JPH0630443A - Moving vector detecting/movement compensating device - Google Patents

Abstract

PURPOSE:To shorten a movement detecting time by improving the accuracy and compression efficiency of a moving vector of a chrominance signal by circuit size less than twice of convensional constitution. CONSTITUTION:A movement detector 1 respectively inputs a luminance signal and a chrominance signal in a current frame to a luminance signal moving vector forming part 10 and a chrominance signal moving vector forming part 11. The forming part 11 supplies a difference vector to a chrominance signal moving vector forming part 20A in a movement compensating part 2 and the forming part 10 supplies a luminance signal moving vector to the forming part 20A and a variation determining part 20C. The forming part 20A scales the luminance signal moving vector, synthesizes the scaled vector with a chrominance signal difference vector to form a chrominance signal moving vector and supplies the formed moving vector to the determining part 20C.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention, when performing motion compensation of a video signal, performs motion compensation of a video signal by obtaining a motion vector of a color signal by using a motion vector obtained by motion detection from a luminance signal. The present invention relates to a motion vector detection and / or motion compensation device suitable for use in a device.

[0002]

2. Description of the Related Art Generally, a motion vector of a video signal is created by pattern matching for each block based on current frame data and previous frame data. This block handles the luminance signal and the color signal included in the video signal as a two-dimensional group in which the luminance signal is 4 × 2 and the color signal is 1, respectively. The circuit has the configuration shown in FIG. The current frame data is supplied to the intra-block error calculation unit 40 and the position information output unit 41. Also,
The previous frame data is supplied to the intra-block error calculation section 40 via the search range variable delay 42. The in-block error calculation unit 40 supplies the error calculated in the block based on the supplied current frame data and previous frame data to the minimum value detection unit 43.

The position information output unit 41 supplies position information based on the current frame data to the read position changing unit 44. The read position changing unit 44 sends control signals for changing the read position to the search range variable delay 4 respectively.
2, the minimum value detection unit 43 and the vector creation unit 45 are supplied. The in-search range variable delay 42 delays and supplies the previous frame data in accordance with the input of the control signal, so that the read position of the previous frame data supplied to the intra-block error calculation unit 40 is set to the position of the current frame data. I am adjusting.

The minimum value detecting section 43 also detects the minimum value according to the control signal from the read position changing section 44 and outputs it to the vector creating section 45. The vector creation unit 45
The detection output of the minimum value is created and output as a motion vector under the control of the read position changing unit 44.

The video signal is composed of a luminance signal and a chrominance signal. When calculating the motion vector for each signal, motion vectors for the luminance signal and the color signal are created from the current frame data and the previous frame data of the luminance signal and the current frame data and the previous frame data of the color signal as shown in FIG. It These motion vectors are created by separate circuits.

Further, the motion vector of the color signal included in the video signal is the number of samples of the brightness signal and the color signal detected using the motion vector detected from the brightness signal included in the video signal using the motion vector detection device. It is calculated using the method of reducing the vector according to the ratio of. This method can be performed in a state in which the number of bits of the motion vector to be transmitted is small by scaling the value of each component of the motion vector obtained only from the luminance signal according to the ratio of the number of samples.

[0007]

However, although the former method described above can obtain optimum motion vectors by both methods, the circuit scale required for creating these motion vectors is large. For example, as shown in FIG. It can be seen that the circuit configuration shown in FIG. 9 requires twice the circuit configuration shown in FIG. 8 in comparison with the circuit configuration shown. Therefore, the detection time required for detecting the motion vector also takes a long time.

In the latter method, the motion vector for the color signal does not match the motion vector for the brightness signal because the number and position of the sampling points of the brightness signal and the color signal for each block are different, that is, for the color signal. Motion vector is not necessarily optimized, and the compression efficiency of the color signal is reduced.

Furthermore, when both the motion vector of the luminance signal and the motion vector of the chrominance signal are transmitted, the number of transmission bits for the motion vector increases, and the compression efficiency of the entire transmission / reception system that receives the transmitted motion vector is Not necessarily good.

Therefore, the present invention has been made in view of such a situation, and the compression efficiency is high even though the motion vector is highly accurate, without significantly increasing the circuit scale for detecting the motion state of the video signal. It is an object of the present invention to provide a motion vector detection and / or motion compensation device capable of transmitting / receiving high quality data.

[0011]

A motion vector detecting device according to the present invention is a motion vector detecting device for inputting a video signal including a luminance signal and a color signal and detecting a motion vector in the video signal. A motion vector indicating the motion state of the luminance signal is output, and the number of samples of the luminance signal included in the video signal is set to be larger than the number of samples of the color signal to sample the luminance signal and the color signal. The luminance signal motion detecting means for reducing and outputting the motion vector by the ratio of the number, and the sample number of the color signal is smaller than the reduced motion vector from the luminance signal motion detecting means and the sample number of the luminance signal. And a motion vector that minimizes the error of the color signal sampled by setting to By becoming a preparative solves the problems described above.

The motion vector of the color signal included in the video signal in the motion detecting apparatus is indicated by the vector of the difference from the reference pixel in the representative points set in the video signal, thereby solving the above-mentioned problem.

Further, the motion compensation device is created by using the motion vector obtained from the luminance signal and the motion vector with the number of samples of the color signal being smaller than the number of samples of the luminance signal included in the input video signal. A motion compensation device for receiving the motion vector of the color signal and compensating the motion of the video signal, the brightness signal predicting means for predicting the motion of the brightness signal by receiving the motion vector of the brightness signal; The above problem is solved by including a color signal predicting unit that receives the motion vector of the color signal and predicts the motion of the color signal.

In the above motion compensator, the motion vector of the color signal included in the video signal is indicated by the vector of the difference from the reference pixel in the representative points set in the video signal, thereby solving the above-mentioned problem.

The motion vector detecting and compensating device is a motion vector detecting device and a motion compensating device for compensating the motion by using a motion vector in which a motion state including a luminance signal and a chrominance signal is detected. A motion vector indicating the state of motion is output, and the number of samples of the luminance signal included in the video signal is set to be larger than the number of samples of the color signal to set the ratio of the number of samples of the luminance signal and the color signal. The luminance signal motion detecting means for reducing and outputting the motion vector and the reduced motion vector from the luminance signal motion detecting means are used to calculate the color signal of the color signal from the motion vector and the number of samples of the luminance signal. Set the number of samples to a small value and create a motion vector of the color signal by vector composition with the motion vector that minimizes the error of the color signal sampled. Video signal sending means having a color signal motion creating means, a brightness signal predicting means for receiving a brightness signal motion vector supplied from the video signal sending means and predicting the motion of the brightness signal, and the video signal. Color signal predicting means for predicting the motion of the color signal by receiving the motion vector for the color signal supplied from the sending means, and the motion of the video signal based on the outputs from the luminance signal predicting means and the color signal predicting means. The above-mentioned problem is solved by having a video signal receiving means including a motion compensating means for performing compensation.

In the motion vector detecting and motion compensating apparatus, the motion vector of the color signal included in the video signal is represented by the vector of the difference from the reference pixel in the representative points set in the video signal, whereby the above-mentioned problem is solved. To solve.

[0017]

In the motion vector detecting and / or motion compensating device of the present invention, the motion vector detecting device detects the luminance signal and the chrominance signal included in the input video signal, and calculates the motion vector of each signal detected by the video signal. When detecting, the number of samples of the color signal is smaller than the number of samples of the luminance signal included in the video signal, and the motion detecting unit for the luminance signal detects the motion state of the brightness signal and outputs the motion vector. Along with this, the motion vector is reduced by the ratio of the luminance signal and the color signal to output a motion vector, and the motion creating means for the color signal uses the motion vector from the motion detecting means for the brightness signal and the number of samples of the brightness signal. The motion vector of the color signal is obtained by vector-synthesizing with the motion vector that minimizes the error of the sampled color signal by setting the number of samples of the color signal to be small. To create a Le.

Further, in the motion vector detecting and / or motion compensating apparatus of the present invention, the motion compensating apparatus sets the number of samples of the chrominance signal smaller than the number of samples of the luminance signal included in the video signal to obtain the motion obtained from the luminance signal. When the vector and the motion vector of the color signal created by using the motion vector are received to compensate for the motion of the video signal, the motion vector of the brightness signal is received by the brightness signal prediction means and the motion vector of the brightness signal is received. The motion compensation means receives the motion vector of the color signal in the color signal prediction means to predict the motion of the color signal, and the motion compensation means in the motion compensation means based on the outputs from the luminance signal prediction means and the color signal prediction means. The motion compensation of the video signal is performed.

The motion vector detecting and compensating device, when compensating for the motion using the motion vector detecting the motion state including the luminance signal and the color signal, has the above-mentioned color rather than the number of samples of the luminance signal included in the video signal. The color signal generated by the motion vector detecting means for color signals using the output signal from the motion vector detecting means for the brightness signal, by detecting the motion state of the brightness signal as a motion vector with a small number of signal samples. And the motion vector of the brightness signal is output from the video signal sending means to the video signal receiving means, and the brightness signal predicting means of the video signal receiving means receives the brightness signal motion vector to detect the motion of the brightness signal. Predict,
Further, the color signal predicting means receives the color signal motion vector to predict the motion of the color signal, and based on the outputs from the luminance signal predicting means and the color signal predicting means, the motion signal compensating means is used to detect the video signal. We are doing motion compensation.

The motion vector detecting and / or motion compensating apparatus of the present invention indicates the motion vector of the color signal included in the video signal by indicating the vector of the difference from the reference pixel in the representative points set in the video signal as the color vector. Improves signal accuracy.

[0021]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Motion vector detection and / or
Alternatively, an embodiment of the motion compensation device will be described with reference to the drawings.

In the motion vector detecting and / or motion compensating apparatus according to the present invention, the motion vector detecting apparatus detects the motion vector in the video signal based on the luminance signal and the color signal included in the input video signal. is there.

The motion vector detecting device of the first embodiment shown in FIG. 1 outputs a motion vector indicating the motion state of the luminance signal, and at the same time determines the number of samples of the luminance signal included in the video signal as the color signal. And a luminance signal motion vector creating section 10 which is a luminance signal motion detecting means for reducing and outputting the motion vector according to the ratio of the number of samples of the luminance signal to the color signal. A motion vector that minimizes the error of the sampled color signal by setting the number of samples of the color signal to be smaller than the number of samples of the reduced brightness motion vector and the sample of the brightness signal from the luminance signal motion vector creation unit 10. And a color signal motion vector creating unit 11 which is a color signal motion creating means for creating a motion vector of the color signal by vector-synthesizing.

The configurations of the luminance signal motion vector generation unit 10 and the color signal motion vector generation unit 11 will be described. The current frame data and the previous frame data of the luminance signal included in the video signal are input to the luminance signal motion vector creation unit 10. These frame data are supplied to the intra-block error calculation unit 10a for each block. Here, the intra-block error calculation unit 10a treats the luminance signal and the color signal as a two-dimensional group in which the luminance signal is 4 × 2 and the color signal is 1, for example. The sample points of the luminance signal Y in the block are indicated by subscripts i and j to indicate the positions of the individual sample points.

The correspondence between the current frame data Y _ij and the previous frame data y _ij is pattern-matched to detect the difference. In order to perform this pattern matching, the intra-block error calculation unit 1 is included in the luminance signal motion vector creation unit 10.
The previous frame data y _ij adjusted by delaying the current frame data Y _ij supplied to 0a via the search range variable delay 11b is supplied to the intra-block error calculation unit 10a.

The current frame data Y _ij is supplied to the intra-block error calculation unit 10a and the position information output unit 10c. The previous frame data y _ij is the variable delay 1 within the search range.
It is supplied to the intra-block error calculation unit 10a via 0b. The in-block error calculation unit 10a calculates an error | Y _ij −y _ij | calculated for the sample points in the block based on the supplied current frame data Y _ij and previous frame data.
Is supplied to the minimum value detection unit 10d.

The position information output unit 10c supplies position information based on the current frame data to the read position changing unit 10e. The read position changing unit 10e outputs a control signal for changing the read position to the search range variable delay 10b, the minimum value detecting unit 10e, and the vector creating unit 10, respectively.
supply to f. The variable delay 10b within the exploration range is
By delaying the previous frame data y _ij according to the input of this control signal and supplying it to the intra-block error calculation unit 10a, the read position of the previous frame data y _ij in the intra-block error calculation unit 10a is set to the current frame data Y _ij. The alignment of is being adjusted. In-block error calculation unit 10a
Is to sample the current frame data and the previous frame data of the luminance signal Y at the position corresponding to the subscripts i and j, calculate the difference between both frame data at this sample point as an error, and supply it to the minimum value detection unit 10d. ing.

The minimum value detecting section 10d also detects the minimum value of the error in the block according to the control signal from the read position changing section 10e and outputs it to the vector creating section 10f. The vector creation unit 10f creates a motion vector in accordance with the control from the read position changing unit 10e and outputs the minimum value detection output. The motion vector of the luminance signal is also supplied to the scaling unit 10g.

The scaling unit 10g sets the ratio of the number of samples of the luminance signal and the color signal in advance. The motion vector of the supplied luminance signal is divided by the two-dimensional horizontal and vertical sample ratios, and the read position variable unit 11 provided in the color signal motion vector creation unit 11 divides the value.
is supplied to a.

The read position changing unit 11a supplies signals for changing the read position to the minimum value detecting unit 11b, the search range variable delay 11c and the vector creating unit 11d, as in the case of the luminance signal. The in-search range variable delay 11c delays the previous frame data c _ij of the color signal according to the input of the control signal from the read position changing section 11a and supplies the delayed frame data c _ij to the in-block error calculating section 11e, thereby calculating the in-block error. The read position of the previous frame data c _ij in the section 11e and the current frame data C _ij are aligned. In-block error calculation unit 11e
At the positions corresponding to the subscripts i and j, the current frame data and the previous frame data of the color signal C are sampled, the difference at this sample point is calculated as an error, and the error is supplied to the minimum value detection unit 11b.

The minimum value detecting section 11b also detects the minimum value of the error in the block in accordance with the control signal from the read position changing section 11a and outputs it to the difference vector generating section 11A in the vector generating section 11d. . This difference vector generation unit 1
1A reads out the detection output of the above-mentioned minimum value, and reads out the position varying unit 11
In accordance with the control from a, a motion vector (referred to as a difference vector) relating to the difference between the color signals is created, and the vector synthesis unit 11B
Is being supplied to. The vector synthesis unit 11B vector-synthesizes the output obtained by scaling the motion vector of the luminance signal and the difference vector, and outputs the result.

The operation of this motion vector detecting device will be described with reference to the schematic diagram shown in FIG. Where ○
The mark represents the sample point of the luminance signal. The cross marks represent the sampling points of the color signal. Further, a box 1 shown by a broken line represents a block of the current frame of the luminance signal, and a box 2 shown by a solid line shows a case where the block obtained by predicting the motion matches the position of the block of the previous frame. Shows. In one block shown in FIG. 2, the luminance signal is composed of 8 dot sample points arranged in 2 rows and 4 columns. On the other hand, the color signal is represented by a 1-dot sample point.

At this time, when the vector from the current frame to the corresponding position of the previous frame is represented as a motion vector, the motion vector of this luminance signal is (3, 1).
That is, the prediction signal in the motion vector detecting device uses the previous frame data at the position represented by this motion vector from the current frame.

Assuming that the predicted signal for the color signal (a) in box 1 is at the position of the color signal (b) in the previous frame, the motion vector of the color signal can be expressed as (1,0). An optimum motion vector can be obtained by separately detecting the motion of the luminance signal and the color signal as described above.
In the circuit shown in (1), the circuit scale is almost doubled, and the detection time is inevitably prolonged. Moreover, the number of bits used to transmit the luminance signal and the color signal increases.

Therefore, a method is generally used in which only the motion vector of the luminance signal is transmitted, the motion vector relating to the color signal is calculated from this luminance signal, and this motion vector is used. In this conventional method, as shown in FIG. 4A described later, the scaling unit 12 scales the motion vector of the luminance signal Y by the ratio of the number of samples of the luminance signal Y and the color signal C to obtain the motion vector of the color signal. Is a method of generating.

A detailed description will be given using the above-described motion vector and the numerical value of the number of samples. The motion vector of the luminance signal is (3,1). The dot configuration of the box is such that the luminance signal is 4 dots in the horizontal direction, 2 dots in the vertical direction, and the color signal is 1 dot. Actually, the ratio Y _S / C _S of the number of samples of the luminance signal to the color signal is 4/1 = 4 in the horizontal direction and 2/1 = 2 in the vertical direction, respectively. When scaling is performed by dividing each corresponding component of the motion vector (3, 1) of the luminance signal by the ratio of the number of samples, the scaling results in the horizontal direction and the vertical direction are 3/4 = 0.75 (1) 1 /2=0.5 (2). The method of the circuit that further processes the data of the motion vector reduced by the scaling result is rounded up,
Or, if you adopt the method of processing signals by rounding off,
The motion vector of the color signal is (1,1), and (0,0) is obtained by the truncation method.

Here, the data corresponding to the color signal (a) in FIG. 2 depends on the method, depending on the motion vector of the color signal, but the position or color of the color signal (a) in the previous frame shown in FIG. Only the position of signal (c) will be chosen. However, the optimum data for the color signal is likely to be the color signal (b) or the color signal (d) in FIG. Therefore, in order to obtain a highly accurate motion vector of the color signal, the scaling unit 10g shown in FIG. 1 reduces the motion vector of the brightness signal by the number of samples of the brightness signal and the color signal. The result of this processing is truncated to obtain the basic vector. When the above numerical values are used, the basic vector is represented by (0,0).

Next, the positions (a), (b), (c) and (d) of the color signal shown in FIG.
It will be represented by two difference vectors (0,0), (1,0), (1,1) and (0,1). At this time, the intra-block error calculation unit 11e calculates the intra-block prediction error from the previous frame data and the current frame data of the color signal input via the in-search range variable delay 11c according to the control from the read position variable unit 11a. To do. The minimum value detection unit 11b compares the errors in the supplied blocks, detects the minimum error, and supplies the data indicating the minimum difference vector to the difference vector generation unit 11A. The difference vector generation unit 11A supplies the minimum difference vector to the vector synthesis unit 11B.
Further, the vector synthesis unit 11B includes a scaling unit 10g.
The fundamental vector from is also supplied. Vector synthesizer 11
B is a motion vector of the color signal by combining the basic vector and the difference vector.

With this configuration, by adding the motion vector having the smallest prediction error to the original basic vector, the motion vector of the color signal with high accuracy can be detected without significantly increasing the circuit scale. Will be able to. As a result, it is possible to prevent the compression efficiency of the color signal from decreasing.

Although the first quadrant in which the reference vector (0, 0) of the color signal is the origin is +1 has been described, it is also possible to set and obtain the difference vector including the minus direction. Also, the range in which the difference vector is set is not limited.

Next, a second embodiment of the motion vector detecting device of the present invention will be described with reference to the block diagram shown in FIG. Here, common parts are given the same reference numerals, and description thereof is omitted. In this embodiment, the difference vector is output as the motion vector of the color signal without performing the vector synthesis in the first embodiment. Here, in the conventional motion vector transmission, 9 bits of the luminance signal Y shown in FIG. 4A and 6 bits of the color signal C obtained by scaling the signal after scaling by the scaling unit 12 are respectively used as the luminance signal motion vector and the color signal. It is necessary to transmit a total of 15 bits as a signal motion vector.

Further, in the case of the first embodiment, the motion vector detecting device scales the brightness signal Y together with 9 bits of the motion vector of the brightness signal Y by the scaling section 12 as shown in FIG. 4B. It is necessary to transmit a total of 17 bits as a motion vector of the color signal C by combining the 6 bits and the 2 bits of the difference vector for the color signal C by the vector combining unit 13. If it is transmitted as it is, the compression efficiency is not good because the number of bits required to transmit the motion vector is increased even if the compression efficiency of the color signal is improved.

Therefore, in the circuit configuration of the second embodiment shown in FIG. 3, instead of transmitting the motion vector of the luminance signal Y and the color signal C in the case of the first embodiment, the motion vector of the luminance signal Y is transmitted. And the difference vector of the color signal is transmitted. As for the motion vector of the color signal, the reference vector (or the representative vector) generated by scaling and the difference vector are vector-synthesized by the vector synthesizing unit 11B. The transmission of the difference vector becomes easy because the reference vector can be easily obtained by scaling, and it is sufficient to supply only the difference vector having a small number of bits corresponding to the motion with respect to the reference vector. Also, the variable delay 11c within the exploration range
Is variably delayed with respect to the previous frame data of the color signal C by the output from the vector synthesizing unit 11B.

In fact, in the above-described embodiment, the difference vector can be represented by 4 bits, and therefore can be represented by 2 bits. Therefore, in this case, the number of bits to be transmitted is 9 bits of the motion vector of the luminance signal Y and 2 bits of the difference vector of the color signal, which is a total of 11 bits, which is significantly smaller than the 17 bits of the transmission bit number. can do.

As described above, the motion vector of the color signal is created by synthesizing the reference vector and the difference vector calculated from the luminance signal Y, and only the difference vector is transmitted for the color signal at the time of transmission, so that the luminance signal Y is obtained. Together with the color signal C,
A highly accurate motion vector can be transmitted with a small number of required bits, and efficient compression is possible as a whole.

Also in this case, the direction and magnitude of the difference vector can be set arbitrarily.

Next, the motion compensation apparatus of the present invention will be described with reference to FIG.
Will be described with reference to. The motion compensating apparatus sets the number of samples of the color signal smaller than the number of samples of the luminance signal included in the input video signal, and the motion vector obtained from the luminance signal and the motion of the color signal created by using the motion vector. The vector is received and the motion of the video signal is compensated.

The motion compensation apparatus shown in FIG. 5 includes a variable amount determining unit 20c which is a luminance signal predicting means for predicting the movement of the luminance signal by receiving the movement vector of the luminance signal Y, and the motion vector of the color signal. And a color signal predicting section 20 which is a color signal predicting means for predicting the movement of the color signal C. The color signal predicting section 20 receives the A motion compensation signal for frame data is output.

The color signal predicting section 20 comprises a scaling section 20a, an adder 20b and a variable amount determining section 20c.

In the color signal predicting section 20, the motion vector of the luminance signal Y is supplied to the scaling section 20c, and the output data concerning the reduced color signal motion vector is supplied from the scaling section 20c to one end of the adder 20b. ing. The difference vector of the color signals is supplied to the other end of the adder 20b. The adder 20b synthesizes these signals and supplies the motion vector of the color signal to the variable amount determination unit 20c.

The variable amount determining unit 20c receives the motion vector of the luminance signal Y, supplies the motion vector of the color signal, and supplies the motion vector of the color signal to the in-search range variable buffer 21 according to these data. Control the data.

By configuring the motion compensator in this way, it is possible to perform transmission of a motion vector with higher accuracy than the transmission of only the luminance signal by adding several bits per vector. In addition, the overall compression efficiency can be increased in motion compensation.

Further, the motion vector detecting and motion compensating apparatus of the present invention will be described with reference to FIG. Here, this motion vector detection and motion compensation apparatus will be described based on an example applied to a motion compensation predictive encoder.

The motion vector detecting and motion compensating apparatus is applied to compensate the motion by using the motion vector in which the motion state including the luminance signal and the color signal is detected. The motion vector detection device and the motion compensation device are
A motion vector indicating the motion state of the luminance signal Y is output, and the number of samples of the luminance signal Y included in the video signal is set to be larger than the number of samples of the color signal C so that the luminance signal Y and the luminance signal Y are included. A luminance signal motion vector creating section 10 which is a luminance signal motion detecting means for reducing and outputting the motion vector in accordance with the ratio of the number of samples to the color signal C, and a reduction from the luminance signal motion vector creating section 10. Using the motion vector, this motion vector and the motion vector that sets the number of samples of the color signal C smaller than the number of samples of the luminance signal Y to minimize the error of the sampled color signal C are vector-combined. And a motion detection unit 1 having a color signal motion vector creation unit 11 that is a color signal motion creation unit that creates a motion vector of the color signal C, and a video signal transmission unit. A motion prediction coding unit, a variable amount determination unit 20c that is a brightness signal prediction unit that receives a motion vector for a brightness signal supplied from the motion prediction coding unit, and predicts the motion of the brightness signal, and the motion prediction code. From the color signal predicting unit 20 that is a color signal predicting unit that predicts the motion of the color signal by receiving the motion vector for the color signal supplied from the converting unit, the variable amount determining unit 20c, and the color signal predicting unit 20. It comprises a motion compensating unit 2 which is a motion compensating unit for compensating the motion of the video signal based on the output, and a motion decoding unit which is a video signal receiving unit.

The color signal predicting unit 20 of the motion compensating unit 2 is variable with the color signal motion vector generating unit 20A which creates the color signal motion vector C from the color signal difference vector and the luminance signal Y motion vector as described above. The quantity determination unit 20c is used.

The motion compensation predictive encoder shown in FIG. 6 is an encoder on the video signal transmitting side. This motion compensation predictive encoder supplies the input luminance signal Y and color signal C to one end of the motion detector 1 and the subtractor 30. Subtractor 3
0 supplies a motion compensation signal for compensating the shift due to the motion between frames to the other end side. The subtractor 30 performs subtraction processing so as to suppress the value of the difference output with respect to the input signal, and outputs it to the discrete cosine transform (DCT) unit 31. The discrete cosine transform (DCT) unit 31 performs orthogonal transform on the supplied signal and supplies it to the quantizer 32. The quantizer 32
The transformed data is quantized and the discrete cosine coefficient is output. This discrete cosine coefficient is supplied to a circuit for Huffman coding or Reed-Solomon code. The discrete cosine coefficient is supplied to the inverse discrete cosine transform (IDCT) unit 34 via the inverse quantizer 33 as data for motion compensation of the next luminance signal and chrominance signal. The inverse discrete cosine transform (IDCT) unit 34 converts the original signal and supplies it to one end of the adder 35. The output from the motion compensation unit 2 is input to the other end of the adder 35. By this addition, the motion compensation component subtracted by the subtractor 30 is added to become the same signal as the luminance signal and the color signal input to the subtractor 30. This signal contains an error amount of the quantizer 32 and the inverse quantizer 33, is supplied to the frame memory 36, and is temporarily held. The data output from the frame memory 36 is sent to the motion compensation unit 2 and the motion detection unit 1. This data is supplied as the data of the previous frame supplied to the motion compensation predictive encoder.

As described above, the motion detecting section 1 inputs the luminance signal and the color signal of the current frame to the luminance signal motion vector creating section 10 and the color signal motion vector creating section 11, respectively. The color signal motion vector creation unit 11
The difference vector is supplied to the color signal motion vector generation unit 20A in the motion compensation unit 2. In addition, the motion vector of the brightness signal from the brightness signal motion vector creation unit 10 includes the color signal motion vector generation unit 20A and the variable amount determination unit 20.
is being supplied to c. Color signal motion vector generation unit 20A
In the above, scaling is performed to reduce the motion vector of the luminance signal according to the ratio of the number of samplings, and the motion vector of the color signal is generated by combining with the difference vector of the color signal and is supplied to the variable amount determination unit 20c.

The variable amount determining unit 20c supplies a control signal for motion compensation to the in-search range variable buffer 21 according to the motion vector of the luminance signal and the motion vector of the color signal. The variable buffer 21 within the search range is the frame memory 3
The data of the previous frame from 6 is compensated for the motion, and the data of the previous frame is supplied to the other end of the subtractor 30 and the other end of the adder 35. In this way, the data of the previous frame is made to act on the data of the current frame. In this way, the motion compensation predictive encoder outputs the discrete cosine coefficient to perform predictive encoding.

The block circuit shown in FIG. 7 shows a motion compensation predictive decoder which is a decoding unit for the motion compensation predictive encoder. This motion compensation predictive decoder is a discrete cosine transform (DCT) from the motion compensation predictive encoder side.
The coefficient, the motion vector of the luminance signal Y, and the difference vector of the color signal C are input. The discrete cosine transform coefficient is subjected to the inverse discrete cosine transform (IDCT) through the inverse quantizer 33.
It is supplied to the section 34. This inverse discrete cosine transform (ID
The (CT) unit 34 supplies the output signal to one end of the adder 35. The motion vector of the luminance signal Y and the difference vector of the color signal C are supplied to the motion compensation unit 2.
Here, the luminance signal Y is input to the variable amount determination unit 20c,
The difference vector between the luminance signal Y and the color signal is supplied to the color signal motion vector generation unit 20A. The operation of the motion compensating unit 2 is similar to that described above.
At 0A, the motion vector of the color signal is generated by combining the reference vector and the difference vector. The motion vector of the generated color signal is supplied to the variable amount determination unit 20c. The variable amount determination unit 20c outputs the variable amount to the search range variable buffer 21. The previous frame data from the frame memory 36 is supplied to the other end side of the adder 35 after the variable amount is changed by the search range variable buffer 21. The adder 35 uses an inverse discrete cosine transform (IDC
T) output from the unit 34 and the variable buffer 2 within the search range
Luminance signal Y decoded by adding and synthesizing outputs from 1
And the color signal C is output. With this configuration, motion compensation of the decoder is performed.

With this configuration, motion vectors of both the luminance signal and the chrominance signal can be obtained without doubling the circuit scale, and in particular, the motion vector of the chrominance signal simply calculated. It becomes possible to supply the motion vector of the color signal with higher accuracy than the accuracy. This improves the compression efficiency of the color signal. Further, with this configuration, the accuracy can be maintained even if the search range of the motion vector of the color signal is narrowed, so that the calculation amount of the prediction error can be reduced and the time required for the predictive coding and the predictive decoding can be shortened. You can also plan.

Also in this case, the direction and magnitude of the difference vector can be set arbitrarily.

[0062]

As is apparent from the above description, according to the motion vector detecting apparatus of the present invention, a motion vector for inputting a video signal including a luminance signal and a chrominance signal and detecting a motion vector in the video signal. A detecting device, which outputs a motion vector indicating a motion state of the luminance signal, sets the number of samples of the luminance signal included in the video signal to be larger than the number of samples of the color signal, and outputs the luminance signal. And a luminance signal motion detecting means for reducing and outputting the motion vector at a ratio of the number of samples of the color signal, and the reduced motion vector from the luminance signal motion detecting means and the number of samples of the luminance signal. Create a motion vector of the color signal by vector-synthesizing with the motion vector that minimizes the error of the sampled color signal by setting the number of samples of the color signal to be small. By becoming a motion creating means color signal, without increasing the circuit scale significantly, it becomes possible to detect the motion vector with high accuracy a color signal. As a result, it is possible to prevent the compression efficiency of the color signal from decreasing. By transmitting only the difference vector for the color signal at the time of transmission, it is possible to transmit the motion vector with high accuracy with a small number of required bits together with the luminance signal Y and the color signal C, and achieve efficient compression as a whole. It can be carried out.

Further, according to the motion compensation device, the number of samples of the color signal is smaller than the number of samples of the luminance signal included in the input video signal, and the motion vector obtained from the luminance signal and the motion vector are used. A motion compensation device that receives the motion vector of the created color signal and compensates the motion of the video signal, and a luminance signal prediction unit that receives the motion vector of the luminance signal and predicts the movement of the luminance signal. , A color signal predicting means for receiving the motion vector of the color signal and predicting the motion of the color signal is provided, so that the transmission of the motion vector having higher accuracy than the transmission of only the luminance signal is performed by one vector. This can be done by adding a few bits per time. In addition, the overall compression efficiency can be increased in motion compensation.

Further, according to the motion vector detecting and compensating device, the motion vector detecting device and the motion compensating device for compensating the motion by using the motion vector detecting the motion state including the luminance signal and the chrominance signal, While outputting a motion vector indicating the motion state of the luminance signal, the number of samples of the luminance signal included in the video signal is set to be larger than the number of samples of the color signal, and the luminance signal and the color signal A luminance signal motion detection means for reducing and outputting the motion vector according to the ratio of the number of samples, and a reduced motion vector from the luminance signal motion detection means are used to determine the motion vector and the number of samples of the luminance signal. The number of samples of the color signal is set to a small value and the motion vector that minimizes the error of the sampled color signal is vector-synthesized to generate the motion of the color signal. A video signal sending means having a color signal motion creating means for creating a vector; and a luminance signal predicting means for receiving the luminance signal motion vector supplied from the video signal sending means and predicting the movement of the luminance signal. A color signal predicting means for receiving a color signal motion vector supplied from the video signal sending means to predict the motion of the color signal, and the video based on the outputs from the luminance signal predicting means and the color signal predicting means. By including the video signal receiving means including the motion compensating means for performing the motion compensation of the signal, it is possible to obtain the motion vectors of both the luminance signal and the chrominance signal without doubling the circuit scale. , It becomes possible to supply the motion vector of the color signal with higher accuracy than the accuracy of the motion vector simply calculated for the color signal. This improves the compression efficiency of the color signal. Further, with this configuration, the accuracy can be maintained even if the search range of the motion vector of the color signal is narrowed, so that the calculation amount of the prediction error can be reduced and the time required for the predictive coding and the predictive decoding can be shortened. You can also plan.

[Brief description of drawings]

FIG. 1 is a diagram showing a schematic block configuration of a first embodiment of a motion vector detection device of the present invention.

FIG. 2 is a schematic diagram for explaining the operation of the motion vector detection device shown in FIG.

FIG. 3 is a diagram showing a schematic block configuration of a second embodiment of the motion vector detection device of the present invention.

FIG. 4 is a schematic diagram illustrating the relationship between the number of transmission bits in transmission of a motion vector.

FIG. 5 is a diagram showing a schematic block configuration of a motion compensation device of the present invention.

[Fig. 6] Fig. 6 is a diagram showing a schematic block configuration in the case where the motion vector detection and motion compensation device of the present invention is applied to a motion compensation predictive encoder.

FIG. 7 is a diagram showing a schematic block configuration in the case where the motion vector detection and motion compensation device of the present invention is applied to a motion compensation predictive decoder.

FIG. 8 is a block diagram showing a schematic configuration of a conventional motion detection device.

FIG. 9 is a diagram showing a schematic circuit showing a configuration for obtaining a luminance signal and a color signal in a conventional motion detection device.

[Explanation of symbols]

10 ..... Luminance signal motion vector creation unit 11 ..... Color signal motion vector creation unit 10a, 11e ..... In-block error calculation unit 10b, 11c, 21 ... Variable delay within search range 10c ..... Position information output unit 10d, 11b ..... Minimum value detection unit 10e , 11a ..... Reading position changing section 10f, 11d ..... Vector creating section 10g, 20a ..... Scaling section 11A .. Difference vector generation unit 11B ..... Vector synthesis unit 20 .................. Color signal prediction unit 20b .................. Adder 20c ・ ・ ・ ・ ・ ・ ・ ・ ・ ・ Variable amount determination unit

Claims (6)

[Claims] 1. A motion vector detection device for inputting a video signal including a luminance signal and a color signal and detecting a motion vector in the video signal, the motion vector detecting device outputting a motion vector indicating a motion state of the luminance signal. , The luminance that is set by setting the number of samples of the luminance signal included in the video signal to be larger than the number of samples of the color signal and reducing the motion vector by the ratio of the number of samples of the luminance signal and the color signal The signal motion detection means, and the error of the sampled color signal is set to a minimum by setting the sample number of the color signal smaller than the sample number of the reduced motion vector and the brightness signal from the luminance signal motion detection means. And a motion vector detecting means for creating a motion vector of the color signal by vector-synthesizing the motion vector with the motion vector. 2. The motion detecting device according to claim 1, wherein the motion vector of the color signal included in the video signal is represented by a vector of a difference from a reference pixel in a representative point set in the video signal. 3. A motion vector obtained from the luminance signal and a motion of the color signal generated by using the motion vector, with the number of samples of the color signal being smaller than the number of samples of the luminance signal included in the input video signal. A motion compensation device that receives a vector and a motion of a video signal, and a motion compensation device that receives a motion vector of the brightness signal and predicts a motion of the brightness signal, and a motion vector of the color signal. And a motion compensating unit for compensating the motion of the video signal based on outputs from the luminance signal predicting unit and the color signal predicting unit. A motion compensation device characterized by the following. 4. The motion compensation apparatus according to claim 3, wherein the motion vector of the color signal included in the video signal is represented by a vector of a difference from a reference pixel in a representative point set in the video signal. 5. A motion vector detecting device and a motion compensating device for compensating for a motion using a motion vector detecting a motion condition including a luminance signal and a chrominance signal, the motion vector indicating a motion condition of the luminance signal. And the number of samples of the luminance signal included in the video signal is set larger than the number of samples of the color signal to reduce the motion vector by the ratio of the number of samples of the luminance signal and the color signal. By using the luminance signal motion detecting means for outputting as a result and the reduced motion vector from the luminance signal motion detecting means, the number of samples of the color signal is set smaller than the number of samples of the motion vector and the luminance signal. And a color signal motion creating means for creating a motion vector of the color signal by vector-synthesizing the sampled motion vector with the motion vector that minimizes the error of the color signal. Video signal sending means, brightness signal predicting means for predicting the motion of the brightness signal by receiving the motion vector for the brightness signal supplied from the video signal sending means, and motion for the color signal supplied from the video signal sending means A video comprising a color signal predicting means for receiving a vector to predict the motion of a color signal, and a motion compensating means for compensating the motion of the video signal based on the output from the luminance signal predicting means and the color signal predicting means. A motion vector detecting and compensating device, comprising: a signal receiving means, 6. The motion vector detecting and compensating apparatus according to claim 5, wherein the motion vector of the color signal included in the video signal is represented by a vector of a difference from a reference pixel at a representative point set in the video signal. .