 for detecting the high frequency component.

The generation area detecting means 1 is an A / D converter 1 for converting a video signal (analog signal) into a digital signal.
The difference (1) detecting means 1b for obtaining a high-frequency component (1) as a rough change by taking the difference between a and the video signal delayed by a relatively large time with respect to this digitally converted video signal, and the difference (1) detecting means 1b. The binarizing means 1c for binarizing the obtained high frequency component (1) data and detecting the area where the high frequency component (1) is actually obtained, and the autofocus area set by the autofocus area setting means 6 (conventional technique). And the range obtained by counting the binarized area based on the area obtained by the binarizing means 1c, that is, the range of the data of the high frequency component (1), that is, a rough change region. Area count means 1d for a binarized area for detecting the range of The rough change area is an area in which a high-frequency component of a video signal has a level change, and is for detecting an area to be subjected to image processing for focusing.

The high-frequency component detecting means 2 includes the A / D
Difference (2) detecting means for obtaining a high-frequency component (2) as a minute change by taking the difference between the converter 1a and the video signal delayed by a time smaller than the difference (1) with respect to the digitally converted video signal. 2a, a comparator 2c for comparing the video signal at a white level, a saturation region detector 2d for detecting a saturation region based on the output of the comparator 2c, the saturation region and the binarized area. Data exclusion means 2b for excluding external data from the obtained high frequency component (2) data.

The data excluding means 2b and the autofocus area setting means 6 include a high frequency component (2) excluding the data outside the saturation area and the binarized area and the high frequency component (based on the autofocus area). The difference amount adding means 3 as the high frequency component amount detecting means for detecting the amount of 2) is connected. The difference amount adding means 3 and the area counting means 1d for the binarized area divide the added difference amount by the area count value to detect a high frequency component amount correction for detecting a high frequency component amount per unit area (unit area). The means 4 is connected, and the judging means 5 is connected to the high frequency component amount correcting means 4.

The determination means 5 combines the storage means 5a for storing the output from the high frequency component amount correction means 4 in the memory for each screen from the far point to the near point, and the peak value in the stored data. Focus position detecting means 5b which determines a focus position and sends a signal for moving the optical system to this focus position to the motor.
It consists of and.

Next, the operation of the automatic focusing device having such a configuration will be described below. First, the camera 8 images a candle as shown in FIG. 8A, for example. In this case, the optical system 9 is sequentially moved from the far point to the near point, and the following processing is performed at each position, but for convenience of explanation,
The case where the optical system 9 is at the in-focus position and the case where the optical system 9 is at, for example, the near point side (even the far point side is possible) from the in-focus position will be described in parallel.

Here, when the optical system 9 is at the in-focus position, the video signal has a steep rising edge as shown in FIG. As shown in FIG. 3A, the rising edge of the video signal at the position is blurred and becomes gentle.

The difference between this video signal and the video signal delayed by a certain amount of time is taken, and a high frequency component (1) is obtained. The high-frequency component (1) is, for example, as shown in FIG. 2B at the in-focus position and as shown in FIG. 3B at the position closer to the in-focus position. That is, comparing the high-frequency component (1) between the in-focus position and the near-point side position, the in-focus position has a higher level and a narrower range than those at the near-point side position.

This high frequency component (1) is binarized. That is, the area where the high frequency component (1) is actually obtained is obtained, and is shown in the range shown in FIG. 2 (b) at the focus position and shown in FIG. 3 (b) at the position closer to the focus position. It will be the range to be.

Area counting is performed based on this area. Therefore, the area count value at the in-focus position is considerably smaller than that at the near point side position.

Further, with respect to the above video signal, a difference (1)
The difference from the video signal delayed by a smaller time is taken to obtain the high frequency component (2). This high frequency component (2) is
The focus position is as shown in FIG. 2C, and the position nearer to the focus position is as shown in FIG. 3C. That is, since the high frequency component (2) is the result of delaying the high frequency component (1) by a short time, the high frequency component (2) at the in-focus position is the high frequency component (1) at the same position. As the level becomes smaller, the range becomes narrower. This also applies to the position on the near point side.

For the high frequency component (2), data outside the saturation area and the binarized area is excluded. That is,
The range shown in FIG. 2C is excluded at the in-focus position, and the range shown in FIG. 3C is excluded at a position closer to the in-focus position. This exclusion of the saturated region is to improve the detection accuracy of high frequency components.

Then, the high frequency component (2) excluding the data outside the saturation region and the binarized area is added to obtain the high frequency component amount. That is, in the in-focus position, as shown in FIG.
The area of the shaded area in (c) is obtained, and the area of the shaded area in FIG. 3 (c) is obtained at a position closer to the in-focus point.

Here, in the prior art, the peak value of the differential waveform (the peak value of FIG. 2 (b) or (c), FIG.
(Corresponding to the peak value of (b) or (c)), or a value obtained by adding the differential value in the focus area (corresponding to the area of FIG. 2C, the area of FIG. 3C) from the far point to the near point Since it was a method of determining the peak value among these as the in-focus position, it is clear from the figure that either the peak value of the differential waveform or the value obtained by adding the differential value in the focus area is taken. Also, the peak value on the near point side becomes larger than that at the in-focus position (see FIG. 4), which results in erroneous recognition of the in-focus position.

However, the following processing is further performed in this embodiment. That is, the value obtained by adding the high frequency component (2) (high frequency component amount) is divided by the area count value of the binarized area. Therefore, the amount of high frequency components per unit area is detected.

This high frequency component amount per unit area is stored for each screen from the far point to the near point (the high frequency component amount per unit area is actually the focus position and the near point side as described above. The position is not something that is required in parallel).
Therefore, as shown in FIG. 4, the high-frequency component amount (the sum of the high-frequency components (2); that is not divided by the area count value of the binarized area) is large at the in-focus position and small at the periphery thereof. , Far point and near point other than the in-focus position are larger than the in-focus position, but high frequency component (1)
The area (area count value of the binarized area) is small at the in-focus position and very large at the far point and near point other than the in-focus position, as shown in FIG. When the level of the amount of high-frequency component is obtained, the level becomes low at the far point and near point other than the in-focus position and peaks at the in-focus position, as shown in FIG. And
The peak point of the amount of high frequency components per unit area is set as the focus point, and the optical system 9 is moved to the peak position (focus position).

As described above, in the present embodiment, when the subject is saturated to a white level with high brightness such as flame immediately, the level becomes large at the far point and the near point other than the in-focus point. By using the amount of high-frequency components per unit area instead of the amount of high-frequency components, the level at the far point and near point other than the in-focus point is set to a relatively low level with respect to the level at the in-focus point. Since the peak value of the output up to the near point is set as the in-focus point, the in-focus position can be accurately detected, and the reliability can be improved.

Since the high frequency component in the saturated region is excluded from the high frequency components to improve the detection accuracy of the high frequency component, the effect of the present invention can be made more effective. .

Although the invention made by the present inventor has been specifically described based on the embodiments, the present invention is not limited to the above embodiments, and various modifications can be made without departing from the scope of the invention. Needless to say, for example, in the above-described embodiment, the subject 7 is a candle flame, but the subject is not limited to this.

[0037]

As described above, according to the automatic focusing apparatus of the first aspect, the contrast signal is detected based on the video signal input from the camera, and the contrast signal is used to change from the far point to the near point. When the contrast signal per unit area of the area with a large contrast signal between is obtained and the subject is saturated to a white level immediately with high brightness like flame, for example, far point and near point other than the in-focus point. By using the contrast signal per unit area instead of using the contrast signal whose level becomes large, the level at the far point and the near point other than the in-focus point is made relatively low with respect to the level at the in-focus point. Since the peak point of the contrast signal per unit area is set as the in-focus point, the in-focus position can be accurately detected and the reliability can be improved.

According to the automatic focusing device of claim 2,
The generation area detection means finds the generation area of the high frequency component from the video signal, the high frequency component detection means detects the high frequency component in the generation area from the video signal, and the high frequency component amount detection means detects the high frequency component. The amount of the high frequency component in the generation area is detected using the output from the means, and the high frequency component amount correction means divides the output from the high frequency component amount detection means by the output from the generation area detection means to determine the unit area. When the amount of high frequency component per hit is detected and the subject is saturated to a white level immediately with high brightness like flame, for example, the amount of high frequency component that becomes large at the far point and near point other than the in-focus point is calculated. By using the high frequency component amount per unit area without using, the far point other than the in-focus point,
The level at the near point is set to a relatively low level with respect to the level at the in-focus point, and the determination means determines that the peak value of the output from the far point to the near point of the high-frequency component correction means is the in-focus point. Since it is a thing, it is possible to accurately detect the in-focus position,
It is possible to improve reliability.

According to the automatic focusing device of claim 3,
In addition to the second aspect, the high frequency component detecting means is configured to exclude the high frequency component in the saturation region to further improve the detection accuracy of the high frequency component. Therefore, the effect of the second aspect is made more effective. It becomes possible.

[Brief description of drawings]

FIG. 1 is a block diagram showing a configuration of an automatic focusing device according to an embodiment of the present invention.

2A and 2B are signal diagrams for explaining the operation of the same device when the optical system is located at the in-focus position, where FIG. 2A is a video signal, FIG. 2B is a difference (1) signal, and FIG. Are the differential (2) signals respectively.

FIG. 3 is each signal diagram for explaining the operation of the above-mentioned device when the optical system is separated from the in-focus position on the far point side or the near point side. FIG. ) Represents the difference (1) signal, and (c) represents the difference (2) signal.

FIG. 4 is a corresponding diagram showing a difference amount added value when the optical system is moved from a far point to a near point.

FIG. 5 is a correspondence diagram showing a binarized area count value when the optical system is moved from a far point to a near point.

FIG. 6 is a correspondence diagram showing a difference amount added value per unit area when the optical system is moved from a far point to a near point.

FIG. 7 is a schematic configuration diagram showing an automatic focusing device in the related art.

8A and 8B show an object, in which FIG. 8A shows the case where the optical system is located at the in-focus position, and FIG. 8B shows the optical system at the far point side or the near point side with respect to the in-focus position. This is for the case where they are separated.

FIG. 9 shows a differential waveform image corresponding to the images shown in FIGS. 8A and 8B, where FIG. 9A shows the case where the optical system is located at the in-focus position, and FIG. Shows the case where the optical system is located at the far point side or near point side with respect to the in-focus position.

FIG. 10 is a correspondence diagram showing a contrast signal when the optical system is moved from a far point to a near point.

[Description of Reference Signs] 1 generation area detection means 2 high frequency component detection means 3 high frequency component amount detection means 4 high frequency component amount correction means 5 determination means 8 camera

Claims (3)

[Claims] 1. An automatic focusing device for detecting a contrast signal from a video signal input from a camera and performing automatic focusing based on a peak value between the far point and the near point of the contrast signal. The automatic focusing device is characterized in that the contrast signal per unit area of the area having a large contrast signal from the far point to the near point is obtained, and the peak point of the contrast signal per unit area is set as the in-focus point. 2. An automatic focusing device which detects a high frequency component signal from a video signal from a far point to a near point input from a camera and detects a focus position based on the high frequency component of the video signal. A generation area detection means for obtaining a generation area of a high frequency component from the video signal; a high frequency component detection means for detecting a high frequency component in the generation area from the video signal; and an output from the high frequency component detection means High frequency component amount detecting means for detecting the amount of high frequency components in the generation area, and high frequency for detecting the amount of high frequency components per unit area by dividing the output from this high frequency component amount detecting means by the output from the generation area detecting means. An automatic focusing device comprising: a component amount correcting means; and a determining means for setting a peak value of an output from a far point to a near point of the high frequency component amount correcting means as a focus point. 3. The automatic focusing device according to claim 2, wherein the high frequency component detecting means excludes high frequency components in the saturation region.