JPH06217263A - Motion correction system interpolation signal generating device - Google Patents

Abstract

PURPOSE:To eliminate the distortion of an interpolation signal by selecting an interpolation signal generated from a motion-uncorrected television signal when the value of a motion vector exceeds a specific value. CONSTITUTION:A comparing circuit 43 compares the value ¦V¦ of the motion vector detected by a motion vector detecting circuit 13 with the threshold value Talpha outputted from a threshold value generating circuit 42 to decide whether or not ¦V¦ exceeds Talpha. The decision result of this comparing circuit 43 is supplied to an adaptive motion interpolation switching control circuit 41. This control circuit 41 generates an adaptive switching coefficient beta on the basis of inter-field difference signals S4 and S5 when receiving the decision result indicating that ¦V¦ does not exceed Talpha or forcibly sets beta to 0 when receiving the decision result indicating the Ta is exceeded. Namely, when ¦V¦ exceeds the threshold value Talpha, an uncorrected field interpolation signal S2 is wholly selected irrelevantly to the levels of the inter-field difference signals S4 and S5.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a motion compensation type interpolation signal generating apparatus for subjecting a television signal to motion compensation based on a motion vector and generating a signal for interpolation from the television signal.

[0002]

2. Description of the Related Art In recent years, in the television broadcasting system, the conversion of the television system has been actively performed with the internationalization of the television broadcasting.

One of the conversion items of the television system is conversion of the number of fields. This conversion is performed by thinning or inserting fields. However, if you simply do this, the movement will be unnatural. Therefore, usually, a plurality of fields are weighted and added to synthesize a new field.

However, even if a plurality of fields are weighted and added, the blurring of the edge portion occurs due to the conversion in the moving image area of the screen.

For this reason, it is desired that the interpolating signal generating device for field conversion be capable of synthesizing a new field so that such bleeding does not occur.

In order to meet this demand, conventionally, the position of the moving image area in a new field is predicted based on the motion vector, the position of the moving image area is corrected based on this prediction result, and the interpolation signal is output from this corrected output. A so-called motion-compensated interpolation signal generating device for generating a signal is considered.

FIG. 2 is a block diagram showing the configuration of a conventional motion compensation type interpolation signal generating apparatus.

The illustrated apparatus detects a motion vector,
A portion (13 to 1) that performs motion correction on the input signals PS and FS.
6), a portion (18 to 19) for generating the interpolated signal S1 from the signals PS and FS subjected to motion correction, and a portion (20) for generating the interpolated signal S2 from the signals PS and FS not subjected to motion correction. 22 to 22) and a portion (23 to 25, 27 to 31) that adaptively switches between the two interpolation signals S1 and S2.

With such a configuration, the coordinates of the moving image area can be matched between the television signal PS of the current field and the television signal FS of the previous field, so that the above-mentioned bleeding does not occur. Interpolation signal S1
Can be generated.

[0010]

However, in the conventional motion compensation type interpolation signal generating apparatus described above,
There were the following problems.

That is, let us consider a case where the camera catches a moving object (a circle in this figure) and is panning as shown in FIG. 3 (a). In this case, in the television image, the moving object is the still image area and the background is the moving image area.

In such a television image, image distortion hardly occurs when the background is flat (when the level of the image signal does not change) or when the background movement is small.

However, if the background is not flat and the motion is large, the motion vector suddenly changes behind the moving object (circle in the figure). This causes distortion in the interpolated signal S1 generated from the motion-corrected signals PS and FS.

Even in this case, the interpolated signals S1 and S2 are
If the other interpolated signal S2 is selected by the adaptive switching of, the true interpolated signal S3 will not be distorted.

However, in this case, since the background which is the shadow of the moving object appears on the screen of the current field, the accuracy is lowered in the adaptive switching using the difference value between fields as a basic parameter.

As a result, since the interpolated signal S1 is selected, the true interpolated signal S3 is distorted as shown in FIG. 3 (b).

Therefore, according to the present invention, even when a moving object moving at a high speed is panned under an uneven background, the motion compensation system can prevent distortion of the interpolation signal. It is an object of the present invention to provide an interpolation signal generation device of

[0018]

In order to achieve the above object, the present invention is generated from a television signal that is not forcibly subjected to motion compensation when the magnitude of a motion vector exceeds a predetermined magnitude. The interpolation signal is selected.

[0019]

According to the above construction, when the magnitude of the motion vector exceeds a predetermined magnitude, the interpolation signal generated from the television signal that has not been subjected to motion correction is selected, so that the motion vector is rearranged behind the moving object. Even if the motion vector suddenly changes, the true interpolated signal will not be distorted.

[0020]

Embodiments of the present invention will be described in detail below with reference to the drawings. FIG. 1 is a block diagram showing the configuration of an embodiment of the present invention.

The apparatus of FIG. 1 produces two interpolated signals,
This is the same as the apparatus of FIG. 2 in that a true interpolation signal is generated by adaptively switching these.

However, in the apparatus of FIG. 2, adaptive switching is controlled based on the difference value between the fields, whereas in the apparatus of FIG. 1, control is performed based on this difference value and the magnitude of the motion vector. The difference is. Therefore, in FIG. 1, the same parts as those in FIG. 2 are designated by the same reference numerals.

In the figure, reference numeral 11 is an input terminal to which a digitized current field television signal (hereinafter referred to as "current field signal") PS is supplied. 1
Similarly, 2 is a digitized front field television signal (hereinafter referred to as "pre-field signal") F
It is an input terminal to which S is supplied.

Reference numeral 13 denotes the two television signals P mentioned above.
This is a motion vector detection circuit that detects a motion vector V in the moving image area based on S and FS. In this case, any method may be used as the motion vector detecting method, but as an example, for example, a television image is divided into blocks of m pixels × n lines (m and n are integers), There is a detection method for each block.

Incidentally, this method is disclosed in JP-A-55-1.
The pattern matching method described in JP-A-62683 and JP-A-55-162684 and the iterative gradient method described in JP-A-60-158786 are known.

Reference numeral 14 is a weighting factor α which defines the motion vector V detected by the motion vector detection circuit 13 by the distance between two input fields and one output field.
It is a motion vector correction circuit that performs correction based on Where α is, for example, the weighting factor of the previous field,
(1-α) is the weighting factor of the current field.

Reference numeral 15 is a motion correction memory for shifting the coordinates of the current field signal PS to the predicted position of the new field based on the motion vector (1-α) V corrected by the motion vector correction circuit 14. is there.

Similarly, 16 is a motion vector correction circuit 1
4 is a motion correction memory for shifting the coordinates of the previous field signal FS to the predicted position of the new field based on the motion vector αV corrected by 4.

Reference numeral 17 denotes a multiplication circuit which multiplies the current field signal PS, which has been motion-corrected by the motion correction memory 15, by a weighting factor (1-α). Similarly, reference numeral 18 is a multiplication circuit for multiplying the front field signal FS subjected to the motion correction by the motion correction memory 16 by the weighting factor α.

Reference numeral 19 is an addition circuit for adding the multiplication outputs of the multiplication circuits 17 and 18. Output S1 of this adder circuit 19
Is an interpolation signal generated by the motion-corrected television signals PS and FS (hereinafter referred to as "motion-compensation field interpolation signal").

Reference numeral 20 is a multiplication circuit for multiplying the current field signal PS supplied to the input terminal 11, that is, the current field signal PS which is not subjected to motion correction, by a weighting factor (1-α). Similarly, 21 is a multiplication circuit that multiplies the front field signal FS, which is not subjected to motion correction, by the weighting factor α.

Reference numeral 22 is an addition circuit for adding the multiplication outputs of the multiplication circuits 20 and 21. The output S2 of this adder circuit 22
Are the television signals PS and FS without motion compensation.
The interpolated signal generated by the above method, in other words, the interpolated signal generated by linear interpolation (hereinafter, referred to as “linear field interpolated signal”).

Reference numeral 23 is a multiplication circuit for multiplying the motion correction field signal S1 output from the addition circuit 19 by the adaptive switching coefficient β. Similarly, 24 is an adder circuit 22.
For the linear field interpolation signal S2 output from
It is a multiplication circuit that multiplies the adaptive switching coefficient (1-β).

Reference numeral 25 is an adder circuit for obtaining the true field interpolation signal S3 by adding the multiplication outputs of the multiplication circuits 23 and 24. 26 is the field interpolation signal S
3 is an output terminal supplied.

Reference numeral 27 is a subtraction circuit for obtaining the level difference between the television signals PS and FS output from the motion correction memories 15 and 16 in pixel units. 28 obtains the absolute value of the subtraction output of the subtraction circuit 27, and cumulatively adds it by the number of pixels in the motion vector detection block, so that the inter-field difference value signal (hereinafter, “motion correction field It is an absolute value conversion / accumulation circuit for obtaining S4).

Reference numeral 29 is a subtraction circuit for obtaining the level difference between the television signals PS and FS which are not subjected to motion correction in pixel units. The subtraction circuit 30 obtains the absolute value of the subtraction output of the subtraction circuit 29, and cumulatively adds it by the number of pixels in the motion vector detection block. This is an absolute value conversion / accumulation circuit that obtains S5).

Reference numeral 41 is an adaptive motion interpolation switching control circuit which, like the conventional control circuit 31, generates adaptive switching coefficients β, (1-β) based on the inter-field difference signals S4, S5.

Reference numeral 42 is a threshold value generating circuit for generating a predetermined threshold value Tα. Here, the threshold value Tα is set so that Tα <| V | max, where | V | max is the maximum value of the magnitude | V | of the motion vector V that can be detected.
Although not shown in the figure, the threshold value generation circuit 42 is configured so that the threshold value Tα can be appropriately adjusted by the user, for example.

Reference numeral 43 denotes the magnitude | V | of the motion vector V detected by the motion vector detection circuit 13 and the threshold generation circuit 4
This is a comparison circuit that determines whether or not | V | exceeds Tα by comparing the magnitude with the threshold value Tα output from 2

The judgment result of the comparison circuit 43 is supplied to the adaptive motion interpolation switching control circuit 41. When the control circuit 41 receives the determination result that | V | does not exceed Tα, it generates the adaptive switching coefficient β based on the inter-field difference signals S4 and S5 and receives the determination result that it has exceeded. In this case, this β is forcibly set to 0.

The operation of the above configuration will be described.

The field signals PS and FS supplied to the input terminals 11 and 12 are supplied to the motion vector detection circuit 13. By this, for each predetermined detection block,
The motion vector V of the moving image area is detected.

This motion vector V is supplied to the motion vector correction circuit 14 and is corrected based on the weight coefficient α.
Motion vector αV, (1-α) obtained by this correction
V is supplied to the motion correction memories 151 and 16.

As a result, the coordinates of the television signals PS and FS supplied to the input terminals 11 and 12 are α.
It will be shifted by V, (1-α) V. as a result,
This means that the coordinates of the moving image areas of the televisions PS and FS are matched.

The outputs read from the memories 16 and 17 are weighted by the multiplication circuits 17 and 18 and the addition circuit 19 based on the weighting coefficient α. As a result, the motion correction field interpolation signal S1 is obtained.

Similarly, the field signals PS and FS supplied to the input terminals 11 and 12 are weight-added by the multiplication circuits 20 and 21 and the addition circuit 22 based on the weight coefficient α. Thereby, the linear field interpolation signal S2 is obtained.

The field interpolated signals S1 and S2 thus obtained are weight-added by the multiplication circuits 23 and 24 and the addition circuit 25 based on the adaptive control coefficient β. As a result, the true field interpolation signal S3 is obtained.

The adaptive control coefficient β is generated based on the inter-field difference signals S4 and S5 when the magnitude | V | of the motion vector V is smaller than the threshold value Tα. Accordingly, in this case, based on the levels of the signals S4 and S5,
The selection amount of the field interpolation signals S1 and S2 is determined.

That is, if the level of the signal S4 is higher than the level of the signal S5, the motion correction field interpolation signal S
This means that the quality of 1 is worse than the quality of the uncorrected field interpolation signal S2. Therefore, in this case, the selection amount of the non-correction field interpolation signal S2 is increased.

On the contrary, if the level of the signal S4 is smaller than the level of the signal S5, it means that the quality of the motion compensation field interpolation signal S1 is better. Therefore, in this case, the selection amount of the motion correction field interpolation signal S1 is increased.

On the other hand, when | V | exceeds the threshold value Tα, the uncorrected field interpolation signal S2 is entirely selected regardless of the levels of the inter-field difference signals S4 and S5. As a result, even if the motion vector V suddenly changes behind the moving object, the true field interpolation signal S3 will not be distorted.

As described above in detail, in this embodiment, when the magnitude | V | of the motion vector V exceeds the threshold value Tα, only the non-correction field interpolation signal S2 is forcibly selected. Even if the motion vector V suddenly changes behind the moving object, the true field interpolation signal S3 can be prevented from being distorted.

In this embodiment, Tα <| V | ma
Since the threshold value Tα is set to be x, the comparison circuit 4
3 can be reduced.

That is, it is desirable to control the adaptive switching based on the field difference signals S4 and s5. For this purpose, it is necessary to set Tα = | V | max and widen the area where the adaptive switching is controlled based on the field difference signals S4 and s5.

However, when the magnitude | V | of the detected motion vector V approaches | V | max, the accuracy of the motion vector V decreases. Therefore, when Tα = | V | max, the possibility of erroneous determination is increased as a result of the determination by the motion vector V having poor accuracy.

On the other hand, in this embodiment, Tα <| V
Since | max is set, the determination can be performed in a region where the motion vector V is stable. This makes it possible to reduce erroneous determination, that is, erroneous operation of the comparison circuit 43.

Although one embodiment of the present invention has been described in detail above, the present invention is not limited to such an embodiment.

For example, in the above embodiment, the user sets the threshold T
The case where α can be adjusted has been described. However, in the present invention, the threshold value Tα may be fixed.

In the above embodiment, the case where the magnitude | V | of the motion vector V and the threshold value Tα are directly compared has been described. However, in the present invention, these may be decomposed into the vertical direction and the horizontal direction, and determination may be made for each direction.

Further, in the above embodiment, the case where the interpolation processing is performed for each detection block of the motion vector V has been described, but the interpolation processing may be performed for each pixel. However, in this case, it is necessary to consider an interpolation error due to noise.

Further, although the case where the present invention is applied to inter-field interpolation has been described in the above embodiment, the present invention is
It can also be applied to other interpolations in the time axis direction, for example, interframe interpolation.

Further, in the above embodiment, the case where the present invention is applied to the field conversion apparatus has been described, but the present invention can also be applied to other apparatuses. For example, in the high-efficiency coding method, it can be applied to the case where the field thinned out on the transmission side is reproduced on the reception side. It can also be applied to a motion compensation type device in a high-definition system.

In addition to the above, it goes without saying that the present invention can be variously modified and implemented without departing from the scope of the invention.

[0064]

As described above, according to the present invention,
PROBLEM TO BE SOLVED: To provide an interpolation signal generation device of a motion correction method capable of preventing distortion of an interpolation signal even when a moving object moving at high speed is panned under an uneven background. can do.

[Brief description of drawings]

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

FIG. 2 is a block diagram showing a configuration of a conventional device.

FIG. 3 is a diagram for explaining a conventional problem.

[Explanation of symbols]

11, 12 ... Input terminals, 13 ... Motion vector detection circuit,
14 ... Motion vector correction circuit, 15, 16 ... Motion correction memory, 17, 18, 20, 21, 23, 24 ... Multiplication circuit, 19, 22, 25 ... Addition circuit, 26 ... Output terminal, 2
7, 29 ... Subtraction circuit, 28, 30 ... Absolute value conversion / accumulation circuit, 41 ... Adaptive motion interpolation switching control circuit, 42 ... Threshold value generation circuit, 43 ... Comparison circuit.

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Hiroshi Hirabayashi 2-2-1 Jinnan, Shibuya-ku, Tokyo Within Japan Broadcasting Corporation (72) Gen Sonehara 2-2-1 Shinnan, Shibuya-ku, Tokyo Within the sending association

Claims (1)

[Claims] 1. A motion vector detecting means for detecting a motion vector of a moving image area from a television signal, and a correcting means for correcting the motion of the television signal by the motion vector detected by the motion vector detecting means. First interpolation signal generating means for generating a first interpolation signal using the tv signal which has been motion-corrected by the correcting means, and second interpolation using a television signal not corrected by the correcting means. Second interpolation signal generating means for generating a signal, interpolation signal switching means for adaptively switching the first and second interpolation signals based on the qualities of the first and second interpolation signals, and the vector detecting means for detecting. Switching control means for selecting only the second interpolation signal by the interpolation signal switching means when the magnitude of the motion vector exceeds a predetermined level. Inner interpolation signals generating device motion correction method characterized by comprising the.