KR100516711B1 - Device and method for motion vector detection - Google Patents

Abstract

The present invention makes it possible to realize motion vector detection with high precision for an image having various features.

The motion vector detection unit 30 detects a motion vector of a block to be encoded from a degree of correlation with candidate blocks within a search range. The search range storage memory 20 stores the pixel data of the original search range, and supplies the pixel data of the search range to the motion vector detector 30 according to the memory address MA. The control unit 40 is supplied from the motion vector detection unit 30, sets the size and position of the search range based on the stored motion vector and the correlation evaluation value, and the pixel width in the horizontal and vertical directions of the search range. And offset values representing relative positions to the original search range. The address generator 10 generates a memory address MA which is assigned to the search range storage memory 20 based on the pixel width and the offset value of the search range generated by the controller 40.

Description

Motion vector detection device and motion vector detection method {DEVICE AND METHOD FOR MOTION VECTOR DETECTION}

The present invention relates to motion vector detection used in motion compensation predictive encoding, which is a compression encoding method of digital video data.

As a method of realizing image coding of a moving image, there is a method of reducing temporal redundancy by using information (motion vectors) indicating which part of the current screen has moved from one place of the previous screen. One method for extracting this motion vector is a block matching method.

16 shows the principle of the block matching method. The target frame image (target frame), which is the object of encoding, is compared with the search frame image (search frame) for searching the motion vector, so that the block most similar to the encoding target block of the target frame (that is, the highest correlation) (the best matching block) ) Is extracted from within the search range of the search frame to detect the motion vector. This block matching method is widely used for motion compensation prediction of picture compression coding.

In the block matching method, a correlation between a plurality of candidate blocks in a search range of a search frame is calculated for one encoding target block, and a candidate block having the highest correlation is selected as the best matching block, and the best matching block and the encoding target The difference in the position of the block is detected as a motion vector.

As an evaluation value of a correlation degree, the square error (Square Error) and the absolute value error (Absolute Error) are used normally. Since the difference value between the block to be encoded and the candidate block becomes smaller as the correlation is higher, when the accumulated value of the square error or the accumulated value of the absolute value error is used as the evaluation value, the higher the correlation, the smaller the evaluation value. do.

The purpose of motion vector detection is to detect from the search frame a block that can most effectively reduce the amount of code in motion compensated prediction coding. Therefore, the detection accuracy of the motion vector is one of the important performances of the motion vector detection apparatus. However, motion vector detection is a process that requires an operation amount, especially among image coding algorithms. When this is realized by hardware, there is a problem that the power consumption increases as the processing time increases or the hard scale increases. In the present state, real-time processing by software is very difficult. From these backgrounds, motion vector detection with high detection accuracy and high processing efficiency is required.

As a conventional motion vector detection apparatus, a search range is reduced for an image having a small motion, and a processing time for detecting a motion vector is shortened, while a search range is increased for an image having a large motion, thereby improving processing efficiency without decreasing detection accuracy. There is a plan.

For example, the motion vector detection apparatus disclosed in Japanese Patent Application Laid-open No. Hei 8-32969 adaptively adjusts the size of the search range according to the means for performing correlation evaluation at the same position as the processing block and the correlation evaluation result at the same position. Means for performing selection control and means for performing motion vector detection by evaluating the correlation between the candidate blocks in the search range, and reducing the search range as the correlation of the same position increases. As a result, the motion vector can be detected with high accuracy.

However, the conventional motion vector detection apparatus has the following problems.

In images taken by cameras or the like, so-called pan images in which the entire screen moves in a constant direction are frequently used. In the pan image, the object itself reflected on the screen does not change and its position moves in a constant direction.

In the conventional motion vector detection apparatus, since the size of the search range is set by correlation evaluation at the same position as the processing block, for example, since the correlation at the same position is low for the fan image, the search range is always large. Will be set. In addition, since the same position correlation evaluation is performed before motion vector detection, there is a problem that the processing becomes complicated.

In view of the above problems, an object of the present invention is to make it possible to efficiently and accurately perform motion vector detection on an image having various features, and in particular, to perform motion vector detection on a pan image with high speed and accuracy. .

In order to solve the above problems, the present invention seeks to provide a motion vector detection apparatus, which is used to calculate an evaluation value indicating a degree of correlation between an encoding target block of a target frame and a candidate block within a search range of a search frame. Motion vector detecting means for detecting a motion vector for the encoding target block based on the evaluation value, and pixel data in the original search range are stored, and the pixel data is read in accordance with a given memory address, and the motion vector is detected. The original search range storage means to be supplied to the means, a first search width which is a pixel width in the horizontal direction of the search range, a second search width which is a pixel width in the vertical direction, and a relative position with respect to the original search range of the search range. Control means for generating an offset value, and the original search range based on the first and second search widths and offset values generated by the control means. Is provided with an address generating means for generating a memory address which is given to the memory means.

According to the above arrangement, based on the first and second search widths and the offset values generated by the control means, the address generating means generates a memory address corresponding to the search range. Using the pixel data read from the original search range storage means in accordance with this memory address, the motion vector detecting means calculates an evaluation value indicating the correlation between the encoding target block and the candidate block in the search range, and detects the motion vector. Thus, the size and position of the search range can be freely set, and motion vector detection can be performed efficiently and with high accuracy.

In the motion vector detecting apparatus, the motion vector detecting means supplies the detected motion vector and the correlation evaluation value corresponding to the motion vector to the control means, and the control means is provided from the motion vector detecting means. The first and second search widths and offset values are generated based on the supplied motion vectors and the correlation evaluation values.

According to the above configuration, since the first and second search widths and the offset values of the search range are generated based on the motion vectors and the evaluation values already detected, the first and second search widths and the offset values are generated before the motion vector detection to determine the size and position of the search range. There is no need to perform the processing. In addition, since the motion vector indicates the motion of the image, the search range is set based on the motion vector and the corresponding correlation evaluation value, so that the proper search range is set, particularly for a pan image in which the image moves in a constant direction. This becomes possible.

The control means in the motion vector detection apparatus uses a block at the same position as or near the encoding target block in the frame preceding the target frame as the range setting reference block, and the correlation evaluation value corresponding to the range setting reference block. The size of the search range is set based on.

According to the above configuration, the degree of association with the search frame of the encoding target block is determined by a correlation evaluation value corresponding to a range setting reference block at the same position or in the vicinity of the encoding target block in the frame preceding the target frame. Can be predicted to some extent, and by setting the size of the search range based on the correlation evaluation value of this range setting reference block, it is possible to set an appropriate search range particularly for a pan image in which the image moves in a constant direction.

The control means in the motion vector detecting apparatus uses a block at the same position as or near the encoding target block in the frame preceding the target frame as the range setting reference block and is based on the motion vector corresponding to the range setting reference block. It is assumed that the position of the search range is set.

According to the above configuration, the motion of the encoding target block can be predicted to some extent by a motion vector corresponding to the range setting reference block at the same position or in the vicinity of the encoding target block in the frame before the target frame. By setting the position of the search range by the motion vector of the setting reference block, it is possible to set an appropriate search range, particularly for a pan image in which the image moves in a constant direction.

The control means in the motion vector detection apparatus uses a block at the same position as or near the encoding target block in the frame preceding the target frame as the range setting reference block, and the correlation evaluation value corresponding to the range setting reference block. When the correlation is relatively low, the original search range is set as the search range, while when the correlation degree is relatively high, the size of the search range is set smaller than the original search range, and the position of the search range is set to the range setting reference block. It is set based on the corresponding motion vector.

The offset value in the motion vector detection apparatus is a component of a vector whose end point is the search point in the original search range and whose end point is the search point in the search range.

The address generating means in the motion vector detecting apparatus compares the count value of the first counter with a first counter for counting a clock and compares the count value of the first search width with the first counter. A comparator for resetting and outputting a count-up signal, a second counter for counting the count-up signal, and the original value according to a predetermined calculation formula based on the offset value and the count value of the first and second counters. It is assumed that a memory address conversion section for generating a memory address supplied to the search range storage means is provided.

Further, in the motion vector detecting apparatus, the control means may include the first subsample number and the vertical sub-number indicating the subsample degree in the horizontal direction of the search range in addition to the first and second search widths and offset values. Generating a second subsample number indicating a sample degree, wherein the address generating means generates the first and second search widths and the first and second search widths generated by the control means to generate a memory address given to the original search range storage means. And based on the second subsample number and the offset value.

The present invention also provides a method for detecting a motion vector with respect to an encoding target block, comprising: a first step of determining whether a portion similar to the encoding target block exists in a predetermined region of a search frame; A second step of setting a first area including this similar part in the search frame and smaller than the predetermined area as a search range for motion vector detection when it is determined that there is a similar part in the first step; When it is determined in step 1 that there is no similar part, a third step of setting an area larger than the first area as a search range for motion vector detection is provided.

According to the above arrangement, it is determined whether or not there is a portion similar to the encoding target block in a predetermined region in the search frame, and when it is determined, the region includes the similar portion and is smaller than the predetermined region. Is set as a search range for motion vector detection, so that an appropriate search range can be set, particularly for a pan image in which the image moves in a constant direction. In this way, motion vector detection can be performed efficiently and with high accuracy.

In the motion vector detection method, in the first step, a block is set as a range setting reference block at or near the same position as the encoding target block in the frame before the target frame, and the range setting reference block is detected. The determination is made based on the correlation evaluation value corresponding to the motion vector, and in the second step, the position of the search range is set based on the motion vector detected with respect to the range setting reference block.

Further, in the motion vector detection method, the first step includes a unit area including at least a part of an encoding target block and the same size as the unit region provided in the vicinity of the same position as the encoding target block in the search frame. Predetermined pixel operations are performed on the plurality of comparison areas of the plurality of comparison areas, and the determination is performed based on these calculation results, and the second step is the unit of the plurality of comparison areas from the calculation result of the predetermined pixel operation. The most similar thing to the area is obtained, and the position of the search range is set based on the position of this comparison area.

The present invention also provides a motion vector detection method, comprising: setting a search range for detecting a motion vector, and determining a motion vector based on a correlation between candidate blocks within the search range with respect to an encoding target block. And a step of detecting the motion range, wherein the search range setting step uses a block at the same position as or near the encoding target block in the frame preceding the target frame as the range setting reference block, and detects the motion with respect to the range setting reference block. The size of the search range is set based on the correlation evaluation value corresponding to the vector.

According to the above-described configuration, the degree of association with the search frame of the encoding target block is based on the correlation evaluation value corresponding to the range setting reference block at the same position or in the vicinity of the encoding target block in the frame before the target frame. Can be predicted to some extent, so that the appropriate search range can be set by setting the size of the search range based on the correlation evaluation value of this range setting reference block.

The present invention also provides a motion vector detection method, comprising: setting a search range for detecting a motion vector, and determining a motion vector based on a correlation between candidate blocks within the search range with respect to an encoding target block. And a step of detecting the motion range, wherein the search range setting step uses a block at the same position as or near the encoding target block in the frame preceding the target frame as the range setting reference block, and detects the motion with respect to the range setting reference block. The position of the search range is set based on the vector.

According to the above configuration, the motion of the encoding target block can be predicted to some extent by a motion vector corresponding to the range setting reference block at the same position or in the vicinity of the encoding target block in the frame preceding the target frame. By setting the position of the search range by the motion vector of the range setting reference block, an appropriate search range can be set.

The above and other objects and features and advantages of the present invention will become more apparent from the following detailed description taken in conjunction with the accompanying drawings.

(Example)

Hereinafter, embodiments of the present invention will be described with reference to the drawings.

(First embodiment)

FIG. 1 is a diagram showing three frame images consecutive in time, and shows a so-called pan image photographed by moving a camera from right to left. In fan images, when viewed for a very short time, the object on the screen is moving at a constant speed. Therefore, in the present embodiment, when the motion vector MV for the encoding target block CB of the target frame (frame 1) is detected, it is the same as the encoding target block CB in the search frame (frame 0). The size of the search range is based on the block AB of the position as the range setting reference block and based on the motion vector BMV detected with respect to the range setting reference block AB and the correlation evaluation value BAE at that time. And position. As a result, as shown in FIG. 1, the search range for motion vector detection can be set by narrowing to the position where the image of the encoding object block CB exists in the search frame. Therefore, motion vector detection can be performed accurately and at high speed.

2 is a schematic diagram showing a search range for motion vector detection according to the present embodiment. In this embodiment, the size of the search range is two types. FIG. 2 shows an example of a search range for a coding target block CB of a general size (16 pixels in the horizontal direction and 16 pixels in the vertical direction) in an image coding standard such as MPEG, and FIG. The search range of the horizontal direction 46 pixels and the vertical direction 46 pixels as an example of a large search range W is shown, (b) has shown the search range of the horizontal 30 pixels and 30 pixels of a vertical direction as an example of a small search range S. .

The position of the small search range S is indicated according to the relative position with respect to the large search range W. FIG. In this case, a vector having the upper left corner as the search point in the search range W as the start end and the upper left corner as the search point in the search range S as the end is referred to as offsets (Hoff, Voff). According to the offset value, it is assumed that the position of the search range S is shown. At this time, for example, when the image is a still image and the search range S is set at the position of the encoding target block CB, the offset value is as shown in FIG.

Hoff = 8

Voff = 8

Becomes The pixel width in the horizontal direction of the search range is set to the search horizontal (H) pixel width as the first search width, and the pixel width in the vertical direction of the search range is set to the search vertical (V) pixel width as the second search width.

3 is a diagram showing the configuration of a motion vector detection apparatus according to the present embodiment. 3 to 10 are address generation units as address generation means for generating a memory address MA based on search H pixel width, search V pixel width, and offset, and 20 are original search ranges for storing image data of the original search range. The search range storage memory 30 as the storage means calculates an evaluation value indicating the correlation between the encoding target block and the candidate block in the search range, and outputs a motion vector corresponding to the maximum correlation and an evaluation value corresponding to the motion vector. The motion vector detection unit 40 serving as a motion vector detection unit determines the search H pixel width, the search V pixel width, and the offset from the motion vector and the evaluation value output from the motion vector detection unit 30, and sends them to the address generator 10. The control unit 50 as a control means for outputting is an encoding target block storage memory for storing image data of the encoding target block. Here, the large search range W of 46 x 46 pixels is used as the original search range.

4 is a flowchart illustrating the operation of the controller 40. An operation of the controller 40 will be described with reference to the flowchart of FIG. 4 and FIG. 1.

Now, a search range for motion vector detection is set for the encoding target block CB in the target frame. The control unit 40 displays a search H pixel width and a search V, which display a search range based on the motion vector BMV of the range setting reference block AB of the search frame and the correlation evaluation value BAE at that time. The small width and the offset are supplied to the address generator 10.

First, in step S1, the correlation evaluation value BAE of the range setting reference block AB is compared with the threshold value AE0. As a result of this comparison, when the evaluation value BAE is larger than the threshold AE0, a portion similar to the encoding target block CB in a predetermined region of the search frame is low because the correlation corresponding to the motion vector BMV is low. The search range W is set as the search range for motion vector detection without determining the search range and narrowing the search range, i.e., the large search range W of 46 x 46 pixels (step S2). In this case, the search H pixel width and the search V pixel width are both "46" and the offset is (0, 0).

When the correlation evaluation value BAE of the ranging reference block AB is not larger than the threshold AE0, the encoding target block ( It is determined that there is a portion similar to CB), and a small search range S of 30 x 30 pixels is set as a search range for motion vector detection (step S3). And the magnitude of the motion vector BMV of the range setting reference block AB in step S4. And threshold MV0.

As a result of this comparison, when the magnitude of the motion vector BMV is larger than the threshold value MV0, it is determined that the motion of the encoding target block CB is large, and the offset is set using the components of the motion vector BMV (step S5). ). Specifically, the values of the components of the offsets Hoff and Voff are set as follows.

Hoff = x + 8

Voff = y + 8

That is, in this case, both the search H pixel width and the search V pixel width are "30", and the offset is (x + 8, y + 8).

On the other hand, when the magnitude of the motion vector BMV is not larger than the threshold value MV0, the motion of the encoding target block CB is determined to be small and a still picture, and the offset is set to (8, 8). In this case, both the search H pixel width and the search V pixel width are set to "30", and the offset is set to (8, 8).

The address generator 10 generates the memory address MA based on the search H pixel width, the search V pixel width, and the offset output from the control unit 40. The motion vector detection unit 30 performs motion vector detection on the encoding target block CB using pixel data of the search range read from the search range storage memory 20 (step S10).

5 is a block diagram showing the configuration of the address generator 10. As shown in Fig. 5, the address generator 10 compares the H counter 11 as the first counter for counting the clock with the count value of the search H pixel width and the H counter 11, and when H matches, In addition to resetting the count value of the counter 11, the H comparator 12 as a comparator for outputting the count up signal CU and the V counter 13 as a second counter for counting the count up signal CU And a V comparator 14 which compares the search V pixel width and the count value of the V counter 13 and generates an operation signal for terminating the operation of the motion vector detection unit 30 when the match is made, and the offset value and the H counter. And a memory address conversion section 15 that generates a memory address MA of the search range storage memory 20 based on the count values of the (11) and V counters (13). In response to the start signal supplied from the control unit 40, the H counter 11 and the V counter 13 are reset, and the address generator 10 starts operation.

The operation of the address generator 10 shown in FIG. 5 will be described with reference to FIGS. 6 to 8.

FIG. 6 is a diagram showing the relationship between the original search range stored in the search range storage memory 20 and the search range for motion vector detection. Here, as shown in FIG. 6, the search range storage memory 20 stores pixel data of 12 pixels x 8 pixels horizontally as the original search range, and the search H pixel width is 8 as the search range for motion vector detection. It is assumed that a search range with a search V pixel width of 6 and an offset of (2, 1) is set.

FIG. 7 is a diagram illustrating a result of mapping two-dimensional pixel data illustrated in FIG. 6 to an one-dimensional address space of the search range storage memory 20. The address generator 10 generates a memory address MA for reading pixel data of a search range for motion vector detection from the search range storage memory 20.

8 is a timing diagram illustrating an operation of the address generator 10. When the start signal is input as shown in Fig. 8, the H counter 11 starts counting the clock. When the count value of the H counter 11 reaches a value corresponding to the search H pixel width, the H comparator 12 outputs a count up signal CU, whereby the count value of the H counter 11 is reset. . The V counter 13 counts the count up signal CU. When the count value of the V counter 13 reaches a value corresponding to the search V pixel width, the V comparator 14 outputs an operation signal.

When the offsets Hoff and Voff are input, the memory address conversion unit 15 moves from the horizontal pixel width Hmax and the vertical pixel width Vmax of the original search range stored in the search range storage memory 20. The address SA0 serving as the starting point of the search range for vector detection is obtained. This starting address SA0 is obtained by the following equation.

SA0 = Hmax × Voff + Hoff

In the example of FIG. 6

SA0 = 12 × 1 + 2 = 14

It becomes

The memory address conversion section 15 obtains the memory address MA based on the following equation.

MA = SA0 + Hmax × (Count value of the V counter 13)

     + Count value of H counter 11

As a result, the memory address MA changes as shown in FIG. For example, as shown in Fig. 7, the memory address MA is obtained by adding the count value of the H counter 11 to the starting address SA0 from the starting address SA0 to the search H pixel width. By the next clock input, the count value of the H counter 11 becomes zero, and the count value of the V counter 13 becomes one. At this time, the memory address MA is

MA = 14 + 12 × 1 + 0 = 26

Becomes

As described above, according to the present embodiment, motion vectors can be detected efficiently and efficiently with respect to the fan image.

In the present embodiment, a block at the same position as the encoding target block in the search frame is used as the range setting reference block. Instead, a block near the same position as the encoding target block in the search frame is used as the ranging reference block. You may use it. Alternatively, a block at the same position as or adjacent to the encoding target block in the frame before the search frame may be used as the range setting reference block.

In the present embodiment, the size of the search range is set to two types, but the size of this or more types may be set. Further, the search H pixel width and the search V pixel width can be changed by using the magnitude of the motion vector component of the range setting reference block as an evaluation index.

(Modification of the first embodiment)

In the first embodiment, subsamples may be merged and applied. In this case, in addition to the search H pixel width, the search V pixel width, and an offset value, the control unit 40 controls the horizontal subsample number DH as the first subsample number in the search range and the vertical direction as the second subsample number. The subsample number DV is input to the address generator 10.

9 is a flowchart showing the operation of the control unit 40 according to the present modification. The difference from FIG. 4 is that the presence or absence of a subsample is specified in addition to setting the size and position of the search range. That is, in step S2A of setting the large search range W and in the step S5A of setting the small search range S other than the position of the encoding target block CB, the subsample is specified and the small search range S is encoded. In step S6A, which is set at the position of the target block CB, no subsample is specified.

10 is a schematic diagram showing a search range for motion vector detection according to the present modification. Here, the number of subsamples DV in the horizontal direction and the number of subsamples DV in the vertical direction are both two. In Fig. 10, (a) shows an example of a large search range W in which subsamples are specified, (b) shows an example of a small search range S in which subsamples are specified, and (c) designates no subsample. An example of a small search range S is shown. The setting of the small search range S at the position of the encoding object block CB in step S6A is because it is determined that the motion of the encoding object block CB is minute. In this case, since it is not desirable to reduce the accuracy of the search, as shown in Fig. 10C, no subsample is designated.

The operation of the address generator 10 according to the present modification will be described with reference to FIGS. 11 to 13.

FIG. 11 is a diagram showing the relationship between the original search range stored in the search range storage memory 20 and the search range for motion vector detection. Here, as shown in FIG. 11, as the search range for motion vector detection, the search H pixel width is 6, the search V width is 4, the offset is (0, 0), and the subsamples (DH, DV) are all 2 It is assumed that the search range is set.

FIG. 12 is a diagram illustrating a result of mapping the two-dimensional pixel data shown in FIG. 11 to the one-dimensional address space of the search range storage memory 20.

13 is a timing diagram illustrating an operation of the address generator 10. In this modification, the memory address conversion section 15 obtains the memory address MA based on the following equation.

MA = SA0 + Hmax × (Count value of V counter 13) × DV

     + (Count value of H counter 11) x DH

As a result, the memory address MA changes as shown in FIG.

In addition, the present invention can be easily applied to motion vector detection corresponding to the rotation and enlargement / reduction of an image by setting the subsamples DH and DV.

(2nd Example)

According to the second embodiment of the present invention, a plurality of blocks having the same size as that of the encoding target block are provided in the vicinity of the same position as the encoding target block in the search frame, and the predetermined blocks are assigned to the encoding target block and the plurality of blocks set in the search frame. Each pixel operation is performed, and the size and position of the search range for motion vector detection are set based on these calculation results. Here, accumulation operation of pixel data in a block is performed as a predetermined pixel operation.

14 is a diagram showing an example of block setting in the search frame according to the present embodiment. In Fig. 14, nine blocks SB0 to SB8 of 16x16 pixels as a comparison area are set in a search frame with respect to the encoding target block CB of 16x16 pixels as a unit area.

The motion vector detecting apparatus according to the present embodiment is configured as shown in FIG. 3 as in the first embodiment. However, the control part 40 does not input the motion vector and evaluation value output from the motion vector detection part 30, but inputs image data of each frame instead.

15 is a flowchart showing the operation of the control unit 40 in the motion vector detection apparatus according to the present embodiment. First, in step S21, accumulation operation of the pixel data in the block is performed on the encoding target block CB and each of the blocks SB0 to SB8 set in the search frame. In step S22, the minimum value SAmin of the difference between the accumulation operation result AC (CB) of the encoding target block and the accumulation operation result AC (SBn) (n = 0 to 8) of each block of the search frame. And threshold SA0 are compared. As a result of this comparison, when the minimum value SAmi is larger than the threshold SA0, a large search of 46x46 pixels is determined without narrowing the search range by determining that there is no portion similar to the encoding target block CB in a predetermined area of the search frame. The range W is set as a search range for motion vector detection (step S23). In this case, the search H pixel width and the search V pixel width are both "46" and the offset is (0, 0).

When the minimum value SAmi is not larger than the threshold SA0, it is determined that there is a portion similar to the encoding target block CB in a predetermined area of the search frame, and a small search range S of 30 x 30 pixels is detected in the motion vector. It is set as the search range for (step S24). In step S25, when the absolute value of the difference between the accumulation operation result of the block SBk and the accumulation operation result of the encoding target block CB becomes minimum, the offset is set so that the search range S is set at the position of the block SBk. Set it.

For example, when the block SB0 is set at the same position as the encoding target block CB, and the blocks SB1 to SB8 are set around it, the accumulation operation result of the block SB0 and the encoding target block CB. When the absolute value of the difference between the accumulated operation results in is minimized, set (8, 8) as the offset. In addition, when the absolute value of the difference between the accumulation operation result of the block SB8 located at the bottom right of the block SB0 and the accumulation operation result of the encoding target block CB becomes minimum, the offset is (16, 16). Set.

As described above, according to the present embodiment, since the offset is set without using the motion vector detected in the previous frame, the offset can be set from the original frame in which the motion vector is detected, so that the motion vector is more efficiently than the first embodiment. Detection can be performed.

In addition, a part of the encoding target block may be set as a unit region, and a plurality of comparison regions having the same size as the unit region may be set, and predetermined pixel operations may be performed on them. It is also possible to include a region near the encoding target block in this unit region. The unit area may be in any shape.

In addition to the accumulation operation as a predetermined pixel operation, for example, an accumulation operation of a difference between adjacent pixels, a dispersion operation of pixels, a frequency distribution operation of pixels, and the like may be used. The calculation result may be weighted according to the position of the region.

As described above, according to the present invention, since the size and position of the search range can be freely set, the motion vector detection can be performed efficiently and with high accuracy. Further, since the search range is set on the basis of the motion vectors already detected and the correlation evaluation value corresponding thereto, the search range can be set appropriately, particularly for a pan image in which the image moves in a constant direction.

In addition, since a region including a portion similar to the encoding target block in the search frame is set as a search range for motion vector detection, an appropriate search range can be set, particularly for a pan image in which the image moves in a constant direction.

In addition, since the size and position of the search range are set by the motion vector and the correlation evaluation value corresponding to the range setting reference block at the same position or in the vicinity of the encoding target block in the frame before the target frame, the appropriate search range is set. This becomes possible.

Preferred embodiments of the present invention are disclosed for purposes of illustration, and those skilled in the art will be able to make various modifications, changes, substitutions and additions through the spirit and scope of the present invention as set forth in the appended claims.

BRIEF DESCRIPTION OF THE DRAWINGS Fig. 1 is a view for explaining a first embodiment of the present invention, showing three frame images in successive time, each representing a fan image.

Fig. 2 is a schematic diagram showing a search range in motion vector detection according to the first embodiment of the present invention, where (a) is an example of a large search range and (b) is an example of a small search range.

3 is a block diagram of a motion vector detection apparatus according to an embodiment of the present invention.

4 is a flowchart showing the operation of the control unit according to the first embodiment of the present invention.

5 is a block diagram showing a configuration of an address generator.

6 is a view showing a relationship between a maximum search range stored in a search range storage memory and a search range for motion vector detection;

FIG. 7 illustrates a result of mapping two-dimensional pixel data shown in FIG. 6 to a one-dimensional address space of a search range storage memory; FIG.

Fig. 8 is a timing diagram showing the operation of the address generator according to the first embodiment of the present invention.

9 is a flowchart showing the operation of the control unit according to the modification of the first embodiment of the present invention.

10 is a schematic diagram showing a search range in motion vector detection according to a modification of the first embodiment of the present invention.

FIG. 11 is a diagram showing a relationship between a circle search range stored in a search range storage memory and a search range for motion vector detection; FIG.

FIG. 12 is a view showing a result of mapping two-dimensional pixel data shown in FIG. 11 to a one-dimensional address space of a search range storage memory; FIG.

Fig. 13 is a timing diagram showing the operation of the address generator according to the modification of the first embodiment of the present invention.

Fig. 14 is a diagram showing an example of setting a comparison area in a search frame according to the second embodiment of the present invention.

15 is a flowchart showing the operation of the control unit according to the second embodiment of the present invention.

16 illustrates block matching.

             Explanation of symbols on main parts of drawing

CB: Encoding block AB: Range setting reference block

BMV: Motion vector of range setting reference block

BAE: Correlation evaluation value of range setting reference block

Hoff, Voff: Offset value MA: Memory address

CU: count up signal

DH: Number of subsamples in the horizontal direction (number of first subsamples)

DV: Number of subsamples in the vertical direction (number of second subsamples)

SB0 to SB8: Block (comparative area)

10: address generator (address generating means)

11: H counter (first counter) 12: H comparator (comparator)

13: V counter (second counter) 15: Memory address conversion unit

20: search range storage memory (original search range storage means)

30: motion vector detection unit (motion vector detection means)

40 control part (control means)

Claims (13)

(correction) A motion vector for calculating a correlation value indicating a correlation between an encoding target block of a target frame and a candidate block within a search range of a search frame, and detecting a motion vector for the encoding target block based on these correlation evaluation values Detection means, Original search range storage means for storing pixel data of the original search range and reading pixel data in accordance with a given memory address and supplying the pixel data to the motion vector detecting means; Control means for generating a first search width that is a pixel width in the horizontal direction of the search range and a second search width that is a pixel width in the vertical direction, and an offset value indicating a relative position with respect to the original search range of the search range; And address generation means for generating a memory address assigned to said original search range storage means on the basis of the first and second search widths and offset values generated by said control means. (correction) The method of claim 1, The motion vector detection means, Supplying the detected motion vector and the correlation evaluation value corresponding to the motion vector to the control means, The control means, And generating a first and a second search width and an offset value based on the motion vector and the correlation evaluation value supplied from the motion vector detecting means. The method of claim 2, The control means, Setting a size of a search range on the basis of a correlation evaluation value corresponding to the range setting reference block, with a range setting reference block having a block at or near the encoding target block in the frame preceding the target frame; Motion vector detection device, characterized in that.  The method of claim 2, The control means, The position of the search range is set based on the motion vector corresponding to the range setting reference block, and the block at the same position as or near the encoding target block in the frame before the target frame is set. Motion vector detection device characterized in that. The method of claim 2, The control means, When the block at the same position as or near the encoding target block in the frame preceding the target frame is a range setting reference block, and the correlation degree indicated by the correlation evaluation value corresponding to the range setting reference block is relatively low, the original search is performed. While the range is set as the search range, when the size is relatively high, the size of the search range is set smaller than the original search range, and the position of the search range is set based on the motion vector corresponding to the range setting reference block. Motion vector detection device characterized in that. The method of claim 1, The offset vector is a motion vector detection device characterized in that it is a vector component whose starting point is the search starting point of the original search range and whose ending point is the search point. The method of claim 1, The address generating means, A first counter for counting a clock, A comparator configured to reset the count value of the first counter and output a count up signal when the first search width and the count value of the first counter are matched to each other; A second counter for counting the count up signal; And a memory address conversion unit for generating a memory address supplied to the original search range storage unit according to a predetermined calculation formula based on the offset value and the coefficient values of the first and second counters. Device. The method of claim 1, The control means, Generating a first subsample number indicating a horizontal subsample degree and a second subsample number indicating a vertical subsample degree in addition to the first and second search widths and an offset value, The address generating means, Generation of a memory address assigned to the original search range storage means is performed based on the first and second search widths, the first and second subsample numbers generated by the control means, and an offset value. Motion vector detector. (correction) A method of detecting a motion vector with respect to an encoding target block, A first step of determining whether there is a part similar to the encoding target block in a predetermined region of a search frame; A second step of setting, as a search range for motion vector detection, a first area including this similar part and smaller than the predetermined area in the search frame when it is determined that there is a similar part in the first step. Wow, And a third step of setting an area larger than the first area as a search range for motion vector detection when it is determined that there is no similar portion in the first step. The method of claim 9, The first step is, Based on the correlation evaluation value corresponding to the motion vector detected with respect to the motion vector detected with respect to this encoding object block in the frame before or similar to the said encoding object block in the frame, and the area | region set to the above, Is to make a judgment, The second step, And a position of a search range is set based on a motion vector detected with respect to the range setting reference block. The method of claim 9, The first step is, Predetermined pixel operations are performed on a unit region including at least part of an encoding target block and a plurality of comparison regions of the same size as the unit region provided in the vicinity of the same position as the encoding target block in the search frame. The above determination is made based on these calculation results. The second step, Obtaining the most similar to the unit region among the plurality of comparison regions from the operation result of the predetermined pixel operation, and setting the position of the search range based on the position of the comparison region. . Setting a search range for motion vector detection; Detecting a motion vector with respect to an encoding target block based on a correlation with a candidate block in the search range, The search range setting step, Search based on the correlation evaluation value corresponding to the motion vector detected with respect to the motion vector detected with respect to this range setting reference block which makes the block which is the same position or the vicinity of this encoding object block in the frame before the target frame. A motion vector detection method, characterized in that the size of the range is set. Setting a search range for motion vector detection; Detecting a motion vector with respect to an encoding target block based on a correlation with a candidate block in the search range, The search range setting step, Setting a position of a search range on the basis of a motion vector detected with respect to the range-referenced block as a range-setting reference block with a block at or near the encoding target block in the frame preceding the target frame. Motion vector detection method characterized in that.