JP2010109917A - Motion vector detecting apparatus and motion vector detecting method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To improve accuracy of a motion vector search range in motion vector detection. <P>SOLUTION: A motion vector detecting apparatus includes: an important region detecting means for sequentially detecting an important region of an image for each of an input image to be encoded and a reference image; an important region moving amount detecting means for detecting the moving amount of the important region detected by the important region detecting means; a global vector detecting means for calculating a global vector as an overall moving amount of the image from the image to be encoded and the reference image; and a search area determining means for determining a motion vector search area in the reference image on the basis of predetermined conditions in a block to be encoded that is obtained by dividing the image to be encoded into a plurality of pixels on the basis of the moving amount of the important region calculated by the important region moving amount detecting means and the overall moving amount of the image calculated by the global vector detecting means. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to a motion vector detection apparatus and a motion vector detection method, and more particularly to a technique suitable for use in detecting a motion vector in video data encoding processing.

  As a method of encoding a moving image signal with high efficiency, redundant information is removed by using a motion amount (motion vector) between images using an image to be encoded and an image that precedes and follows on the time axis as a reference image. A method for reducing the transmission code amount is known.

  For example, in an encoding method such as MPEG2 or H.264, a method is used in which an image is divided into blocks each composed of a plurality of pixels, and a motion vector of the image is detected for each block.

  A block matching method is known as a method for detecting the motion vector described above. As shown in FIG. 8, in the block matching method, an encoding target block 802 obtained by dividing the encoding target image 801 into a plurality of pixels and a block 804 having the same coordinates as the encoding target block 802 in the reference image 803 are determined. To do.

  Further, the block 806 having the highest similarity with the encoding target block 802 is searched from a predetermined motion vector search area 805 in the reference image 803. Finally, the motion amount 807 between the block 804 and the block 806 is detected as a motion vector.

  Here, it is best to designate the entire reference image 803 as the predetermined motion vector search area 805 described above from the viewpoint of the code amount. However, searching for the entire reference image due to the recent increase in image size increases the amount of computation and is not realistic from the viewpoint of processing time and power consumption.

  Therefore, as a method for reducing the amount of calculation, a method is known in which a rough search is first performed, and a subsequent search is performed based on the search result (see, for example, Patent Document 1). With the method described in Patent Document 1, it is possible to expect a reduction in the amount of calculation, but if the search area is set incorrectly in the initial search, the result also affects the subsequent stage. For this reason, there is a problem that an erroneous motion vector is finally detected, and a high compression effect cannot be expected.

  As the above-described motion vector search area determination method, a method is known in which a motion amount (global vector) is calculated for the entire image or for each divided area, and the search area is determined based on the result of the global vector. .

Japanese Patent Laid-Open No. 10-164586

  However, in the search area determination method using the global vector as described above, there is a problem that when the movement direction of the subject and the movement direction of the background are different, the desired value is different. An example where the above problem occurs will be described with reference to FIG.

  FIG. 5A shows an encoding target image 501 and a reference image 502 that is an image immediately before the encoding target image. Further, FIG. 5B is a diagram showing a global vector 503 that is a movement amount of the entire image to be encoded 501 and the reference image 502.

  As shown in FIG. 5, the subject (person) moves from left to right in the screen between the reference image 502 and the encoding target image 501, but the background moves from right to left. In such a case, there is a possibility that the search area is wrong.

  In the case shown in FIG. 5, the ratio of the background moving from right to left in the entire screen is larger than the area of the person moving from left to right in the opposite direction to the background. The global vector that is the amount of motion is as indicated by an arrow 503.

  Here, as shown in FIG. 5C, when the block 504 in the region indicating the person in the encoding target image is the encoding target block, the position of the encoding target block 504 is set as the origin. Then, an area indicating the position of a predetermined pixel in the horizontal and vertical directions around the position indicated by the global vector 503 is set as a motion vector search area.

  In the case of FIG. 5C, an incorrect position is searched by showing the opposite direction of the original person's movement. Therefore, even if the block with the smallest error in the search area is calculated, it is different from the original search area, so that a motion vector is erroneously detected and a high compression effect cannot be expected.

  The present invention has been made in view of the above-described problems, and an object of the present invention is to improve the accuracy of a motion vector search range in target search area motion vector detection.

  The motion vector detection apparatus of the present invention includes an important area detection unit that sequentially detects an important area of an image in each of an input encoding target image and a reference image, and the important area detected by the important area detection unit. An important region motion amount detection unit for detecting a motion amount, a global vector detection unit for calculating a global vector indicating an overall motion amount between the encoding target image and a reference image, and the important region motion amount detection unit The motion in the reference image in the encoding target block obtained by dividing the encoding target image into a plurality of pixels based on the amount of motion of the important region calculated by the global vector calculated by the global vector detecting unit. Search area determining means for determining a vector search area based on a predetermined condition.

  In the motion vector detection method of the present invention, an important region detection step of sequentially detecting an important region of an image in each of an input encoding target image and a reference image, and the important region detected in the important region detection step An important region motion amount detecting step for detecting a motion amount; a global vector detecting step for calculating a global vector indicating an overall motion amount between the encoding target image and the reference image; and the important region motion amount detecting step. The motion in the reference image in the encoding target block obtained by dividing the encoding target image into a plurality of pixels based on the amount of motion of the important region calculated in step 1 and the global vector calculated in the global vector detection step And a search area determining step for determining a vector search area based on a predetermined condition.

The program according to the present invention includes an important area detection step of sequentially detecting an important area of an image in each of an input encoding target image and a reference image, and an amount of motion of the important area detected in the important area detection step. An important area motion detection process to detect;
A global vector detecting step for calculating a global vector indicating an overall amount of motion between images from the encoding target image and a reference image; a motion amount of the important region calculated in the important region motion amount detecting step; A search for determining a motion vector search area in a reference image based on a predetermined condition in an encoding target block obtained by dividing the encoding target image into a plurality of pixels based on a global vector calculated in a global vector detection step The area determining step is executed by a computer.

  According to the present invention, it is possible to improve the prediction accuracy by improving the accuracy of the motion vector search range in the motion vector detection of the important region, and thereby to improve the image quality.

(First embodiment)
Hereinafter, the configuration of a video encoding device using a preferred embodiment of the present invention will be described with reference to the block diagram of FIG.
As illustrated in FIG. 1, in the video encoding device 100 according to the present embodiment, a difference calculation unit 101 calculates a difference value between an input image and a predicted image generated by the predicted image generation unit 109. The predicted image generation method in the predicted image generation unit 109 will be described later.

  Next, in the orthogonal transform unit quantization unit 102, the orthogonal value conversion processing by the integer precision discrete cosine transform and the discrete Hadamard transform is performed on the difference value data between the input image output from the difference calculation unit 101 and the predicted image in units of pixel blocks. I do. Further, quantization processing is performed on the data after orthogonal transformation. The pixel block unit is 8 × 8 pixel or 4 × 4 pixel block unit.

  Next, the entropy encoding unit 103 performs entropy encoding processing such as CAVLC or CABAC on the input quantized data, and outputs encoded data. CAVLC is Context-based Adaptive Variable Length Coding. CABAC is Context-based Adaptive Binary Arithmetic Coding.

Next, processing related to prediction image generation will be described.
The inverse quantization unit inverse orthogonal transform unit 104 performs inverse quantization processing and inverse orthogonal transform processing on the quantized data output from the orthogonal transform unit quantization unit 102. Furthermore, the data obtained by adding the prediction image output from the prediction image generation unit 109 by the addition unit 110 is output to the deblocking filter processing unit 105 and the memory A106.

  Further, the data subjected to the deblocking filter processing in the deblocking filter processing unit 105 is stored in the memory B107. Here, the data before the deblocking process is stored in the memory A106, and the data after the deblocking filter is stored in the memory B107. The deblocking filter processing unit 105 smoothes only the block boundary of integer conversion and suppresses the occurrence of block noise.

  Next, the prediction method determination unit 108 determines the prediction method of the prediction image using the input image and the data stored in the memory A 106 and the memory B 107. For in-screen coding, data before deblocking filter processing stored in the memory A 106 is used to make a predetermined pixel block unit (for example, 16 × 16 pixel block unit, 8 × 8 pixel block unit, 4 × 4 unit). In pixel block units). Furthermore, encoding efficiency is calculated | required for every predetermined prediction mode.

  On the other hand, for inter-screen coding, the coding efficiency is obtained using the input image and the data after the deblocking filter processing stored in the memory B 107, and the prediction method that optimizes the coding efficiency is determined. In the present embodiment, when the input image is an I slice, only intra-frame coding is employed, and when the input image is a P slice and B slice, the coding efficiency is more optimal than intra-frame coding and inter-frame coding. Select a prediction method.

  By performing the processing as described above, it is possible to output a prediction method and encoding parameters that can obtain optimal encoding efficiency. The predicted image generation unit 109 generates a predicted image according to the prediction method determined by the prediction method determination unit 108. Then, the prediction image generated by the prediction image generation unit 109 is output to the addition unit 110.

Next, the configuration example of the prediction method determination unit 108 in FIG. 1 will be described in more detail with reference to FIG.
2, 201 is equal to the prediction method determination unit 108 shown in FIG. Further, it is assumed that the memory A 106 and the memory B 107 in FIG. 1 are the same as the memory A 207 and the memory B 200 in FIG.

First, the inter-screen prediction search area determination unit 202 in FIG. 2 will be described.
The inter-screen prediction search area determination unit 202 determines a search area (motion vector search region) in the reference image in inter-screen prediction.

  The important part detection unit 203 detects a desired important part in each of the input image which is the current image and the reference image read from the memory B200. When an important part is detected in each of the input image and the reference image, information on the coordinate position in the screen of the important part is also sent to the important part motion amount detection unit 204. Further, when an important part is detected, detection accuracy or reliability of the detected important part is detected, and these pieces of information are also sent to the important part motion amount detection unit 204.

  Next, when the important part motion amount detection unit 204 receives an important part undetected signal from the important part detection unit 203, the important part undetected signal is similarly sent to the search area determination unit 205. Further, when the important part detection signal is received, the important part is detected from the difference value between the images using the coordinate position information of the important part of the input image and the reference image received from the important part detection unit 203 in the same manner as the signal. Is calculated. The important part motion amount and the important part detection signal described above are sent to the search area determining unit 205.

  Next, the global vector detection unit 206 calculates the motion amount of the entire image from the input image and the reference image read from the memory B 200, and sends the result to the search area determination unit 205. Note that the global vector may be calculated for each divided region when the image is divided into several regions, not limited to the entire image. The global vector is a value indicating the overall amount of motion between the encoding target image and the reference image, and is calculated by comparing the entire image or a wide range portion.

  In the search area determination unit 205, when it is determined that there is no important part based on the signal indicating the presence / absence of the important part received from the important part motion amount detection unit 204, the global vector received from the global vector detection unit 206 is obtained. The search area is determined based on this. Then, the determined search area information is sent to the prediction method determination unit 209. Also, in the signal indicating the presence / absence of an important part received from the important part motion amount detection unit 204, when it is determined that there is an important part, the input image input to the prediction method determination unit 201 is within the important part detection area. It is determined whether or not.

  Here, when the input image that has been input is outside the important part detection area, the search area is determined based on the global vector received from the global vector detection unit 206. If the input image is within the important part detection area, the search area is determined based on the important part motion amount information received from the important part motion amount detection unit 204, and the determined search area information is predicted. The data is sent to the method determination unit 209.

  Next, the intra prediction encoding mode determination unit 208 determines an intra encoding prediction mode from the input image and the reference image read from the memory A 207, and sends it to the prediction method determination unit 209. The detailed configuration of the intra prediction encoding mode determination unit 208 is not directly related to the technology targeted by the video encoding device 100 of the present embodiment, and thus detailed description thereof is omitted. The prediction method determination unit 209 selects and outputs an optimal prediction method based on the intra prediction information and the inter prediction information.

  In the above-described embodiment, the method of detecting a motion vector using the motion vector search area (motion vector search area) determination method using the global vector has been described. For example, a method may be used in which the motion vector of the encoding target block is predicted from the motion vectors of the encoded blocks around the encoding target block, and the motion vector search area is determined using the predicted motion vector. .

(Second Embodiment)
Hereinafter, a configuration example of the motion vector detection device of the present invention will be described with reference to the block diagram of FIG.
First, the face area detection unit 301 performs important area detection that sequentially detects the important areas of the image. In the present embodiment, face detection is performed as an important region (important part) in the encoding target image and the reference image. Then, as a detection result, a face area motion amount detection unit 302 provided as an important region motion amount detection unit includes a face detection presence / absence signal, coordinate position information in the image of the face detection region, and information indicating reliability in face detection. Send to.

  When the face area motion amount detection unit 302 detects a face in the encoding target image and the reference image from the face detection presence / absence signal received from the face area detection unit 301, the face area motion amount detection unit 302 uses the respective coordinate position information to determine the face between the images. The amount of movement of the area is calculated. The amount of movement of the face area is calculated from the difference in coordinate position. Then, the amount of movement of the face area, the face detection presence / absence signal, and information indicating the reliability in face detection are sent to the search area determination unit 304. In addition, the global vector detection unit 303 calculates a global vector that is the motion amount of the entire image from the encoding target image and the reference image, and outputs the global vector to the search area determination unit 304.

  Next, the search area determination unit 304 determines a motion vector search area in the reference image in the encoding target block obtained by dividing the encoding target image into a plurality of pixels. Detailed processing of the search area determination unit 304 will be described later.

  Next, the reference block generation unit 305 generates a reference block obtained by dividing the reference image into a plurality of pixels. Here, a reference block is generated from the search area determined by the search area determination unit 304.

  Next, the difference calculation unit 306 calculates a difference value in pixel units between the encoding target block and the reference block generated by the reference block generation unit 305. Finally, the motion vector calculation unit 307 uses the result of calculating the difference value for each reference block in the search area to determine the reference block having the smallest difference value, and moves the difference value with the reference block to the motion. Determine and output as a vector.

Next, the process performed by the search area determination unit 304 described above will be described in detail using the flowchart of FIG.
As shown in FIG. 4, first, in step S401, a signal indicating presence / absence of face detection input by the face area motion amount detection unit 302 in FIG. 3 is detected in the encoding target image and the reference image. It is determined whether or not a signal indicating this is input. If a face detection signal is input as a result of this determination, the process proceeds to step S402. If no face detection signal is input, the process proceeds to step S403.

  In step S402, it is determined whether all or part of the encoding target block is within the face detection area. As a result of this determination, if it is determined that all or part of the encoding target block is within the face detection area, the process proceeds to step S404. If it is determined in the above-described determination that it is not within the face detection area, that is, if the encoding target block is not included in the face detection area, the process proceeds to step S403.

  In step S403, a motion vector search area is determined using the global vector input from the global vector detection unit 303 in FIG. In step S404, a motion vector search area is determined and output using the motion amount of the face detection area input from the face detection motion amount detection unit 302 shown in FIG.

Next, the search area determination processing in step S403 and step S404 in the flowchart of FIG. 4 will be described with reference to FIGS.
First, with reference to FIG. 6, a description will be given of a motion vector search area determination process using a global vector, which is a process performed in step S403 in the flowchart of FIG.

  FIG. 6A is a diagram illustrating a global vector 603 that is a movement amount of the entire image in the encoding target image 601 and the reference image 602. Further, FIG. 6B is a diagram showing a motion vector search area 605 centered on the position indicated by the global vector 603 with the encoding target block 604 as the origin.

  As shown in FIGS. 6A and 6B, the motion vector search area determination method for the encoding target block outside the face detection region is determined by a method using a global vector.

Next, the motion vector search area determination process using the motion amount of the face detection area performed in step S404 in the flowchart of FIG. 4 will be described with reference to FIG.
In FIG. 7A, face detection areas 703 and 704 are shown for the encoding target image 701 and the reference image 702, respectively.

  Further, FIG. 7B shows the face detection area motion amount 705 in the encoding target image 701 and the reference image 702. The face detection area motion amount 705 is calculated from the difference value of the coordinates of the face detection areas 703 and 704 in each image.

  Next, in FIG. 7C, in the encoding target block 706 in the face detection area, the motion vector search area centering on the position 707 indicated by the face detection area motion amount 705 with the encoding target block 706 as the origin. 704 is shown. As described above, the motion vector search area 704 is determined using the face detection area motion amount 705 for the encoding target block in the face detection area.

Next, control of a motion vector search area determination method using reliability in face detection will be described with reference to the flowchart of FIG.
In step S901 in FIG. 9, a signal indicating the presence or absence of face detection input by the face region motion amount detection unit 302 in FIG. 3 is input as a signal indicating that a face has been detected in the encoding target image and the reference image. It is determined whether or not it has been done. As a result of this determination, if it is input, the process proceeds to step S902. If no face is detected as a result of face detection in the encoding target image or the reference image, the process proceeds to step S904.

  In step S902, it is determined whether all or part of the encoding target block is within the face detection area. As a result of this determination, if it is determined that all or part of the encoding target block is within the face detection area, the process proceeds to step S903. If NO in step S902, that is, if the encoding target block is not included in the face detection area, the process proceeds to step S904.

  In step S903, it is determined whether or not the input face detection reliability is equal to or higher than a predetermined threshold. As a result of the determination, if the reliability of face detection is equal to or greater than a predetermined threshold value, the process proceeds to step S905. If the face detection reliability is less than a predetermined threshold value, the process proceeds to step S904.

  Next, in step S904, a motion vector search area is determined using the global vector input from the global vector detection unit 303 in FIG. In step S905, a motion vector search area is determined and output using the motion amount of the face detection area input from the face detection motion amount detection unit 302 shown in FIG.

Next, control of the motion vector search area determination method using the comparison result between the motion amount of the face detection area and the global vector will be described with reference to the flowchart of FIG.
As shown in FIG. 10, in step S1001, a signal indicating presence / absence of face detection input by the face area motion amount detection unit 302 in FIG. 3 indicates that a face is detected in the encoding target image and the reference image. It is determined whether or not the indicated signal is input. As a result of this determination, if it is input, the process proceeds to step S1002. If no face is detected as a result of face detection in the encoding target image or the reference image, the process proceeds to step S1004.

  In step S1002, it is determined whether all or part of the encoding target block is within the face detection area. As a result of this determination, if it is determined that all or part of the encoding target block is within the face detection area, the process proceeds to step S1003. If the determination is negative, that is, if the encoding target block is not included in the face detection area, the process proceeds to step S1004.

  In step S1003, a difference value between the motion amount of the face detection area and the global vector is calculated, and it is determined whether or not the difference value is equal to or greater than a predetermined threshold value. If the result of this determination is that the difference value is greater than or equal to a predetermined threshold value, the process moves to step S1005. If the difference value is less than a predetermined threshold value set in advance, the process proceeds to step S1004.

  Next, in step S1004, a motion vector search area is determined using the global vector input from the global vector detection unit 303 in FIG. In step S1005, a motion vector search area is determined and output using the motion amount of the face detection area input from the face detection motion amount detection unit 302 shown in FIG.

Next, a control method for dividing the encoding target image into a predetermined number and determining a motion vector search area for each divided image will be described.
In the present embodiment, the ratio of the face detection area is determined for each divided area (each divided block), and if the ratio of the face detection area is equal to or greater than a predetermined threshold value, A motion vector search area determination method using the amount of motion of the face detection area is used.

  If the ratio of the face detection area in the divided area is less than a predetermined threshold value, a motion vector search area determination method using a global vector is used for the divided area. For each area, an optimal reference block is selected for the search area obtained by the motion vector search area determination method, and a motion vector is detected.

Next, processing when a plurality of face areas are detected in the face area detection unit 301 of FIG. 3 will be described.
When a plurality of face areas are detected as described above, the face area motion amount detection unit 302 in FIG. 3 detects each face detection area in the encoding target image and each face detection area in a plurality of reference images. The amount of motion in the face detection area is calculated. Then, only the minimum value among the calculated motion amounts of the plurality of face detection areas is assigned.

  Next, in the search area determination unit 304 in FIG. 3, when the encoding target block is entirely or partially included in the face detection area, the motion amount of the face detection area assigned to the face detection area is calculated. And a motion vector search area is obtained. The above means realizes effective motion vector detection even when a plurality of faces are detected in face area detection.

(Another embodiment according to the present invention)
Each means constituting the motion vector detection device in the above-described embodiment of the present invention can be realized by operating a program stored in a RAM or ROM of a computer. This program and a computer-readable recording medium recording the program are included in the present invention.

  In addition, the present invention can be implemented as a system, apparatus, method, program, storage medium, or the like, and can be applied to a system composed of a plurality of devices. Moreover, you may apply to the apparatus which consists of one apparatus.

  In the present invention, a software program (in the embodiment, a program corresponding to the flowcharts shown in FIGS. 4, 9, and 10) that executes each step in the motion vector detection method described above is directly transferred to a system or apparatus. Alternatively, it is supplied from a remote location. In addition, this includes a case where the system or the computer of the apparatus is also achieved by reading and executing the supplied program code.

  Accordingly, since the functions of the present invention are implemented by computer, the program code installed in the computer also implements the present invention. In other words, the present invention includes a computer program itself for realizing the functional processing of the present invention.

  In that case, as long as it has the function of a program, it may be in the form of object code, a program executed by an interpreter, script data supplied to the OS, and the like.

  Various recording media can be used as a recording medium for supplying the program. For example, floppy (registered trademark) disk, hard disk, optical disk, magneto-optical disk, MO, CD-ROM, CD-R, CD-RW, magnetic tape, nonvolatile memory card, ROM, DVD (DVD-ROM, DVD- R).

  As another program supply method, a browser on a client computer is used to connect to an Internet home page. The computer program itself of the present invention or a compressed file including an automatic installation function can be downloaded from the homepage by downloading it to a recording medium such as a hard disk.

  It can also be realized by dividing the program code constituting the program of the present invention into a plurality of files and downloading each file from a different homepage. That is, a WWW server that allows a plurality of users to download a program file for realizing the functional processing of the present invention on a computer is also included in the present invention.

  In addition, the program of the present invention is encrypted, stored in a storage medium such as a CD-ROM, distributed to users, and key information for decryption is downloaded from a homepage via the Internet to users who have cleared predetermined conditions. Let It is also possible to execute the encrypted program by using the key information and install the program on a 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 performs part or all of the actual processing. Also, the functions of the above-described embodiments can be realized.

  Further, the program read from the recording medium is written in a memory provided in a function expansion board inserted into the computer or a function expansion unit connected to the computer. Thereafter, the CPU of the function expansion board or function expansion unit performs part or all of the actual processing based on the instructions of the program, and the functions of the above-described embodiments are realized by the processing.

1 is a block diagram of a video encoding device according to a first embodiment. FIG. It is a block diagram which shows 1st Embodiment and demonstrates the structural example of a prediction method determination part. It is a block diagram which shows 2nd Embodiment and shows the structural example of the motion vector detection part using face detection. It is a flowchart which shows 2nd Embodiment and demonstrates the procedure of a motion vector search area determination process. It is a figure which shows a prior art example and demonstrates the case where a to-be-photographed object and the movement direction of a background differ. It is a figure which shows 2nd Embodiment and demonstrates the determination process of the motion vector search area using a global vector. It is a figure which shows 2nd Embodiment and demonstrates a motion vector search area determination process. It is a figure explaining the block matching method used as a method of detecting a motion vector. It is a flowchart which shows 2nd Embodiment and demonstrates the control procedure of the motion vector search area determination method using the reliability in face detection. 10 is a flowchart illustrating a control procedure of a motion vector search area determination method using a comparison result between a motion amount of a face detection region and a global vector according to the second embodiment.

Explanation of symbols

DESCRIPTION OF SYMBOLS 100 Video coding apparatus 101 Difference calculation part 102 Orthogonal transformation part Quantization part 103 Entropy coding part 104 Inverse quantization part Inverse orthogonal transformation part 105 Deblocking filter process part 106 Memory A
107 Memory B
108 Prediction method determination unit 109 Prediction image generation unit 110 Addition unit 200 Memory B
201 prediction method determining unit 202 inter-screen prediction search area determining unit 203 important part detecting unit 204 important part motion amount detecting unit 205 search area determining unit 206 global vector detecting unit 207 memory A
208 Intra-screen predictive coding mode determination unit 209 Prediction method determination unit 301 Face region detection unit 302 Face region motion amount detection unit 303 Global vector detection unit 304 Search area determination unit 305 Reference block generation unit 306 Difference calculation unit 307 Motion vector calculation unit

Claims (12)

An important area detecting means for sequentially detecting an important area of the image in each of the input encoding target image and the reference image;
An important region motion amount detection means for detecting a motion amount of the important region detected by the important region detection means;
Global vector detecting means for calculating a global vector indicating an overall motion amount between images from the encoding target image and a reference image;
An encoding target block obtained by dividing the encoding target image into a plurality of pixels based on the amount of motion of the important region calculated by the important region motion amount detection unit and the global vector calculated by the global vector detection unit. And a search area determining means for determining a motion vector search area in the reference image based on a predetermined condition.   The search area determining means determines a motion vector search area based on the amount of motion of the important area calculated by the important area motion amount detecting means when all or part of the encoding target block is included in the important area. The motion vector detection device according to claim 1, wherein the motion vector detection device is determined. Having reliability detection means for detecting the accuracy or reliability of the detection result in the important area detection means,
The search area determining unit is configured to calculate the amount of motion of the important region calculated by the important region motion amount detecting unit when the accuracy or reliability detected by the important region detecting unit is equal to or higher than a preset threshold. The motion vector detection device according to claim 1, wherein a motion vector search region is determined. A difference calculating means for calculating a difference value between the amount of motion of the important region and the global vector;
The search area determining means determines the amount of motion of the important area calculated by the important area motion amount detecting means when the difference value between the amount of motion of the important area and the global vector is equal to or greater than a preset threshold value. The motion vector detection device according to claim 1, wherein a motion vector search region is determined. A determination means for determining a ratio of the important area for each divided area;
The search area determining means determines a motion vector search area based on the amount of motion of the important area calculated by the important area motion amount detecting means when the proportion of the important area is equal to or greater than a predetermined threshold. The motion vector detection device according to claim 1, wherein the motion vector detection device is determined. In each of the input encoding target image and the reference image, an important region detection step of sequentially detecting an important region of the image,
An important region motion amount detection step for detecting a motion amount of the important region detected in the important region detection step;
A global vector detection step of calculating a global vector indicating an overall motion amount between images from the encoding target image and a reference image;
An encoding target block obtained by dividing the encoding target image into a plurality of pixels based on the amount of motion of the important region calculated in the important region motion amount detection step and the global vector calculated in the global vector detection step. And a search area determining step of determining a motion vector search area in the reference image based on a predetermined condition.   In the search area determination step, when all or a part of the encoding target block is included in the important region, a motion vector search region is determined based on the amount of motion of the important region calculated in the important region motion amount detection step. The motion vector detection method according to claim 6, wherein the motion vector detection method is determined. A reliability detection step of detecting the accuracy or reliability of the detection result in the important region detection step;
In the search area determination step, when the accuracy or reliability detected in the important region detection step is greater than or equal to a preset threshold value, the search area determination step uses the motion amount of the important region calculated in the important region motion amount detection step. The motion vector detection method according to claim 6, wherein a motion vector search region is determined. A difference calculating step of calculating a difference value between the amount of motion of the important region and the global vector,
In the search area determination step, when the difference value between the motion amount of the important region and the global vector is equal to or larger than a preset threshold value, the search area determination step uses the motion amount of the important region calculated in the important region motion amount detection step. The motion vector detection method according to claim 6, wherein a motion vector search region is determined. A determination step of determining the proportion of the important area for each divided area;
The search area determination step determines a motion vector search region based on a motion amount of the important region calculated in the important region motion amount detection step when a ratio of the important region is equal to or more than a predetermined threshold value set in advance. The motion vector detection method according to claim 6, wherein the motion vector detection method is determined. In each of the input encoding target image and the reference image, an important region detection step of sequentially detecting an important region of the image,
An important region motion amount detection step for detecting a motion amount of the important region detected in the important region detection step;
A global vector detection step of calculating a global vector indicating an overall motion amount between images from the encoding target image and a reference image;
An encoding target block obtained by dividing the encoding target image into a plurality of pixels based on the amount of motion of the important region calculated in the important region motion amount detection step and the global vector calculated in the global vector detection step. And a search area determining step for determining a motion vector search area in a reference image based on a predetermined condition.   A computer-readable storage medium storing the program according to claim 11.

Images