JPH1032822A - Motion vector detector - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a motion vector detector which can contrive effectiveness of enhancement effect for detection precision brought about by using plural candidate vectors and can enhance prediction effect between frames for a color difference signal. SOLUTION: A motion vector detector 100 is provided with a luminance signal preprocessing part 122 which performs down sampling of an input luminance signal in the case of a system where only one motion vector is sent for one block, in motion vector detection which uses a multi-step method for prediction between motion compensation frames, and in a motion vector detection part 124, a luminance signal and a color difference signal are evaluated with the same evaluating means by a similarity calculating part and a candidate vector calculating part, similarity is searched as an evaluating value, and plural candidate vectors are selected by re-evaluating similarity between blocks by using each similarity.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a motion vector detecting device for detecting a motion compensation vector in moving image compression, and more particularly to a motion vector detecting device using block matching.

[0002]

2. Description of the Related Art JPEG (Jo
int Photograghic Expert Group) or MPEG (Moving
Picture Expert Group).

[0003] MPEG is MPEGI, MPEGII, M
Three levels of PEGIV standards are being considered. MPE
The purpose of GI is to compress moving images that can be transmitted through a 1.5 Mbps communication line, and is considered to be mainly used for videophones, videoconferencing, and the like. In MPEGI, the current NTSC video image is 320 × 240
Treat as pixel resolution and make up one frame 2
Data of only one of the fields is used.
MPEG II aims at compression that can be transmitted over a communication line exceeding 10 Mbps, and targets moving picture transmission by ISDN or the like and digital video. MPEG IV is intended for low bit rates.

The feature of MPEG is that DCT (Discrete Cos
ine Transform (discrete cosine transform) is to perform inter-frame prediction processing for compression in the time axis direction in addition to still image compression, but the prerequisites for moving image compression include random access to frames and fast forward. Playback and rewind playback (reverse direction). Therefore, inter-frame prediction in MPEG employs both forward and backward directions. MPE
Even for G, MC (motion compensation) + DCT is basically used. The block size for motion compensation is 16 × 16
(However, there is also an 8 × 8 mode), DCT is performed on 8 × 8 blocks. This motion compensation is performed with half-pixel accuracy. Motion compensation with half-pixel accuracy is performed not only by examining a position shifted in pixel units on a reference frame used for prediction, but also by generating a position between pixels by interpolation and performing matching.

[0005] In a moving image compression apparatus that involves prediction in the time direction, prediction is performed by motion compensation in order to reduce a reduction in prediction efficiency due to movement of a camera PAN or a subject. This motion compensation detects a motion vector of a target region between a frame of interest (encoding target frame) and a reference frame (for example, a previous frame), and sets a position shifted by the motion vector in the reference frame as a reference pixel. This is a method of transmitting a difference (prediction error) from a target pixel as a prediction value. For example, in motion compensation prediction, the motion of a moving object is predicted based on the motion vector of the prediction source image, and the motion is compensated for in the original image. The motion compensation is a method of searching for a place having the smallest difference in the vicinity of the position of the block in the previous image in units of 16 × 16 pixels, and taking the difference therefrom, thereby reducing the data to be sent further. Yes, as a means for detecting a motion vector, generally, a search is performed within a certain range from the original position of a partial image to be subjected to motion compensation, and the error is minimized (ie,
A place where the degree of approximation is the highest) is detected, and this is used as a prediction signal.

[0006] In addition, a conventional moving picture compression apparatus (CCITT H.261 or MPEG.
eo), when encoding the generated motion vector, the difference between the generated motion vector and the motion vector of a nearby partial image (usually, the partial image processed immediately before) is calculated, and only the difference is encoded. Is becoming

[0007] In conventional moving image compression apparatus of this kind, M
As a C (motion compensation) detection method, see, for example,
There is one disclosed in JP-A-189297.

In the motion vector detection method described in the above publication, in order to accurately detect the motion vector of the color difference signal, the position of the motion vector obtained from the luminance signal is corrected for the color difference signal, and the corrected vector is used as a reference vector. A motion vector of a color difference signal is detected from the reference vector and a plurality of candidate vectors at peripheral positions.

FIG. 5 is a block diagram showing a motion vector detecting method described in the above publication.

In FIG. 5, a motion vector detecting device 10
Represents a current frame memory 11, a previous frame memory 12, a frame delay unit 13, a luminance signal motion vector detection unit 14,
It is composed of a color difference signal motion vector detection unit 15, a reference vector calculation unit 16, and a control unit 17. 18 is an input terminal,
19 is an output terminal.

The input terminal 18 receives the image data of the current frame. The current frame memory 11 stores the image data of the current frame, and the frame delay unit 13 stores the current frame image data delayed by one frame in the previous frame memory 12.

In order to detect a motion vector, the control unit 1
In accordance with the instruction from 7, the necessary luminance data of the macro block of the current frame and the luminance data of the reference block of the previous frame are read from the current frame memory 11 and the previous frame memory 12. The read data of both blocks are input to the luminance signal motion vector detection unit 14, and the luminance signal motion vector detection unit 14 obtains the motion vector of the luminance signal by an all point search method using block matching or the like.

The motion vector of the luminance signal calculated by the luminance signal motion vector detecting section 14 is input to a reference vector calculating section 16 to calculate a reference vector necessary for detecting a motion vector of a color difference signal.

On the basis of this reference vector, the control unit 17 obtains the chrominance signal pixel data of the macroblock of the current frame and the reference block of the previous frame necessary for detecting the motion vector of the chrominance signal from the current frame memory 11 and the previous frame memory 12. Is read. The color difference signal motion vector detection unit 15 detects a color difference signal motion vector based on the pixel data of the read color difference signal, and outputs a color difference signal motion vector from the output terminal 19.

[0015]

However, such a conventional motion vector detecting device has the following problems.

That is, in motion vector detection used for inter-frame predictive coding in moving picture coding, a method is used in which a motion vector of a luminance signal is obtained, and a vector obtained by correcting the position of the vector is used as a motion vector of a color difference signal. H. In a standard such as H.261 or $ PEG, only a motion vector of a luminance signal is transmitted. In a system that can only transmit a motion vector of a luminance signal, when processing a vector obtained by correcting the position of a motion vector of a luminance signal as a motion vector of a color difference signal,
There is a problem that the effect of the inter-frame prediction on the color difference signal is hard to appear.

In the case where a motion vector is detected in multiple stages, when an erroneous detection is performed in the initial stage, the problem that the error propagates affects not only the luminance signal but also the color difference signal. There is a defect.

An object of the present invention is to provide a motion vector detecting device which can improve the detection accuracy by using a plurality of candidate vectors, and can improve the inter-frame prediction effect on a color difference signal. And

[0019]

A motion vector detecting apparatus according to the present invention divides a frame to be coded into predetermined blocks, and detects the position where the difference is smallest among the preceding frames for each block. In the motion vector detecting device, processing means for down-sampling the input luminance signal, evaluation value calculating means for performing the same evaluation of the luminance signal and the color difference signal by the processing means to calculate an evaluation value, and calculating the similarity as the evaluation value, Vector selecting means for re-evaluating the similarity between blocks using each similarity and selecting a plurality of candidate vectors.

Further, the processing means may process the pixel data of the luminance signal among the pixel data so that the pixel data has the same number of pixels as the color difference signal and pixels at the same spatial position.

The motion vector detecting device according to the present invention divides a frame to be encoded into predetermined blocks, and detects a position where the difference is smallest among the preceding frames for each block. A luminance signal similarity calculating means for calculating a similarity evaluation value from the luminance signal; a color difference signal similarity calculating means for calculating a similarity evaluation value from the color difference signal; and a similarity evaluation and a color difference signal obtained from the luminance signal. A vector selecting means for selecting a plurality of candidate vectors from the obtained similarity evaluation is provided.

In the motion vector detecting device, the motion vector detection may be a motion vector detection using a multi-step method in motion compensated inter-frame prediction. Even in the motion vector detection using the Full Pel + Half Pel method, a motion vector with pixel accuracy is detected in the first step, and a motion vector at a half pixel position is obtained based on the candidate vector detected in the first step in the second step. Good.

Further, the motion vector detecting device may be a motion vector detecting device that can send only one motion vector per block.

[0024]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The motion vector detecting device according to the present invention can be applied to a motion vector detecting device such as a moving image storage device using a motion detection prediction signal of a moving image.

FIG. 1 is a block diagram showing the overall processing procedure of the motion vector detecting device according to the first embodiment of the present invention.

In the first embodiment, the processing is shown in a case where the method is adapted to a method using a plurality of candidate vectors in a Full Pel + Half Pel system, which is a system for detecting a motion vector in a multi-stage process. What is Full Pel + Half Pel method?
First, the method used in the PEGII Test Model
In this method, a motion vector with pixel accuracy is detected in the first step, and a motion vector at a half-pixel position is obtained based on the candidate vector detected in the first step in the second step.

In FIG. 1, a motion vector detecting device 10
0 is the current frame memory 111 and the previous frame memory 11
2. Frame delay unit 121, luminance signal preprocessing unit 122
(Processing means), control unit 123 and motion vector detection unit 1
24. 101 is an input terminal, 102
Is an output terminal. In FIG. 1, a chain arrow indicates a data flow of a luminance signal, and a broken arrow indicates a data flow of a color difference signal.

The current frame memory 111 is an image memory for storing the current image data of the current frame and reading it out in block units.

The previous frame memory 112 is an image memory for storing the previous image data of the previous frame and reading it out in block units.

The frame delay section 121 delays the current frame image data by one frame.

The luminance signal preprocessing section 122 processes the pixel data of the luminance signal among the pixel data so that the pixel data has the same number of pixels as the chrominance signal and pixels at the same spatial position.

The control unit 123 controls the current frame memory 1
11 and the previous frame memory 112 to output a control signal to the current frame memory 111 and the previous frame memory 11.
2 to read the macroblock of the current frame and the block data of the reference frame.

The motion vector detecting section 124 detects a motion vector based on the luminance signal processed by the luminance signal preprocessing section 122 and pixel data in a block of a chrominance signal spatially located at the same position as the luminance signal. Perform detection.

As described above, the input terminal 101 is connected to the current frame memory 111 and the frame delay unit 121,
The frame delay unit 121 is connected to the previous frame memory 112. Further, the current frame memory 111 and the previous frame memory 112 are connected to the luminance signal preprocessing unit 122 and the motion vector detection unit 124. Further, the control unit 123 is connected to the current frame memory 111 and the previous frame memory 112, and the motion vector detection unit 124 is connected to the output terminal 102.

FIG. 2 is a block diagram showing the structure of the motion vector detecting section 124.

In FIG. 2, the motion vector detecting section 124
Is composed of a pixel accuracy motion vector detection unit 210 and a half-pixel accuracy motion vector detection unit 220. The pixel accuracy motion vector detection unit 210 includes a similarity calculation unit 211 (evaluation value calculation unit) and a candidate vector selection unit 212 ( Vector selection means).

The pixel precision motion vector detecting section 210
Is to detect a pixel-accurate motion vector in the first stage in the Full Pel + Half Pel system.

The half-pixel precision motion vector detector 220
In the Full Pel + Half Pel method, a motion vector at a half pixel position is detected based on a candidate vector detected in the first stage in the second stage.

The similarity calculator 211 calculates the similarity between the macroblock of the current frame and the block of the reference frame.

The candidate vector selection unit 212 determines candidate vectors for performing half-pixel precision motion vector detection using the calculated evaluation values for all the vectors to be tried.

As described above, inside the motion vector detecting section 124, there are the pixel precision motion vector detecting section 210 and the half pixel precision motion vector detecting section 220, and the input data is input to the similarity calculating section 211. Similarity calculator 21
1 is connected to the candidate vector selection unit 212, and the candidate vector selection unit 212 is connected to the half-pixel precision motion vector detection unit 220.

Next, the operation of the motion vector detecting device 100 configured as described above will be described.

Overall Operation The input terminal 101 receives pixel data of the current frame. This input pixel data is stored in the current frame memory 111.
And the data delayed by one frame by the frame delay unit 121 is stored in the previous frame memory 112. Therefore, the pixel data of the current frame necessary for detecting the motion vector is stored in the current frame memory 111.
And the pixel data of the frame to be referred to is stored in the previous frame memory 112.

The control unit 123 reads the macroblocks of the current frame and the block data of the reference frame necessary for detecting the motion vector from the frame memories 111 and 112.

The pixel data of the luminance signal among the read pixel data is input to the luminance signal preprocessing unit 122, and is processed so as to have the same number of pixels and spatially the same pixels as the color difference signal.

The luminance signal processed by the luminance signal preprocessing unit 122 and the pixel data in the block of the color difference signal spatially located at the same position as the luminance signal are obtained by the motion vector detection unit 12.
4 and a motion vector is detected.

Operation of Motion Vector Detection Unit 124 (FIG. 2) In the motion vector detection unit 124, the similarity calculation unit 211 uses the pixel data obtained by processing the input pixel data of the luminance signal to generate the current frame. Is calculated between the macro block and the block of the reference frame. At the time of calculation, the same processing is performed separately on the luminance signal and the color difference signal. In order to calculate the similarity between blocks, for example, the absolute value of the difference between the pixels in the block is calculated, and the evaluation value is calculated using the value obtained by cumulatively adding the absolute value of the difference for all the pixels as the evaluation value. I do.

The similarity calculation process is performed separately for the luminance signal and the color difference signal, and the similarity between the two blocks is determined using the two evaluation values.

The candidate vector selection unit 212 determines a candidate vector for performing half-pixel precision motion vector detection using the evaluation value obtained as a result of the similarity calculation for all the vectors to be tried. In the determination of the candidate vector, for example, when the absolute value of the cumulatively added difference as the evaluation value is used, when the evaluation value obtained from the luminance signal is Y, and when the evaluation value obtained from the color difference signal is C, H = A × Y + b × C (where a and b are integers) are selected as candidate vectors in order from the trial vector in which the value H obtained is smaller. By setting the number of candidates in advance, the processing amount is kept constant.

The half-precision motion vector detecting section 220 performs processing on the trial vectors listed as candidate vectors. However, when the position information of the candidate vector has information different from the position of the input luminance signal in the luminance signal preprocessing unit 122, the candidate vector is converted into a position-corrected vector, and the corrected vector is used to obtain half-pixel accuracy. Detect a motion vector.

In the motion vector detection with half-pixel accuracy, a block that is spatially shifted by half a pixel is created by interpolation using a block obtained as a candidate vector and its surrounding pixel information. Seeking similarity,
Select the one with the highest similarity from all the blocks,
The difference between the position information of the block and the macroblock of the current frame is output as a motion vector.

As described above, the motion vector estimating apparatus 100 according to the first embodiment has only one motion vector per block in the motion vector estimating using the multi-step method in the motion compensation inter-frame prediction. In the case of a system that cannot send the input luminance signal, a luminance signal pre-processing unit 122 (processing means) for down-sampling the input luminance signal is provided.
1 and the candidate vector selection unit 212 evaluate the luminance signal and the chrominance signal by the same evaluation means, obtain a similarity as an evaluation value, and re-evaluate the similarity between blocks using each similarity, thereby obtaining a plurality of candidate vectors. The same circuit is used for processing the luminance signal and the chrominance signal of the similarity detection unit by performing processing such as downsampling on the luminance signal and using the same number of pixels and the trial vector as the chrominance signal. Can be used.

Since a vector based on the spatial position information of the input chrominance signal is selected as a candidate vector, when detecting a motion vector with half-pixel accuracy, the candidate vector The candidate vector corrected to the position information does not indicate an adjacent block, and the interpolated blocks at the half pixel position do not overlap.

Therefore, it is not necessary to calculate the similarity with the block at the same half-pixel position, and the effect of improving the detection accuracy by using a plurality of candidate vectors can be easily exerted. Also, at the time of candidate vector selection, by taking into account the evaluation value of the color difference signal, the inter-frame prediction effect on the color difference signal is obtained by simply correcting the position of the motion vector obtained by the luminance signal with the motion vector of the color difference signal. Can be improved as compared to the case where

FIG. 3 is a block diagram showing the overall processing procedure of the motion vector detecting device according to the second embodiment of the present invention. As in the first embodiment, an example in which detection is performed using the Full Pel + Half Pel method, which is multi-stage motion vector detection, is shown, and processing in a case where a plurality of candidate vectors are used is shown. In the description of the motion vector detecting device according to the present embodiment, the same components as those of the motion vector detecting device shown in FIG. 1 are denoted by the same reference numerals, and the description of the overlapping portions will be omitted.

In FIG. 3, the motion vector detecting device 30
0 is the current frame memory 111 and the previous frame memory 11
2. It comprises a frame delay unit 121, a control unit 123, and a motion vector detection unit 301.

Here, the input terminal 101 is connected to the current frame memory 111 and the frame delay section 121, and the frame delay section 121 is connected to the previous frame memory 112. In addition, the current frame memory 111 and the previous frame memory 112 are connected to the motion vector detection unit 301. The control unit 123 also stores the current frame memory 11
1 and the previous frame memory 112, and the motion vector detection unit 301 is connected to the output terminal 102.

FIG. 4 is a block diagram showing the structure of the motion vector detecting section 301.

In FIG. 4, a motion vector detecting section 301
Is composed of a pixel accuracy motion vector detection unit 310 and a half-pixel accuracy motion vector detection unit 220.

The pixel-precision motion vector detector 310 includes a luminance signal similarity calculator 311 (luminance signal similarity calculator), a color difference signal similarity calculator 312 (color difference signal similarity calculator), and a luminance signal candidate vector selector. 313, color difference signal candidate vector selection unit 314 and candidate vector selection unit 3
15 is comprised.

The luminance signal candidate vector selection section 313,
The color difference signal candidate vector selection unit 314 and the candidate vector selection unit 315 as a whole are vector selection means 316 for individually selecting a plurality of candidate vectors from the similarity evaluation obtained from the luminance signal and the similarity evaluation obtained from the color difference signal. Is configured.

As described above, the frame memory 111,
Reference numeral 112 is connected to the luminance signal similarity calculator 311 and the color difference signal similarity calculator 312. The luminance signal similarity calculating unit 311 includes a luminance signal candidate vector selecting unit 313.
, And the color difference signal similarity calculation unit 312 is connected to the color difference signal candidate vector selection unit 314. further,
The luminance signal candidate vector selection unit 313 and the chrominance signal candidate vector selection unit 314 are connected to the candidate vector selection unit 315, and the candidate vector selection unit 315 is connected to the half-pixel precision motion vector detection unit 220.

Next, the operation of the motion vector detecting device 300 configured as described above will be described.

Overall Operation The input terminal 101 receives pixel data of the current frame. This input pixel data is stored in the current frame memory 111.
And the data delayed by one frame by the frame delay unit 121 is stored in the previous frame memory 112. Therefore, the pixel data of the current frame necessary for detecting the motion vector is stored in the current frame memory 111.
And the pixel data of the frame to be referred to is stored in the previous frame memory 112.

The control unit 113 reads macroblocks of the current frame and block data of the reference frame required for motion vector detection from the frame memories 111 and 112. The corresponding block data read out is
Input to the motion vector detection unit 301 and output terminal 102
Outputs a motion vector for the macroblock of the current frame.

Operation of Motion Vector Detection Unit 301 (FIG. 4) Of the pixel data of the block input to the motion vector detection unit 301, the luminance signal is input to the luminance signal similarity calculation unit 311 and the luminance signal is Is calculated between the two blocks. Similarly, the similarity of the color difference signal is calculated by the color difference signal similarity calculator 312.

Here, the similarity is calculated by, for example, calculating the absolute value of the difference between the pixels in both blocks and cumulatively adding the absolute value of the difference to all the pixels to indicate the similarity. Find the evaluation value.

Using the evaluation values of the luminance signal and the chrominance signal calculated by the luminance signal similarity calculating section 311 and the chrominance signal similarity calculating section 312, the luminance signal candidate vector selecting section 313 and the chrominance Each candidate vector is selected in the signal candidate vector selection unit 314. When using the evaluation value at the time of selection, the candidate is sequentially selected from the trial vector having the smaller evaluation value. Also, the processing amount can be adjusted by determining the number of candidates in advance.

In the candidate vector selection section 315, it is necessary to correct the vector information between the candidate vector obtained from the luminance signal and the candidate vector obtained from the chrominance signal due to the difference in resolution between the luminance signal and the chrominance signal as input data. Therefore, the position of the candidate vector of the color difference signal is corrected so as to match the position information in the luminance signal.

A candidate vector obtained from the position-corrected color difference signal and a candidate vector obtained from the luminance signal are output as candidate vectors for motion detection with half-pixel accuracy.

As in the first embodiment, the half-pixel precision motion vector detecting section 220 detects a half-pixel precision motion vector using the candidate vector, and outputs the vector having the highest similarity as the motion vector.

As described above, the motion vector detection unit 3 of the motion vector detection device 300 according to the second embodiment
Reference numeral 01 denotes a luminance signal similarity calculation unit 311, which calculates a similarity evaluation value from a luminance signal, a color difference signal similarity calculation unit 312, which calculates a similarity evaluation value from a color difference signal, a luminance signal candidate vector selection unit 313, and a color difference signal candidate. A pixel accuracy motion vector detection unit 310 including a vector selection unit 314 and a candidate vector selection unit 315;
20, and a plurality of candidate vectors are respectively selected from the similarity evaluation obtained from the luminance signal and the similarity evaluation obtained from the chrominance signal. Therefore, the similarity of the luminance signal and the similarity of the chrominance signal are selected. By selecting candidate vectors from each signal using the evaluation value of
The effect of inter-frame prediction on the color difference signal can be improved.

Further, it is possible to prevent the deterioration of the inter-frame prediction effect due to erroneous detection when performing multi-stage detection from affecting the color difference signal.

In each of the above embodiments, Full Pel + Ha
Although an example in which the present invention is applied to motion vector detection in two stages of lf Pel has been described, it is needless to say that the present invention can be applied to a case where a plurality of candidate vectors are selected when another multi-stage detection method such as the 3Step method is used. is there.

When the input pixel data is not in the YCbCr hue system, for example, in the case of RGB, the input pixel data is changed from RGB to Y.
The motion vector is detected after the conversion to CbCr. In this case, when the RGB data is input as it is and the number of candidates is set to 1, the same motion vector as when the RGB to YCbCr conversion is performed can be detected.

In each of the above embodiments, the motion vector detection method may be applied to a moving picture compression apparatus based on, for example, the MPEG algorithm. However, the present invention is not limited to this. It is needless to say that the present invention can be applied to the above device.

Further, it goes without saying that the number and types of circuits and members constituting the motion vector detecting device and the motion vector detecting section are not limited to the above-described embodiment.
It may be realized by software (for example, C language).

[0078]

In the motion vector detecting device according to the present invention, there are provided processing means for down-sampling the input luminance signal, and evaluation value calculating means for performing the same evaluation on the luminance signal and the color difference signal by the processing means and calculating the evaluation value. And a vector selecting means for obtaining a similarity as an evaluation value, re-evaluating the similarity between blocks using each similarity, and selecting a plurality of candidate vectors. The effect of improving the detection accuracy can be achieved, and the effect of inter-frame prediction on the color difference signal can be improved.

Further, in the motion vector detecting device according to the present invention, there are provided a luminance signal similarity calculating means for calculating a similarity evaluation value from a luminance signal, and a chrominance signal similarity calculating means for calculating a similarity evaluation value from a chrominance signal. And a vector selection unit for selecting a plurality of candidate vectors from the similarity evaluation obtained from the luminance signal and the similarity evaluation obtained from the chrominance signal. Thus, it is possible to prevent the deterioration of the inter-frame prediction effect due to erroneous detection when performing multi-stage detection from affecting the color difference signal.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating an overall processing procedure of a motion vector detection device according to a first embodiment to which the present invention has been applied.

FIG. 2 is a block diagram illustrating a configuration of a motion vector detection unit of the motion vector detection device.

FIG. 3 is a block diagram illustrating an overall processing procedure of a motion vector detection device according to a second embodiment to which the present invention has been applied.

FIG. 4 is a block diagram illustrating a configuration of a motion vector detection unit of the motion vector detection device.

FIG. 5 is a block diagram showing a configuration of a conventional motion vector detection device.

[Explanation of symbols]

100, 300 motion vector detecting device, 111 current frame memory, 112 previous frame memory, 121 frame delay unit, 122 luminance signal preprocessing unit (processing means),
123 control unit, 124, 301 motion vector detection unit, 210, 310 pixel precision motion vector detection unit, 2
11 similarity calculator (evaluation value calculator), 212 candidate vector selector (vector selector), 220 half-pixel precision motion vector detector, 311 luminance signal similarity calculator (luminance signal similarity calculator), 312 color difference Signal similarity calculator (color difference signal similarity calculator), 313 luminance signal candidate vector selector, 314 color difference signal candidate vector selector, 315 candidate vector selector, 316 vector selector

Claims (6)

[Claims] 1. A motion vector detecting apparatus that divides an encoding target frame into predetermined blocks and detects a position where a difference between the preceding frame and the difference between the preceding frames becomes smallest. Processing means; evaluation value calculating means for performing the same evaluation of the luminance signal and the color difference signal by the processing means to calculate an evaluation value; obtaining similarity as the evaluation value; and using each similarity to calculate the similarity between blocks. A motion vector detecting device comprising: a vector selecting means for re-evaluating and selecting a plurality of candidate vectors. 2. The image processing apparatus according to claim 1, wherein the processing unit processes the pixel data of the luminance signal among the pixel data so that the pixel data has the same number of pixels as the color difference signal and pixels at the same spatial position. The motion vector detecting device according to claim 1. 3. A motion vector detecting device which divides an encoding target frame into predetermined blocks and detects a position where a difference between the preceding frame and the previous frame is smallest, wherein a similarity evaluation value is calculated from a luminance signal. A luminance signal similarity calculating means for calculating, a color difference signal similarity calculating means for calculating a similarity evaluation value from the color difference signal, and a similarity evaluation obtained from the luminance signal and a similarity evaluation obtained from the color difference signal, respectively. A motion vector detecting device, comprising: vector selecting means for selecting a plurality of candidate vectors. 4. The motion vector detection device according to claim 1, wherein the motion vector detection is a motion vector detection using a multi-step method in motion compensated inter-frame prediction. Motion vector detection device. 5. The motion vector detecting device according to claim 1, wherein in the motion vector detection, a motion vector with pixel accuracy is detected in a first step, and the first step is performed in a second step. The motion vector at the half pixel position based on the candidate vector detected
A motion vector detecting device, wherein the motion vector is detected using the Full Pel + Half Pel method. 6. The motion vector detection device according to claim 1, wherein the motion vector detection is a motion vector detection in which only one motion vector is sent for one block. .