JP3152574B2 - Camera shake correction device and video camera using the same - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a camera shake correction device and a video camera using the same, and more particularly to a camera shake correction device used in a consumer-use camera-integrated VTR and the like, and a video camera using the same.

[0002]

2. Description of the Related Art Conventionally, there has been proposed a method of using a partial motion vector obtained for each detection area by a representative point matching method in order to correct a camera shake component of an imaging apparatus. As a conventional technique for performing camera shake correction using a partial motion vector as described above, there is a technique disclosed in Japanese Patent Application No. 5-156752 proposed by the present applicant.

In this technique, a minimum correlation value, an average correlation value, and position data are obtained for each detection area by an arithmetic circuit.
Given to the microcomputer. The microcomputer obtains a partial motion vector based on the position data of the pixel showing the minimum correlation value. In addition, the microcomputer detects whether the motion vector of each detection area is due to camera shake or other causes based on the average correlation value and the minimum correlation value, and determines whether the detection area is an effective area or an invalid area. I do.

Further, the whole motion vector is detected in the microcomputer by a different method according to the number of effective areas in the detection area. Camera shake is corrected based on the entire motion vector. For example, as shown in FIG. 12, camera shake is corrected by cutting out a part of an image and moving the cutout position in accordance with the entire motion vector.

[0005]

However, in the above-described technique, when an object passes through the detection area, the camera shake correction may malfunction. SUMMARY OF THE INVENTION Therefore, a main object of the present invention is to provide a camera shake correction device which can satisfactorily correct camera shake and a video camera using the same.

[0006]

According to a first aspect of the present invention, a detection area group including a plurality of detection areas is set in a screen, and an entire motion vector is obtained by using partial motion vectors detected from each detection area. In the camera shake correction apparatus, first averaging means for obtaining a first average of partial motion vectors detected from each of the detection areas located at one end of the detection area group and arranged in the first direction, and located at the other end of the detection area group. And a second averaging means for obtaining a second average of the partial motion vectors detected from the respective detection regions arranged in a second direction parallel to the first direction, obtaining an absolute value of a difference between the first average and the second average, First determining means for determining whether or not the absolute value exceeds a first threshold value;
A camera shake correction apparatus comprising means for setting an entire motion vector to zero when a threshold value is exceeded.

According to a second aspect of the present invention, there is provided a camera shake correction apparatus for setting a detection area group including a plurality of detection areas in a screen and correcting a camera shake using a partial motion vector detected from each of the detection areas. Located at one end of the group and the first
First averaging means for obtaining a first average of the partial motion vectors detected from the detection regions arranged in the direction, from the detection regions located on the other end side of the detection region group and arranged in a second direction parallel to the first direction A second averaging means for obtaining a second average of the detected partial motion vectors; obtaining a value obtained by dividing a larger absolute value by a smaller absolute value for each of the absolute values of the first average and the second average; A camera shake correction device comprising: a second determination unit configured to determine whether the value exceeds a second threshold value; and a unit configured to set the entire motion vector to zero when the value exceeds the second threshold value.

A third aspect of the present invention is a video camera using the above-described camera shake correction device.

[0009]

The first average of the partial motion vectors detected from the detection regions located at one end of the detection region group and arranged in the first direction is obtained by the first averaging means. A second average of the partial motion vectors detected from the detection regions located on the other end side of the detection region group and arranged in the second direction parallel to the first direction is obtained by the second averaging means.

For example, the first averaging means obtains, as a first average, the average of the partial motion vectors detected from the detection areas arranged in the uppermost row and in the horizontal direction, and the second averaging means obtains the average in the lowermost row and in the horizontal direction. May be determined as a second average of partial motion vectors detected from the respective detection regions arranged in the range.
Further, the first averaging means calculates the average of the partial motion vectors detected from the detection regions arranged in the leftmost column and in the vertical direction by the first averaging.
The average may be obtained, and the second averaging means may obtain, as the second average, the average of the partial motion vectors detected from the detection regions arranged in the rightmost column and arranged in the vertical direction. Further, if the arrangement state of the detection areas is, for example, rhombic, the first averaging means calculates the average of the partial motion vectors detected from the detection areas located at one end of the detection area group and arranged in the first oblique direction. The second average may be obtained as the first average, and the second average may be obtained as an average of the partial motion vectors detected from the respective detection areas located on the other end side of the detection area group and arranged in the second oblique direction. .

[0011] Then, the first determining means determines the first average and the second average.
The absolute value of the difference from the average is determined, and it is determined whether the absolute value exceeds the first threshold. In the second determining means,
A value obtained by dividing the larger absolute value by the smaller absolute value of each of the first average and the second average is determined, and it is determined whether or not the value exceeds the second threshold.
The first determining means determines that the absolute value exceeds the first threshold value, or the second determining means determines whether the value obtained by the division is equal to the second threshold value.
When it is determined that the threshold value is exceeded, an object that moves other than camera shake enters the detection area and the object exists in a plurality of detection areas. At this time, the entire motion vector is set to zero, and the camera shake is not corrected. In other cases, the entire motion vector is obtained using the partial motion vector.

[0012] Such a camera shake correction device is used for a video camera.

[0013]

According to the present invention, when an object that moves other than a camera shake enters the detection area and an object exists in a plurality of detection areas, the camera shake is prevented from moving. Since the correction is not performed, a malfunction of the camera shake correction can be prevented, and the camera shake correction can be performed well. The above object of the present invention,
Other objects, features and advantages will become more apparent from the following detailed description of embodiments, which proceeds with reference to the accompanying drawings.

[0014]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS A video camera 10 of this embodiment shown in FIG.
Includes a solid-state imaging device 12 such as a CCD that converts a light signal from a subject (not shown) input from a lens 14 into an electric signal. An electric signal from the solid-state imaging device 12 is input to a camera circuit 16. As is well known, the camera circuit 16 includes a sample and hold circuit, and the solid-state imaging device 1
The sample and hold of the electric signal from 2 is performed. The level of the sampled and held electric signal is adjusted by the AGC, and a synchronizing signal is further added by a synchronizing signal adding circuit. Thus, the camera circuit 16 converts the image signal from the solid-state imaging device 12 into an analog video signal. This analog video signal further comprises A
The digital video signal is converted by the / D converter 18. The digital video signal is provided to the motion detection circuit 20. As the motion detection circuit 20, for example, an LSI “L7A0948” manufactured by Sanyo Electric Co., Ltd. is used. Under the control of a memory control circuit 22 included in the same LSI constituting the motion detection circuit 20, digital video signals are written in a field memory 24 in a field sequence.

The motion detecting circuit 20 uses the well-known representative point matching method, for example, to obtain the highest correlation (the minimum correlation value) for each of the four detection areas A, B, C and D shown in FIG. The position of one point and four points around the point and the respective correlation values are calculated. The position data and each correlation value from the motion detection circuit 20 are given to the microcomputer 26.

That is, referring to FIG. 2, motion detection circuit 20 includes an input terminal 28 for receiving a digital video signal from A / D converter 18, and the digital video signal input from input terminal 28 is filtered. The signal is supplied to a representative point memory 32 and a subtraction circuit 34 through 30. The filter 30 is a kind of digital low-pass filter,
It is used to improve the S / N ratio and ensure sufficient detection accuracy with a small number of representative points. The representative point memory 32 extracts a plurality of representative points within each of the detection areas A to D shown in FIG. 3 (in this embodiment, each of the detection areas A to D
Then, 30 representative points are extracted), and the position data and the luminance data are stored. Each detection area 36 (FIG. 4) formed by dividing into 30 is composed of, for example, 32 pixels × 16 rows.

The subtraction circuit 34 calculates the luminance data of the representative point of the previous field given from the representative point memory 32 and the input terminal 2.
8 is subtracted from the luminance data of all the pixels in the current field given by 8 and the absolute value is obtained. That is, a luminance difference is obtained between the luminance data of the current field and the luminance data of the previous field. The obtained luminance difference is given to the accumulation circuit 38. The accumulative addition circuit 38 accumulatively adds (30 in this embodiment) the luminance differences obtained for the pixels at the same position in each detection area 36 in the same detection area to obtain a correlation value, and outputs the correlation value. The correlation value is given to the arithmetic circuit 40,
The arithmetic circuit 40 calculates the minimum correlation value and the average correlation value for each of the detection areas A to D, and obtains the position data of the pixel showing the minimum correlation value for each of the detection areas A to D.
The position data of the pixel indicating the minimum correlation value obtained in this way and the four points around the pixel, the correlation values of these five points, and the average correlation value are supplied from the output terminal 42 to the microcomputer 26 described above. . However, such calculation of the correlation value is executed by the above-described LSI “L7A0948”.

In the microcomputer 26, based on the position data and the correlation value, the motion vector (hereinafter, referred to as the motion vector) of the image on the screen, that is, the entire image field 44 (FIG. 3).
(Referred to simply as “overall motion vector”). First,
Based on the position data of the pixel showing the minimum correlation value, the shift of the pixel showing the minimum correlation value with respect to the representative point is determined, and the shift is used as a partial motion vector. In order to improve the detection accuracy of the partial motion vector, the microcomputer 2
In step 6, interpolation is performed using the correlation values of the four pixels surrounding the pixel having the minimum correlation value, and the position data of the pixel having the minimum correlation value is calculated again. In the microcomputer 26, based on the partial motion vectors,
Equation 4 is determined.

The microcomputer 26 detects whether or not the value obtained by dividing the average correlation value by the minimum correlation value is larger than a predetermined threshold value for each of the detection areas A to D.
It is determined whether the partial motion vectors from .about.D are not erroneously detected by a moving object other than the camera shake and are valid for the determination of the camera shake, that is, whether or not each of the detection areas A to D is an effective area. If the value obtained by dividing the average correlation value by the minimum correlation value is larger than a certain threshold, the detection area is determined to be an effective area.

Specifically, whether or not the area is an effective area is determined as follows. When there is a moving object in the detection area,
The correlation value differs between the portion occupied by the moving object and the portion not occupied, and the portion occupied by the moving object takes various correlation values,
The correlation value generally becomes a large value (the correlation degree is low). Therefore, when there is a moving object in the detection area, there is a high possibility that the minimum correlation value becomes large, and the reliability that the partial motion vector in the detection area is a motion of the camera shake decreases. Using such a partial motion vector,
The whole motion vector is erroneously detected. However, when the average correlation value is large, the reliability of the partial motion vector is high even if the minimum correlation value is large to some extent. On the other hand, when the average correlation value is small, the reliability of the partial motion vector is low unless the minimum correlation value is smaller. Therefore, specifically,
(Average correlation value) / (minimum correlation value)> threshold (for example, 7)
At the time of “condition”, the detection area is determined to be an effective area, and the partial motion vector of the detection area that does not satisfy this condition is not used for calculating the overall motion vector.

Then, in the microcomputer 26,
Using the partial motion vector of the effective area, the amount of motion between the current field and the previous field, that is, the entire motion vector is obtained as described later. The overall motion vector indicates the amount of motion between fields and its direction. Also,
In the microcomputer 26, an integral vector is obtained based on the obtained overall motion vector. The integration vector indicates the amount of shift of the extraction area 46 (FIG. 5) from the center of the image field 44, that is, the amount of correction, and its direction.

The obtained integral vector is supplied to a memory control circuit 22. The memory control circuit 22 determines a read start address of the field memory 24 based on the integrated vector, and uses the address to read the digital video signal stored in the field memory 24. Is read. That is, the memory control circuit 22 controls the field memory 2 according to the integration vector calculated by the microcomputer 26.
4 moves through an extraction area 46 formed by the digital video signals of FIG.

However, the digital video signal read from the field memory 24 as it is, the extraction area 4
6 cannot be moved, the electronic zoom circuit 48 is used.
Referring to FIG. 5, an electronic zoom circuit 48 (FIG. 1) sets an extraction area 46 in which an image is enlarged according to the zoom magnification for the size of image field 44. The position of the extraction area 46 can be freely moved within the range of the image field 44 by changing the read start address of the field memory 24. And
In order to obtain a video signal of the entire image field 44 based on the extracted digital video signal, the image is enlarged using interpolation based on the digital video signal read from the field memory 24.

Thus, the image field 44
By electronically zooming the image of any of the extraction areas 46 in the electronic zoom circuit 48 (FIG. 1), a correctable range 50 corresponding to the difference between the image field 44 and the extraction area 46 can be formed. When camera shake occurs in the video camera 10 as shown in FIG. 6 according to the vibration of the hand of the person who operates the video camera 10, the video camera 10
When the target person exists at the lower left in the image field 44 (FIG. 6).
There are cases where the target person exists in the upper right of the image field 44 (lower in FIG. 6) and the like. Therefore,
By moving the extraction area 46 for each field according to the integration vector calculated by the microcomputer 26, the target person just fits in the extraction area 46 as shown in the right of FIG.

In this way, the digital video signal output from the electronic zoom circuit 48 is converted into an analog signal by the D / A converter 52 and output from the output terminal 54 as a corrected image. The main operation of such a video camera 10 is shown in FIGS. Here, detection areas A to
The case where D is evenly arranged in the image field 44 as shown in FIG. 3 will be described.

First, in step S1 shown in FIG.
It is determined whether there is a detection area that does not satisfy the above conditions. If “YES” in the step S1, the average Vu of the partial motion vectors of the detection areas arranged in the uppermost row and the average Vd of the partial motion vectors of the detection areas arranged in the lowermost row are calculated in the step S3. Next, in step S5, it is determined whether Equation 1 is satisfied.

[0027]

| Vu−Vd | ≧ threshold value b (for example, b = 10) If Expression 1 is satisfied, the overall motion vector is set to zero in step S7. If Expression 1 is not satisfied, the process proceeds to Step S9. In step S9, it is determined whether Equation 2 is satisfied.

[0028]

| Vu | / | Vd | ≧ threshold c (for example, c = 2) where | Vu | ≧ | Vd | | Vd | / | Vu | ≧ threshold c (for example, c = 2) where | When Vu | <| Vd | If Expression 2 is satisfied, the process proceeds to step S7, and the entire motion vector is set to zero. If Expression 2 is not satisfied, the process proceeds to Step S11.

In step S11, the average Vl of the partial motion vectors of the detection regions arranged in the leftmost column and the average Vr of the partial motion vectors of each detection region arranged in the rightmost column are calculated, and the flow advances to step S13. In step S13, it is determined whether Equation 3 is satisfied.

[0030]

| Vl−Vr | ≧ threshold value b (for example, b = 10) If Expression 3 is satisfied, the process proceeds to step S7, and the entire motion vector is set to zero. If Expression 3 is not satisfied, the process proceeds to Step S15. In step S15, it is determined whether Equation 4 is satisfied.

[0031]

| Vl | / | Vr | ≧ threshold c (eg, c = 2) where | Vl | ≧ | Vr | | Vr | / | Vl | ≧ threshold c (eg, c = 2) where | When Vl | <| Vr | If Expression 4 is satisfied, the process proceeds to step S7, and the entire motion vector is set to zero. If Expression 4 is not satisfied, the process proceeds to Step S17.

If "NO" is determined in the step S1, that is, if there is no detection area that does not satisfy the condition, the step S17 is directly performed.
Proceed to. In step S17, for example, as shown in FIG. 8, an overall motion vector is obtained according to the number of effective areas. In FIG. 8, first, in step S21, the number of effective areas is counted. In step S23, it is determined whether or not the number of effective areas is 4, and if not, it is determined in step S25 whether or not the number of effective areas is 3. In steps S23 and S25, if the number of effective areas is 4 and 3, respectively, step S27
Proceed to. That is, if the number of effective areas is three or more, the process proceeds to step S27. In step S27, an arithmetic mean is calculated for each of the partial motion vectors in the horizontal and vertical directions for the effective area. Then, in step S29, the absolute value in the horizontal direction and the absolute value in the vertical direction are obtained for each detection area using the average value obtained in step S27, and the absolute value in the horizontal direction and the absolute value in the vertical direction are obtained for each detection area. The divergence is obtained by adding the absolute value and the absolute value. The absolute value in the horizontal direction is the absolute value of the difference between the horizontal partial motion vector of each detection area and the average value of the horizontal partial motion vectors of the effective area. The absolute value in the vertical direction is
This is the absolute value of the difference between the vertical partial motion vector of each detection area and the average value of the vertical partial motion vectors of the effective area. Then, in step S31, the four divergence values obtained by adding the detection regions are sequentially arranged in ascending order, and two divergence values are selected from the smaller ones. Find the arithmetic mean. Then, in step S33, the arithmetic mean is set as the overall motion vector.

On the other hand, if it is determined in step S25 that the effective area is not 3, then in step S35,
It is determined whether or not the number of effective areas is two. If the number of effective areas is 2, in step S37, a partial motion vector of an arbitrary invalid area is replaced with the entire motion vector obtained one field before, and the process proceeds to step S27. Then, the overall motion vector is determined by the processing in steps S27 to S33. In this embodiment, the partial motion vector is replaced and supplemented for one of the two invalid areas, and the partial motion vector is not supplemented for the remaining invalid area. Therefore, the entire motion vector is determined by the partial motion vectors of the two effective areas and the partial motion vector supplemented by the invalid area.

If it is determined in step S35 that the number of valid areas is not two, it is determined in step S39 whether the number of valid areas is one. If the number of effective areas is 1, the process proceeds to step S41. In step S41, after replacing the partial motion vector of an arbitrary invalid area with the whole motion vector obtained one field before or the whole motion vector obtained two fields before, the process proceeds to step S27.
Then, an overall motion vector is obtained by the processing in steps S27 to S33. In this embodiment, the partial motion vectors of two of the three invalid areas are replaced with the whole motion vectors obtained one field before and two fields before, respectively, to make up for them. Do not supplement the partial motion vector. Therefore, 1
The whole motion vector is determined by the partial motion vector of one effective area and the partial motion vector supplemented by two invalid areas.

When it is determined in step S39 that the number of valid areas is not 1, that is, when it is determined that all the detected areas are invalid areas, in step S43,
The whole motion vector obtained one field before is multiplied by 0.97, and the result is determined as a whole motion vector in step S33. With the video camera operating in this manner, the following effects can be obtained.

For example, as shown in FIG. 9, when an object enters a plurality of detection areas A and C from the horizontal direction,
Since the detection regions A and C become invalid regions, the number of valid regions is reduced, and the resulting overall motion vectors are also unreliable. However, according to this embodiment,
In such a case, for example, Equation 3 or Equation 4 described above is satisfied, so that the entire motion vector is set to zero. Therefore, the whole motion vector, that is, the camera shake detection is not affected by the partial motion vector detected under the influence of the motion of the object.

Although not shown, when an object enters in the vertical direction, Equation 1 or Equation 2 is satisfied, so that the entire motion vector can be similarly set to zero. Further, in the above-described embodiment, the number of detection regions is set to four.
The number of detection areas is not limited to this, and can be set to any number other than four. For example, as shown in FIG. 10, even when the number of detection regions is 5 or more (9 in FIG. 10), the average of partial motion vectors detected from each detection region arranged in the uppermost row and arranged in the horizontal direction is expressed by the following equation (1). Vu of Equation 2 may be used, and the average of the partial motion vectors detected from the detection areas arranged in the lowermost row and in the horizontal direction may be set to Vd of Equations 1 and 2. Similarly, the average of the partial motion vectors detected from the detection regions arranged in the leftmost row and in the vertical direction is calculated as Vl in Expressions 3 and 4.
And the average of the partial motion vectors detected from each detection area arranged in the rightmost row and arranged in the vertical direction is Vr of Expressions 3 and 4.
And it is sufficient. Thus, the present invention can be applied to the case where five or more detection areas are set by using Expressions 1 to 4.

Also in this case, for example, as shown in FIG. 8, the whole motion vector can be obtained by a method corresponding to the number of effective areas. At this time, step S2 shown in FIG.
3, the number of effective areas determined in S25, S35 and S39 is set to an appropriate value. In addition, as shown in FIG.
D may be arranged in a diamond shape. In this case, the average of the respective partial motion vectors of the detection areas A and B is expressed by Equation 1.
And Vu of Equation 2 and the average of the respective partial motion vectors of the detection areas C and D may be Vd of Equations 1 and 2. Also, the average of the respective partial motion vectors of the detection areas B and D may be Vl of Equations 3 and 4, and the average of the respective partial motion vectors of the detection areas A and C may be Vr of Equations 3 and 4. . By using Equations 1 to 4 in this manner, the same effect as in the above-described embodiment can be obtained. This case is particularly effective when an object enters the image field 44 from an oblique direction.

Even when the detection areas are arranged in a diamond shape as shown in FIG. 11, five or more (for example, nine) detection areas are required.
It goes without saying that it may be set. In step S31 shown in FIG. 8, the degree of divergence selected may be one, and the number is arbitrary. In step S43, the constant by which the whole motion vector obtained one field before is multiplied is not limited to 0.97, but is larger than 0 and 1
The following arbitrary coefficients may be used.

Alternatively, whether or not the average correlation value is equal to or more than a predetermined value may be detected for each of the detection areas A to D to determine whether or not the area is an effective area. It may be determined whether or not an area is an effective area only based on the condition relating to the average correlation value, or it is determined that the area is an effective area only when both the above-described conditions of the average correlation value / minimum correlation value and the average correlation value are satisfied. Thus, it may be determined whether or not the area is an effective area based on two conditions.

Specifically, the condition of the average correlation value will be described. When the contrast of the screen is low, since the luminance difference is small, the correlation value is small. For example, when the entire screen is white, the correlation value becomes small. In such a case,
Since the reliability is lost, it is determined to be valid when the average correlation value ≧ the predetermined value. The predetermined value is determined by an experiment.

Further, in the above-described embodiment, the case where the processing is performed on a field basis has been described. However, it is needless to say that the same result can be obtained when the processing is performed on a frame basis. Therefore, in this specification, the term “field” has been described as a term including fields and frames for convenience.

[Brief description of the drawings]

FIG. 1 is a block diagram showing one embodiment of the present invention.

FIG. 2 is a block diagram illustrating an example of a motion detection circuit.

FIG. 3 is an illustrative view showing a principle of electronic zoom and showing a detection area in an image field;

FIG. 4 is an illustrative view showing a principle of electronic zoom and showing representative points and sampling points in a detection area;

FIG. 5 is an illustrative view showing a principle of camera shake correction;

FIG. 6 is an illustrative view showing each block in an image field to which a representative point matching method is applied;

FIG. 7 is a flowchart showing a main operation of this embodiment.

FIG. 8 is a flowchart showing a method of calculating an overall motion vector used in this embodiment.

FIG. 9 is an illustrative view showing a state in which an object enters a detection area from the left.

FIG. 10 is an illustrative view showing a case where five or more detection areas are provided;

FIG. 11 is an illustrative view showing a state in which detection areas are arranged in a diamond shape;

FIG. 12 is an illustrative view showing movement of a cutout position of an image;

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 10 ... Video camera 20 ... Motion detection circuit 22 ... Memory control circuit 24 ... Field memory 26 ... Microcomputer 30 ... Filter 32 ... Representative point memory 34 ... Subtraction circuit 36 ... Detection area 38 ... Accumulation addition circuit 40 ... Arithmetic circuit 44 ... Image field

──────────────────────────────────────────────────続 き Continuation of the front page (56) References JP-A-4-326679 (JP, A) JP-A-4-309078 (JP, A) JP-A-4-213973 (JP, A) JP-A-4- 180369 (JP, A) JP-A-7-38800 (JP, A) JP-A-6-46318 (JP, A) JP-A-6-105211 (JP, A) JP-A-6-284329 (JP, A) (58) Field surveyed (Int. Cl. ⁷ , DB name) H04N 5/222-5/257

Claims (7)

(57) [Claims] 1. A camera shake correction apparatus for setting a detection area group including a plurality of detection areas in a screen and obtaining an entire motion vector using partial motion vectors detected from each of the detection areas. A first averaging means for calculating a first average of partial motion vectors detected from each of the detection areas located on one end side and arranged in the first direction; A second averaging means for calculating a second average of the partial motion vectors detected from the respective detection regions arranged in the parallel second direction; obtaining an absolute value of a difference between the first average and the second average; A camera shake correction apparatus, comprising: first determination means for determining whether the absolute value exceeds a first threshold value; and means for setting the overall motion vector to zero when the absolute value exceeds the first threshold value. 2. A method for obtaining a value obtained by dividing a larger absolute value by a smaller absolute value of each of the first average and the second average, and determining whether the value exceeds a second threshold value. 2. The camera shake correction apparatus according to claim 1, further comprising a second determination unit configured to determine, wherein the overall motion vector is set to zero by the unit even when the value exceeds the second threshold. 3. 3. A camera shake correction apparatus for setting a detection area group including a plurality of detection areas in a screen and correcting a camera shake using a partial motion vector detected from each of the detection areas. First averaging means for calculating a first average of partial motion vectors detected from each of the detection areas located at one end and arranged in the first direction; located at the other end of the detection area group and parallel to the first direction; Second averaging means for calculating a second average of partial motion vectors detected from each of the detection regions arranged in the second direction, wherein a larger one of the absolute values of the first average and the second average is smaller. A second determining means for determining a value obtained by dividing by an absolute value of the second motion vector, and judging whether or not the value exceeds a second threshold value, and a means for setting the total motion vector to zero when the value exceeds the second threshold value Have DOO characterized, image stabilizing unit. 4. The first averaging means calculates a first average of partial motion vectors detected from each detection area arranged in the uppermost row and horizontally, and the second averaging means is arranged in a lowermost row and horizontally arranged. 4. The camera shake correction apparatus according to claim 1, wherein a second average of the partial motion vectors detected from each detection area is obtained. 5. The first averaging means calculates a first average of partial motion vectors detected from each detection area arranged in a leftmost column and vertically, and the second averaging means calculates a first average of partial motion vectors in a rightmost column and a vertical direction. The image stabilization device according to claim 1, wherein a second average of partial motion vectors detected from each of the detection regions arranged in the range is calculated. 6. The first averaging means obtains a first average of partial motion vectors detected from each detection area located at one end of the detection area group and arranged in a first diagonal direction. Is located on the other end side of the detection area group and the first
4. The camera shake correction apparatus according to claim 1, wherein a second average of partial motion vectors detected from each of the detection regions arranged in a second oblique direction parallel to the oblique direction is obtained. 7. A video camera using the camera shake correction device according to claim 1.