JP2007174202A - Motion vector detecting device and method therefor - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To detect motion vector, even for a moving image having a large motion. <P>SOLUTION: Rather than setting a search window about the current macro block as the center, a global vector, representing the motion of the current image and a reference image, and the reliability of the global vector, is used to set the shape and position of the search window in the reference image. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to a motion vector detection apparatus and method, and more particularly to a motion vector detection apparatus and method capable of adaptive motion vector detection in consideration of the magnitude of motion between images.

  In recent years, digitalization of information related to so-called multimedia such as audio signals and video signals has been advancing rapidly, and in response to this, compression coding and decoding techniques for video signals have attracted attention. Since compression encoding and decoding techniques can reduce the storage capacity required for storing video signals and the bandwidth required for transmission, this technique is extremely important for the multimedia industry.

  These compression encoding / decoding techniques compress the information amount / data amount by utilizing the high autocorrelation (ie, redundancy) of many video signals. Video signal redundancy includes temporal redundancy and two-dimensional spatial redundancy, which can be reduced using block-wise motion estimation and motion compensation, while spatial redundancy is It can be reduced using discrete cosine transform (DCT).

  In the MPEG system known as compression encoding / decoding technology, the redundancy of the video signal is reduced by these methods, and the data compression effect of the video frame / field changing with time is improved. The block-by-block motion estimation for reducing temporal redundancy is an operation of finding a block that is closest to a reference frame / field (past frame / field) that is continuously input and a current frame / field. A vector representing the moving direction and moving amount of the corresponding block is called a motion vector. Therefore, motion estimation is synonymous with motion vector estimation.

  In general, a block matching method is used as a method for estimating a motion vector. The block matching method refers to a method in which two images, such as a reference frame / field and a current frame / field, are compared in units of blocks, and motion is estimated in units of blocks based on the degree of matching of signal types. With this block matching method, a motion vector is estimated for each block from the reference frame / field and the current frame / field, and motion compensation prediction is performed using the estimated motion vector. This block matching method is described in Patent Document 1, for example, and will be described with reference to FIG.

  FIG. 2 is a block diagram showing a motion vector detection apparatus using a block matching method. In the figure, the motion vector detection apparatus includes a current frame / field storage unit 20, a reference frame field storage unit 21, a current macroblock storage buffer 22, a reference search window storage buffer 23, and a motion vector search unit 24.

  The current frame / field storage unit 20 and the reference frame / field storage unit 21 store the current frame / field and the reference frame / field, respectively, and are used for motion vector estimation. The current macroblock storage buffer 22 extracts the current macroblock image from the current frame / field storage unit 20. The reference search window storage buffer 23 sets the center of the search area at the center of the current macroblock, and extracts the image from the reference frame field storage unit 21 for the range of the search area (hereinafter referred to as the search window). The motion vector search unit 24 searches for an image of the current macroblock in the search window, and estimates a final motion vector for the current macroblock.

  In addition, as shown in FIG. 3, the block matching method includes a configuration in which the motion vector detection apparatus shown in FIG. 2 includes a motion vector storage unit 35 that stores the motion vector estimated by the motion vector search unit 24. To do. In this configuration, the motion vector estimated for the previous macroblock from the motion vector storage unit 35 is provided to the reference frame / field storage unit 21. The reference frame / field storage unit 21 sets the center position of the search area at a position shifted from the current macro block by the motion vector of the previous macro block, and outputs the reference image to the reference search window storage buffer 23.

  A method of detecting / estimating a motion vector for each macro block as in the block matching method, and performing motion compensation prediction using this motion vector is collectively referred to as local motion compensation in the sense of performing local detection. I will call it. A motion vector in units of blocks is referred to as a local motion vector.

The local motion compensation method described so far is currently the international standard method H.264. 261, H.H. It is used in H.263, MPEG1, MPEG2 and MPEG4, and is the most widely adopted motion compensation method.
JP 04-323780 A (2nd page)

  However, since the local motion compensation method uses a search window consisting of a smaller number of pixels than one frame, when the center of the search area is set at the center of the current macroblock, or the motion vector of the previous macroblock When the center of the search window is positioned using, an image in which the motion between images is fast or the entire screen is panned as in a baseball broadcast does not exist in the search window. In this case, the conventional local motion compensation method does not perform non-intra coding to obtain a difference value between image frames, but instead performs intra coding, so that the compression rate is reduced and the amount of encoded data is large. There is a problem that becomes.

  Accordingly, an object of the present invention is to provide a motion vector detection apparatus and method capable of solving the above-described problems and detecting an appropriate motion vector in consideration of the motion of an image having a large motion between frames or fields. It is to provide.

  Also provided is a motion compensated predictive coding apparatus that can maintain a high compression rate when the motion between frames or fields is small and can perform coding with a high compression rate even when the motion between frames or fields is large. It is a further object of the present invention.

  An object of the present invention is to provide a motion vector detection device that detects a motion vector of each of the plurality of macroblocks by searching each of a plurality of macroblocks constituting the current image in a search region set in a reference image. A global vector determining means for obtaining a global vector which is a motion vector between the current image and the reference image, a reliability determining means for determining the reliability of the global vector, and a determination result of the reliability of the global vector Search area determining means for determining the shape of the search area according to the search area, and a motion vector for searching for a macroblock of the current image within the determined search area of the reference image and detecting a motion vector for the macroblock It is achieved by a motion vector detection device characterized by comprising search means.

  In addition, the above-described object is to detect a motion vector of each of the plurality of macroblocks by searching each of a plurality of macroblocks constituting the current image in a search area set in the reference image. A global vector determining means for determining a global vector which is a motion vector between the current image and the reference image, a reliability determining means for determining the reliability of the global vector, and a reliability of the global vector. Search area determining means for determining the position of the search area according to the determination result, and searching for a macroblock of the current image within the determined search area of the reference image, and detecting a motion vector for the macroblock Also achieved by a motion vector detection device characterized by having motion vector search means It is.

  Furthermore, the above-described object is to detect a motion vector of each of the plurality of macroblocks by searching each of a plurality of macroblocks constituting the current image in a search area set in the reference image. A global vector determination step for obtaining a global vector that is a motion vector between the current image and the reference image, a reliability determination step for determining reliability of the global vector, and reliability of the global vector A search region determining step for determining at least one of a shape and a position of the search region according to a determination result; searching for a macroblock of the current image in the determined search region of the reference image; A motion vector search step for detecting a motion vector. Also achieved by the motion vector detecting method according to.

  According to the present invention, since an appropriate motion vector can be detected in consideration of the motion of an image having a large motion between frames or fields, the accuracy of motion vector detection is improved and the motion is large. Can also improve the compression ratio.

  Hereinafter, the present invention will be described in detail based on preferred embodiments with reference to the drawings.

  FIG. 1 is a block diagram showing a configuration example of a motion vector detection device according to an embodiment of the present invention. 1 includes a reference frame / field storage unit 100, a current frame / field storage unit 101, a global vector determination unit 102, a global vector reliability determination unit 103, and a search window position / shape determination unit 104. In other words, the motion vector search region (search window) is determined using the global vector. Furthermore, a motion vector search unit 105, a motion vector storage unit 106, a reference search window storage buffer 107, and a current macroblock storage buffer 108 are provided, and a motion vector is detected with the above configuration.

  In the configuration, the specific operation will be described below. The reference frame / field storage unit 100 and the current frame / field storage unit 101 store a reference frame / field (reference image) and a current frame / field (current image) for motion vector detection / estimation, respectively.

  The global vector determination unit 102 uses all the reference frame / field pixel values and all the current frame / field pixel values provided from the reference frame / field storage unit 100 and the current frame / field storage unit 101 to generate a reference frame / A global vector indicating the difference in the spatial position of the current frame / field relative to the field is determined.

  The global vector reliability determination unit 103 determines the reliability of the global vector determined by the global vector determination unit 102. The search window determination unit 104 determines the position / shape of the reference search window from the global vector or the previous local motion vector.

  The motion vector search unit 105, the motion vector storage unit 106, the reference search window storage buffer 107, and the current macroblock storage buffer 108 may be the same as the components having the same names described in FIGS.

  The operation of the motion vector detection apparatus according to the present embodiment described above will be described with reference to the flowchart of FIG.

First, the global vector determination unit 102 detects a global vector that is a motion vector between a reference frame / field (reference image) and a current frame / field (current image). Calculate (step S103)
In order to estimate a global vector representing a moving distance and direction in which the reference frame / field and the current frame / field have the maximum correlation, the following MSE (Mean Square Error) (Equation 1) and MAE (Mean Absolute) are used. An evaluation function such as Error (Equation 2) or MAD (Mean Absolute Difference) can be used.

Here, S _cur (m, n) is the (m, n) -th pixel value in the current frame / field, S _ref (m, n) is the reference frame, and the (m, n) -th pixel in the field Indicates the value. (I, j) indicate the spatial position of the current frame / field with respect to the reference frame / field, respectively.

(However, when M and N are the number of horizontal and vertical pixels in one frame / field, m = k × q and n = 1 × r, where k and l are 0 ≦ m ≦ M and 1 ≦ k ≦ M. , 0 ≦ n ≦ N, 1 ≦ l ≦ N, and Q and R satisfy M−k ≦ Q ≦ M and N−l ≦ R ≦ N.)
This evaluation function is based on the difference in pixel values, and determines the vector having the smallest MAE value or MSE value as the global vector.

  FIG. 9 shows an example of a global vector selection method when the MAE value is taken as an example. The reference frame is shifted one pixel at a time in a predetermined direction, and the average of the sum of MAE values is taken for each pixel moving distance. The amount of movement when the average MAE value is the minimum is the global vector selection criterion. For example, if this process is performed also in another direction orthogonal to a predetermined direction and a movement amount that minimizes the average MAE value in this direction is obtained, a global vector is determined from the two movement amounts and the movement direction. Can do.

  Further, the MAE value and the MSE value obtained at this time are used as a global vector reliable value GRV (Global Vector Reliable Value) by the global vector reliability determining unit 103. In this way, after performing the above-described evaluation function calculation, the global vector determination unit 102 determines a motion vector to a position indicating the maximum correlation as a global vector, and the global vector reliability determination unit 103 receives the global vector reliability value GRV. The global vector is transmitted (step S105).

(Determining the reliability of global vectors)
Next, the global vector reliability determination unit 103 determines the reliability of the global vector determined by the global vector determination unit 102 (step S107). An example of this determination method is shown in FIG.

  In this embodiment, the global vector reliability determination unit 103 compares the global vector reliability value GRV transmitted from the global vector determination unit 102 with a preset threshold value, and the value of the global vector reliability value GRV is equal to or less than the threshold value. If so, it is determined that the reliability of the global vector is high. At this time, since the global vector is used for determining the position of the search window, the global vector is transmitted to the search window position / shape determining unit 104.

  On the other hand, if the value of the global vector reliability value GRV is larger than the threshold value, it is determined that the reliability of the global vector is low. In this case, instead of the global vector, the previous local motion vector is transmitted from the motion vector storage unit 106 to the search window position / shape determination unit 104. In this case, a zero motion vector indicating no motion can be transmitted instead of the previous local motion vector. In addition, any motion vector that can be defined by the user may be transmitted instead of the previous local motion vector.

(Determination of search window position and shape)
The search window position / shape determining unit 104 determines the position and shape of the search window according to the transmitted motion vector (according to the global vector (step S109) or the previous local motion vector (step S111)).

First, the center of the search window is designated at a position shifted from the reference block / field macroblock corresponding to the current macroblock by the motion vector. Subsequently, for each of the horizontal direction and the vertical direction, the threshold value set in advance and the value of the motion vector in the horizontal direction and the vertical direction are compared to determine the shape of the search window. Specifically, as shown in FIG. 6, the horizontal threshold value TH _h (−TH _h ) and the vertical threshold value TH _V (−TH _V ) are set, and the horizontal and vertical values of the motion vector are shown in the graph. The shape of the search window is determined depending on where it belongs. For example, when the value x in the horizontal direction of the motion vector is 0 <x <TH _h and the value y in the vertical direction is TH _V <y, the search window belongs to the area shown in FIG. It becomes the shape of. On the other hand, when x <−TH _h and y <−TH _V , the search window has the shape of (v) of FIG. 5 because it belongs to the area of (m) of FIG.

  Thus, the shapes that the search window can take are as shown in FIGS. Further, the shapes of the search windows shown in FIGS. 5 (i) to 5 (v) are determined according to the divided areas as shown in FIGS. 6 (a) to 6 (p).

Next, FIG. 7 (i) shows how the position / shape of the search window is determined in FIG. 6 when, for example, 0 <x <TH _h and TH _V <y (when FIG. 6 (c)). Similarly, FIG. 7 (ii) shows how the position / shape of the search window is determined when x <−TH _h and y <−TH _V (FIG. 6 (m)).

  The current frame / field storage unit 101 extracts the image data of the current macroblock and transmits it to the current macroblock storage buffer 108. At the same time, the current macroblock position information is transmitted to the reference frame / field storage unit 100. Then, the reference frame / field storage unit 100 extracts image data of a necessary search range corresponding to the position of the current macroblock in accordance with the position / shape of the search window determined by the search window position / shape determination unit 104. Transmit to the save buffer 107. FIG. 7 (i) and FIG. 7 (ii) schematically show this state.

  As described above, when the reliability of the global vector is high, the shape is made to follow the global vector, and the position of the search window is determined so as to search around the position indicated by the global vector. The position and shape of the search window are determined so as to search the periphery. FIG. 8 schematically shows this state.

(Search motion vector)
Next, the motion vector search unit 105 searches for a region similar to the macroblock in the search window, and detects / estimates a motion vector corresponding to the current macroblock (step S113). In addition to outputting the estimated motion vector to the outside, it is transmitted to the motion vector storage unit 106 and used for motion vector estimation in the next macroblock when the reliability of the global vector is low.

  Here, if the motion vector of the N × N size macroblock is searched in the range of ± p pixels in the reference frame / field, the size of the search window is (N + 2p) × (N + 2p). Here, after calculating the above-described evaluation function at all positions that can be motion vector candidates, a vector to a position showing the maximum correlation is determined as a motion vector.

  In order to estimate the motion vector having the maximum correlation, the evaluation function such as the above-mentioned MSE (Mean Square Error) (Equation 3), MAE (Mean Absolute Error) (Equation 4), or MAD (Mean Absolute Difference) is used. It is calculated as follows.

Here, S _ref indicates a reference frame / field, S _{cur, k} indicates the current frame, and the k th macroblock in the field. (I, j) indicates the spatial position of the reference frame / field with respect to the k-th macroblock of the current frame / field.

(However, if X and Y are the number of horizontal and vertical pixels of the search window, x = g × u, y = h × v, and g and h are 0 ≦ x ≦ X, 1 ≦ g ≦ X, 0 ≦ (Y ≦ Y, 1 ≦ h ≦ Y is a natural number, and U and V satisfy X−g ≦ U ≦ X and Y−h ≦ V ≦ Y.)
This evaluation function is based on the pixel difference, and the case having the smallest MAE value or MSE value is selected as the final motion vector in the current macroblock.

  The motion vector of the selected macroblock is used for encoding processing in an encoding apparatus (not shown) for encoding the macroblock. The above encoding apparatus is MPEG2 or H.264. H.264 / MPEG4-AVC and other motion compensated predictive encoding devices including a motion vector detecting device according to the present embodiment, or a high-efficiency compression code even for a large motion image. Can be made.

  In the above-described embodiment, one global vector is calculated for one current image / reference image. However, the current image / reference image is divided into a plurality of areas and each area is globally calculated. It is also possible to take a configuration in which a vector (GRV) is calculated and an appropriate search window is determined for each area by a plurality of GRVs in the screen. FIG. 13 shows an example in which an image is divided into four. In FIG. 13, in the area (13-1), the global vector points in the upper right direction, but since GRV is low, the search window is set around the macroblock. On the other hand, in the area (13-2), since the GRV is high and the global vector points to the right, the search window is placed on the right side of the macroblock, and the shape is determined as shown in FIG. 5 (iii). Further, (13-3) is an example in which the GRV is high and the global vector points in the lower left direction, so that the search window is placed at the lower left of the macro block and the shape is determined as (v) in FIG. Further, (13-4) is a case where GRV is high and the global vector points to the upper right direction, so that the search window is placed at the upper right of the macro block and the shape is determined as (v) in FIG. In this way, by dividing an image into a plurality of areas and processing them, it becomes possible to detect a motion vector with higher accuracy following the movement of each area. This is effective for a scene where only a part of the screen is moving greatly.

  As described above, according to the present embodiment, with the above-described configuration, a global vector that is a motion vector between the current image and the reference image is obtained, and a reference search window is set using the global vector, so that the current macroblock is obtained. As compared with the case where the reference search window is set using the center of the reference search window or the previous local motion vector, an appropriate motion vector can be detected even for an image having a large motion. Therefore, by using the motion vector detection apparatus according to the present embodiment, it is not possible to determine a motion vector in a scene with a large motion, and solves the problem in the conventional motion compensation coding system that performs intra coding and increases the amount of code. Thus, it is possible to improve the motion compensation predictive coding efficiency.

  Next, a motion vector detection apparatus according to the second embodiment of the present invention is described. The motion vector detection device according to the second embodiment is different from the configuration of the first embodiment in that the search window position / shape determination unit 104 in the motion vector detection device in FIG. Has the same configuration and function as the first embodiment. Therefore, in the present embodiment, the description will be made with reference to FIG. 1, and the redundant description of the same components will be omitted.

  Hereinafter, the operation of the motion vector detection apparatus of the present embodiment will be described with reference to the flowchart of FIG.

  First, the processing up to global vector determination (steps S103 and S105) is the same as that in the first embodiment.

  In this embodiment, the reliability determination process performed by the global vector reliability determination unit 103 in step S207 and the search window setting process corresponding to the result are different from those in the first embodiment. Specifically, when the global vector reliability determination unit 103 performs reliability determination in step S207, a plurality of threshold values are used. Then, the position / shape of the search window is changed according to the relationship between the reliability value GRV and the plurality of threshold values. In the present embodiment, the case where the global vector reliability determination unit 103 performs reliability evaluation using two threshold values Th1 and Th2 will be described as an example.

  In this case, the global vector reliability value GRV and the threshold values Th1 and Th2 can have the following three relationships.

  And the global vector reliability determination part 103 of a present Example is most reliable when the reliability falls in order of (i), (ii), (iii) of Formula 5, that is, the condition of (ii) is satisfied. Judge as high. Then, the position and shape of the search window are changed according to the determination result (steps S209, S211, and S213).

  Specifically, in the case of the highest reliability (i), the position of the search window is set to the position indicated by the global vector, and the shape of the search window is also changed to follow the direction indicated by the global vector. Next, when the reliability is high (ii), the reliability of the global vector is lower than in the case of (i). Therefore, although the position of the search window is shifted to the position indicated by the global vector, the shape of the search window does not move. Assume that in FIG. 5 (i) used in the state.

  In the case of (iii), since the reliability of the global vector is further lowered, the global vector is not used for positioning the search window, and instead of the global vector, the previous local motion vector or motion from the motion vector storage unit 106 is used. The position of the search window is determined using a zero motion vector meaning that there is no other, or any other motion vector definable by the user.

  In this embodiment, a zero motion vector is used, and a search window is set around the macroblock. Further, the shape of the search window is assumed to be that shown in FIG.

  FIG. 12 schematically shows search window range control according to the reliability of the global vector in this embodiment. As described above, in this embodiment, depending on the reliability of the global vector, both the position and shape of the search window are changed by the global vector, or only the position of the search window is changed by the global vector, or the position and shape of the search window. Determines whether to depend on global vectors.

  The search window position / shape determining unit 104 determines the reference frame / field storage unit 100 and the current frame / field storage unit 101 according to the global vector transmitted from the global vector reliability determination unit 103 and the position / shape of the search window. The images necessary for motion vector estimation are taken out from each of them and transmitted to the reference search window storage buffer 107 and the current macroblock storage buffer 108.

  Then, as described in the first embodiment, the motion vector search unit 105 detects and outputs a macroblock motion vector (step S113).

  In the present embodiment, the position and shape of the search window are determined using the global vector and the global vector confidence value GRV. However, the minimum value among the values obtained by the above Equation 3 or Equation 4 is used as the confidence value. The position and shape of the search window may be determined using a motion vector and its reliability value LRV (Local Vector Reliable Value).

  As described above, according to the present embodiment, since the size of the search window is controlled according to the reliability of the global vector, the possibility of searching for a local motion vector is improved even when the reliability of the global vector is low. Can do. Further, the local motion vector storage unit 106 is not necessary, and a simpler configuration can be realized.

(Other embodiments)
The motion vector detection apparatus described in each of the above embodiments can be suitably applied to a motion compensated prediction encoding apparatus, and as a result, it is possible to improve the compression rate for an image having a large motion.

  In the above-described embodiments, only the motion vector detection device configured by one device has been described, but an equivalent function may be realized by a system configured by a plurality of devices.

  A software program for realizing the functions of the above-described embodiments is supplied from a recording medium directly to a system or apparatus having a computer that can execute the program using wired / wireless communication. The present invention includes a case where an equivalent function is achieved by a computer executing the supplied program.

  Accordingly, the program code itself supplied and installed in the computer in order to implement the functional processing of the present invention by the computer also realizes the present invention. That is, the computer program itself for realizing the functional processing of the present invention is also included in the present invention.

  In this case, the program may be in any form as long as it has a program function, such as an object code, a program executed by an interpreter, or script data supplied to the OS.

  As a recording medium for supplying the program, for example, a magnetic recording medium such as a flexible disk, a hard disk, a magnetic tape, MO, CD-ROM, CD-R, CD-RW, DVD-ROM, DVD-R, DVD- There are optical / magneto-optical storage media such as RW, and non-volatile semiconductor memory.

  As a program supply method using wired / wireless communication, a computer program forming the present invention on a server on a computer network, or a computer forming the present invention on a client computer such as a compressed file including an automatic installation function A method of storing a data file (program data file) that can be a program and downloading the program data file to a connected client computer can be used. In this case, the program data file can be divided into a plurality of segment files, and the segment files can be arranged on different servers.

  That is, the present invention includes a server device that allows a plurality of users to download a program data file for realizing the functional processing of the present invention on a computer.

  In addition, the program of the present invention is encrypted, stored in a storage medium such as a CD-ROM, distributed to the user, and key information for decrypting the encryption for a user who satisfies a predetermined condition is provided via a homepage via the Internet, for example. It is also possible to realize the program by downloading it from the computer and executing the encrypted program using the key information and installing it on the computer.

  In addition to the functions of the above-described embodiments being realized by the computer executing the read program, the OS running on the computer based on the instructions of the program may be part of the actual processing. Alternatively, the functions of the above-described embodiments can be realized by performing all of the processing and performing the processing.

  Furthermore, after the program read from the recording medium is written to a memory provided in a function expansion board inserted into the computer or a function expansion unit connected to the computer, the function expansion board or The CPU or the like provided in the function expansion unit performs part or all of the actual processing, and the functions of the above-described embodiments can be realized by the processing.

It is a block diagram which shows the structural example of the motion vector detection apparatus which concerns on 1st Example of this invention. It is a block diagram which shows the structural example of the conventional motion vector detection apparatus. It is a block diagram which shows another structural example of the conventional motion vector detection apparatus. It is a flowchart explaining operation | movement of the motion vector detection apparatus which concerns on 1st Example of this invention. It is a figure for demonstrating the shape of the search window which can be set. It is a figure for demonstrating the shape determination method of the search window according to the value of the horizontal direction of a global vector, and a perpendicular direction. It is an example of the determination method of the search window position and shape according to the reliability of the global vector in 1st Example. It is a figure for demonstrating the determination method of the search window position and shape according to the reliability of the global vector in 1st Example. It is a figure for demonstrating the determination method of the global vector in a 1st Example. It is a figure for demonstrating the determination method of the reliability of the global vector in a 1st Example. It is a flowchart explaining operation | movement of the motion vector detection apparatus which concerns on 2nd Example of this invention. It is a figure for demonstrating the determination method of the search window position and shape according to the reliability of the global vector in a 2nd Example. It is a figure for demonstrating the determination method of the search window position and shape at the time of dividing | segmenting the image in a 1st Example into a some area, and calculating | requiring a global vector and GRV with respect to each area.

Explanation of symbols

100 current frame / field storage unit 101 reference frame / field storage unit 102 global vector determination unit 103 global vector reliability determination unit 104 search window position / shape determination unit 105 motion vector search unit 106 motion vector storage unit 107 search window storage buffer 108 Current macroblock save buffer

Claims (14)

A motion vector detection device for detecting a motion vector of each of the plurality of macroblocks by searching in a search region set in a reference image for each of a plurality of macroblocks constituting a current image,
A global vector determining means for obtaining a global vector which is a motion vector between the current image and the reference image;
Reliability determination means for determining the reliability of the global vector;
Search area determining means for determining the shape of the search area according to the determination result of the reliability of the global vector;
A motion vector detection device comprising: motion vector search means for searching for a macroblock of the current image in the determined search area of the reference image and detecting a motion vector for the macroblock. A motion vector detection device for detecting a motion vector of each of the plurality of macroblocks by searching in a search region set in a reference image for each of a plurality of macroblocks constituting a current image,
A global vector determining means for obtaining a global vector which is a motion vector between the current image and the reference image;
Reliability determination means for determining the reliability of the global vector;
Search region determining means for determining the position of the search region according to the determination result of the reliability of the global vector;
A motion vector detection device comprising: motion vector search means for searching for a macroblock of the current image in the determined search area of the reference image and detecting a motion vector for the macroblock. A motion vector detection device for detecting a motion vector of each of the plurality of macroblocks by searching in a search region set in a reference image for each of a plurality of macroblocks constituting a current image,
A global vector determining means for obtaining a global vector which is a motion vector between the current image and the reference image;
Reliability determination means for determining the reliability of the global vector;
Search area determining means for determining the shape and position of the search area according to the determination result of the reliability of the global vector;
A motion vector detection apparatus comprising: motion vector search means for searching for a macroblock of the current image in the determined search area of the reference image and detecting a motion vector for the macroblock.   The search area determining means determines the search area according to horizontal and vertical component values of the global vector when the reliability determining means determines that the reliability of the global vector is high. The motion vector detection device according to claim 1, wherein: A motion vector detection device for detecting a motion vector of each of the plurality of macroblocks by searching in a search region set in a reference image for each of a plurality of macroblocks constituting a current image,
A global vector determining means for obtaining a global vector which is a motion vector between the current image and the reference image;
Reliability determination means for determining the reliability of the global vector;
Search area determining means for determining the shape of the search area according to the direction and size of the global vector when the reliability determination means determines that the reliability of the global vector is high;
A motion vector detection apparatus comprising: motion vector search means for searching for a macroblock of the current image in the determined search area of the reference image and detecting a motion vector for the macroblock.   Further, the storage unit stores the detected motion vector, and when the reliability of the global vector is low, the search region determination unit uses the motion vector stored in the storage unit and the shape of the search region and 6. The motion vector detection device according to claim 1, wherein at least one of the positions is determined.   In addition, when the global motion vector is low in reliability, the search area determining unit stores horizontal and vertical component values of the motion vector stored in the storage unit. The motion vector detection apparatus according to claim 1, wherein the shape of the search area is determined from the following.   In addition, when the global motion vector is low in reliability, the search region determination unit uses a zero motion vector or a predefined motion vector to store the shape of the search region. 6. The motion vector detection apparatus according to claim 1, wherein at least one of the position and the position is determined.   In addition, when the global motion vector is low in reliability, the search area determination unit is preset for each of the zero motion vector or the predefined motion vector when the global vector has low reliability. 9. The motion vector detecting apparatus according to claim 8, wherein the shape of the search area is selected.   The current image and the reference image are divided into a plurality of areas, a global vector is determined for each area, and a search region position is determined for each area according to the determination result of the reliability of the global vector. 10. The motion vector detection device according to claim 1, wherein   11. A motion compensated predictive coding apparatus that performs coding by motion compensated prediction of the current image using a motion vector detected by using the motion vector detecting apparatus according to claim 1. A motion vector detection method for detecting a motion vector of each of the plurality of macroblocks by searching in a search area set in a reference image for each of a plurality of macroblocks constituting a current image,
A global vector determination step for obtaining a global vector which is a motion vector between the current image and the reference image;
A reliability determination step of determining the reliability of the global vector;
A search region determination step of determining at least one of the shape and position of the search region according to the determination result of the reliability of the global vector;
A motion vector detection method comprising: searching for a macroblock of the current image within the determined search area of the reference image and detecting a motion vector for the macroblock.   A program causing a computer device to function as the motion vector detection device according to any one of claims 1 to 12.   A computer-readable recording medium storing the program according to claim 13.

Images