CN102905124A - Motion estimation method - Google Patents
Motion estimation method Download PDFInfo
- Publication number
- CN102905124A CN102905124A CN2011102153991A CN201110215399A CN102905124A CN 102905124 A CN102905124 A CN 102905124A CN 2011102153991 A CN2011102153991 A CN 2011102153991A CN 201110215399 A CN201110215399 A CN 201110215399A CN 102905124 A CN102905124 A CN 102905124A
- Authority
- CN
- China
- Prior art keywords
- territory element
- picture
- subregion
- estimated
- motion
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Landscapes
- Compression Or Coding Systems Of Tv Signals (AREA)
Abstract
The invention discloses a motion estimation method which comprises the following steps of: dividing a first picture to be estimated into a plurality of region units, wherein each region unit comprises a plurality of blocks; and distributing a group of motion vector values to each region unit, wherein the group of motion vector values comprise a plurality of motion vector values; and each of a plurality of given motion vector values is distributed to at least one block in each region unit.
Description
Technical field
The present invention relates to a kind of image treatment method, relate in particular to a kind of mobile estimating (motion estimation) method of dynamic image.
Background technology
In the process of image processing, each picture (frame) in the video image is divided into a plurality of macro block (macroblock, MB), and whole condensing routine disassembled into a plurality of rank (stage), wherein every single order is responsible for processing different condensing routines, mobile estimating and motion compensation (motion estimation and motion compensation for example, MEMC), discrete cosine transform (discrete cosine transform, DCT), variable length code (variable length coding, VLC) and reconstruction (reconstruction) supervisor.
In the process of mobile estimating, according to the reference block (reference MB) of mobile object in the priority picture, and be aided with the motion-vector (motion vector) of new generation, predict mobile object in the macro block position of picture to be estimated, wherein the displacement size of macro block and reference block and direction are motion-vector.
With regard to prior art, when producing new motion-vector, often produce new motion-vector with random update mode.This random update mode provides random number sequence output, and advantage is that hard disk is easily realized, yet shortcoming is in same macro block, and different motion-vectors usually can't be rotated fast.In addition, the mobile object in a certain zone in picture, moment needs the calculation of mobile estimating to restrain rapidly when changing mobile route.Yet, if produce new motion-vector with existing random update mode, often fail to make rapidly mobile object to converge to correct vector.
Be with, but provide a kind of mobile estimating method of Fast Convergent that its necessity is arranged.
Summary of the invention
The invention provides a kind of mobile estimating method, can make mobile object converge to rapidly correct vector.
The present invention proposes a kind of mobile estimating method, and it comprises the steps.One first picture to be estimated is divided into a plurality of territory elements, and each territory element comprises a plurality of blocks.One group of motion-vector value is distributed to each territory element.Wherein, this group motion-vector value comprises a plurality of set motion-vector values, and a plurality of set motion-vector each is assigned at least one block in each territory element in worthwhile.
In one embodiment of this invention, in each territory element, the set motion-vector value that wantonly two adjacent a plurality of blocks are assigned with is not identical.
In one embodiment of this invention, each above-mentioned territory element has variable or the fixing number of blocks of sum.
In one embodiment of this invention, above-mentioned mobile estimating method also comprises provides a question blank.Question blank is used for the set motion-vector value that a plurality of blocks of each territory element of record the first picture to be estimated are assigned to separately.
In one embodiment of this invention, above-mentioned mobile estimating method also comprises reference look-up tables, produces a plurality of blocks motion-vector separately of each territory element of the first picture to be estimated.
In one embodiment of this invention, each territory element of the first above-mentioned picture to be estimated has one or more subregion.Each subregion comprises respectively one or more block, and the one to multiple block of the corresponding subregion of wantonly two territory elements, is to concern to be assigned identical set motion-vector value with a rotation map.
In one embodiment of this invention, the above-mentioned motion-vector value of will organizing is distributed to the step of each territory element and is comprised the steps.One seed region unit is provided, and it comprises a plurality of blocks.A plurality of blocks are assigned respectively a plurality of set motion-vector values.Based on the distribution aspect of seed region unit, and respectively for each territory element a plurality of blocks separately, distribute a plurality of set motion-vector values.
In one embodiment of this invention, above-mentionedly distribute the step of a plurality of set motion-vector values to comprise the steps for each territory element a plurality of blocks separately respectively based on the seed region unit.Be one or more first subregion with the seed region dividing elements, each first subregion comprises one or more block.Each territory element is divided into one or more the second subregion.Each second subregion comprises one or more block, and wherein each second subregion one or more block separately corresponds to one or more first subregion one or more block separately.Respectively based on one or more first subregion distribution aspect separately, and determine that for one or more corresponding in each territory element second subregion it distributes aspect respectively.
In one embodiment of this invention, above-mentioned one or more the first subregion one or more block separately, with one or more second subregion one or more block separately in each territory element, be to concern to be assigned identical set motion-vector value with a rotation map.
In one embodiment of this invention, above-mentioned mobile estimating method also comprises the steps.Based on each territory element distribution aspect separately in the first picture to be estimated, and respectively for territory element corresponding in one second picture to be estimated, distribute a plurality of set motion-vector values.
In one embodiment of this invention, a plurality of blocks of each territory element of the first above-mentioned picture to be estimated with a plurality of blocks of each territory element of the second picture to be estimated, are to concern to be assigned identical set motion-vector value with a rotation map.
Based on above-mentioned, in exemplary embodiment of the present invention, mobile estimating method can be guaranteed the motion-vector that block adjacent in the picture to be estimated is corresponding different, when carrying out mobile estimating, shortens mobile object and converges to the required time of correct vector.
For above-mentioned feature and advantage of the present invention can be become apparent, embodiment cited below particularly, and cooperate accompanying drawing to be described in detail below.
Description of drawings
Fig. 1 is the functional-block diagram of the mobile estimating device of first embodiment of the invention.
Fig. 2 is the handled video image of mobile estimating device of Fig. 1.
Fig. 3 (a) is the picture to be estimated of one embodiment of the invention.
Fig. 3 (b) is the vector information of each digital representative of the motion-vector value of Fig. 3 (a).
Fig. 4 (a) is to the method for salary distribution of Fig. 4 (c) for the motion-vector value of the picture to be estimated of one embodiment of the invention.
Fig. 5 is the flow chart of steps of the mobile estimating method of first embodiment of the invention.
Fig. 6 is the handled video image of mobile estimating device of in a second embodiment Fig. 1.
Fig. 7 (a) is a wherein territory element of the first picture to be estimated.
Fig. 7 (b) is the territory element of the territory element of corresponding the first picture to be estimated in the second picture to be estimated.
Fig. 8 is the flow chart of steps of the mobile estimating method of second embodiment of the invention.
Reference numeral:
100: the mobile estimating device
110: image estimation unit
120: to production units
201～205: image frame
201,203,701,704: reference picture
202,702,703: picture to be estimated
400,802,803: territory element
401a, 402a: the first subregion
401b, 402b, 401c, 402c: the second subregion
S600, S602, S604, S900, S902, S904, S906: the step of mobile estimating method
S: mobile object
Embodiment
The first embodiment
In order to promote the picture quality of video image, the turnover rate of present panel TV (refresh rate) promotes, and is to increase to 120Hz by 60 hertz (Hz) for example.Its principle adopts the mobile estimating technology exactly, inserts a picture to be estimated between two reference pictures, is promoted to 120Hz with the turnover rate with panel TV by 60Hz.So, video image is clear and smooth more, the effect that remove ghost to reach, improves dynamic definition.
Fig. 1 is the functional-block diagram of the mobile estimating device of first embodiment of the invention.Fig. 2 is the handled video image of mobile estimating device of Fig. 1.Please refer to Fig. 1 and Fig. 2, in the present embodiment, mobile estimating device 100 is suitable for estimating the image information of a mobile object in video image.Mobile estimating device 100 handled video images comprise a plurality of image frames 201～205.In this video image, the mobile object in a certain zone in picture, when changing its mobile route, mobile estimating device 100 need to carry out mobile estimating to this mobile object, and it is restrained rapidly.For example, in Fig. 2, when one mobile object S moved to the upper right corner by the lower left corner of video image, mobile estimating device 100 needed carry out mobile estimating for the picture to be estimated 202 between the reference picture 201,203, with the position of prediction mobile object S in picture 202 to be estimated.At this, image frame 201 for example is I picture (intra-coding frame, I frame); Image frame 203 for example is P picture (prediction frame, P frame) 203, but the invention is not restricted to this.
Specifically, in the present embodiment, mobile estimating device 100 comprises that image estimation unit 110 and is to production units 120.Mobile estimating device 100 is after receiving video image, and image estimation unit 110 namely carries out mobile estimating to this video image.At this, image estimation unit 110 for example is according to the image information of reference block in the reference picture 201,203 and the motion-vector that provides to production units 120, and the mobile object S that treats estimation picture 202 carries out mobile estimating.
Fig. 3 (a) is the picture to be estimated of one embodiment of the invention.Fig. 3 (b) is the vector information of each digital representative of the motion-vector value of Fig. 3 (a).Please refer to Fig. 3 (a) and Fig. 3 (b), in the present embodiment, to production units 120 picture 202 to be estimated is divided into a plurality of territory elements 400, each territory element comprises a plurality of blocks.At this, estimation picture 202 for example is divided 32 territory elements 400, and each territory element comprises 8 blocks.
Then, to production units 120 one group of motion-vector value is distributed to each territory element.In the present embodiment, this group motion-vector value comprises a plurality of set motion-vector values, this with " 0 " to " 7 " for illustrating.The numeral that each block indicates namely represents the motion-vector value that this block is assigned to.The specific characteristic of this embodiment is, is that during set motion-vector is worthwhile each is distributed at least one block in each territory element to production units 120.In other words, in any territory element, each all is assigned to one or more block in the middle of the motion-vector value " 0 " to " 7 ", and the situation that is not assigned to occurs.Take Fig. 3 (a) as example, in the territory element 400 in the upper left corner, comprise that one group of set motion-vector value of " 0 " to " 7 " is to be assigned to one to one each block in the territory element 400.And other territory element 400 also can be by that analogy, and difference only is that set motion-vector value is assigned to different positions in each territory element 400.
Preferably, in each territory element, the set motion-vector value that wantonly two adjacent a plurality of blocks are assigned with is not identical.In other words, adjacent block corresponds to different motion-vectors.This method of salary distribution helps further to shorten mobile object and converges to the required time of correct vector.In the following Fig. 4 (a) that will utilize to the related description of Fig. 4 (c), provide a kind of better example method of salary distribution of the motion-vector value of picture to be estimated.
Fig. 3 (b) is the vector information of each motion-vector value representative among Fig. 3 (a), but the present invention is not limited to this.For example, the mobile object that " 0 " represents its block corresponding to be estimated is estimated mobile direction upwards, and it is inferior to the right that the mobile object that " 7 " represent its block corresponding to be estimated is estimated mobile direction, by that analogy.In other embodiments, the vector information of each digital representative can also comprise the size of motion-vector except the direction that comprises motion-vector.
Afterwards, also provide a question blank to production units 120, be used for recording the set motion-vector value that a plurality of blocks of each territory element of picture 202 to be estimated are assigned to separately.At this, what the question blank that provides to production units 120 recorded for example is the shown allocation result of Fig. 3 (a).And then, can with reference to this question blank, produce the block motion-vector separately of each territory element of picture 200 to be estimated to production units 120.
In sum, distribute at least one block in each territory element by during set motion-vector is worthwhile each, can guarantee in fixing or Variable Area, produce the motion-vector of each different directions, so can the situation that the motion-vector of some direction fails to rotate not occur such as prior art.Thus, compared to prior art, converging to the required time of correct vector in the present embodiment is significantly to shorten.In addition, in each block unit, adjacent block can also arrange different motion-vector values, therefore can further shorten convergence time again.
It should be noted, in the present embodiment, each territory element has and fixedly adds up to 8 number of blocks, but the present invention is not limited to this.In other embodiments, each territory element also can be the number of blocks with variable sum.
In addition, also it is noted that, in the present embodiment, the block of picture 202 to be estimated is to explain with square block, for example is that the ranks number is the square block of fixing 16 row, 16 row, and namely every delegation, each row all have 16 blocks, but the invention is not restricted to this.In other embodiments, and the line number of each block, columns, or shape, area all can change according to design requirement.In addition, each block of the present embodiment for example comprises 4 * 4 pixels, but the invention is not restricted to this.In other embodiment, each block can comprise the pixel quantity of different numbers.
Fig. 4 (a) is to the method for salary distribution of Fig. 4 (c) for the motion-vector value of the picture to be estimated of one embodiment of the invention.Please refer to Fig. 3 (a) to Fig. 4 (c), in the present embodiment, each territory element of picture 200 to be estimated has one or more subregion.Each subregion comprises respectively one or more block.
Particularly, take Fig. 4 (a) to Fig. 4 (c) as example, it shows respectively among Fig. 3 (a) in the picture 200 to be estimated the wherein territory element 400 of two row blocks.From left to right, these territory elements 400 comprise respectively 2 sub regions 401a, 402a; 401b, 402b; 401c, 402c.In the present embodiment, each subregion comprises respectively 4 blocks, but the invention is not restricted to this, and class is pushed into the situation of other block sums easily, even the block sum of different subregions also can be different in each territory element.
Such as Fig. 4 (a) to shown in Fig. 4 (c), subregion 401a, 401b and 401c all comprise identical set motion-vector value " 4 ", " 0 ", " 1 ", " 5 ", yet these set motion-vector values but are routed to different positions in different subregions.In this embodiment, these set motion-vector value positions in different subregions present a kind of rotation map relation.In specific words, subregion 401b is subregion 401a with rotation map clockwise and gets; Subregion 401c is subregion 401b with rotation map clockwise and gets.Similarly, subregion 402a, 402b and 402c comprise identical set motion-vector value " 7 ", " 6 ", " 2 ", " 3 ", and these set motion-vector value positions in different subregions also exist similar rotation map relation.
In one embodiment, when distributing set motion-vector value to each territory element, can provide first a seed region unit to production units 120, again based on the distribution aspect of this seed region unit, and respectively for each territory element a plurality of blocks separately, distribute set motion-vector value.
Furthermore, for reaching the distribution aspect shown in Fig. 3 (a), the territory element 400 that for example can provide first Fig. 4 (a) to production units 120 is as the seed region unit, and is at least one subregion with the seed region dividing elements, is 2 first subregion 401a and 402a for example.Each first subregion comprises respectively 4 blocks, and is assigned with 4 different set motion-vector values.In other embodiment, seed region unit 400 also can have other to distribute aspect, gets final product so long as satisfy set motion-vector each at least one block of all distributing in the seed region unit 400 in worthwhile.In addition, preferably, in seed region unit 400, the set motion-vector value that wantonly two adjacent a plurality of blocks are assigned with is not identical.
Then, again the territory element 400 of Fig. 4 (b) is divided into 2 second subregion 401b and 402b to production units 120, each second subregion also comprises respectively 4 blocks.Wherein, each the second subregion block separately corresponds to each the first subregion block separately.For example, the second subregion 401b comprises upper left block, lower-left block, bottom right block and upper right block, corresponds to respectively upper left block, lower-left block, bottom right block and the upper right block of the first subregion 401a.Similarly, the second subregion 402b also comprises upper left block, lower-left block, bottom right block and upper right block, also corresponds to respectively upper left block, lower-left block, bottom right block and the upper right block of the first subregion 402a.
Afterwards, respectively based on the first subregion 401a and 402a distribution aspect separately, and determine that for the second corresponding subregion 401b and 402b it distributes aspect respectively to production units 120.In the present embodiment, be to concern to distribute the set motion-vector value with identical to distribute to different subregions with a rotation map to production units 120.That is, the distribution aspect of the second subregion 401b be by the distribution aspect of the first subregion 401a with rotation map clockwise and get, and be assigned to identical set motion-vector value " 4 ", " 0 ", " 1 ", " 5 " in diverse location.Similarly, the distribution aspect of the second subregion 402b be by the distribution aspect of the first subregion 402a with rotation map clockwise and get, and be assigned to identical set motion-vector value " 6 ", " 2 ", " 3 ", " 7 " in diverse location.The present invention is not limited to this.Can by any way, respectively based on the first subregion distribution aspect separately, and determine that for the second corresponding in each territory element subregion it distributes aspect respectively to production units 120.
Although Fig. 4 (a) only shows the producing method of the motion-vector value that the first row to the four of picture 202 to be estimated is listed as to Fig. 4 (c), yet the producing method of the subregion of other territory elements also can be analogized and got by Fig. 4 (a) and the shown method of salary distribution of Fig. 4 (b) in the picture 202 to be estimated, and just repeats no more at this.Therefore, by constantly repeating above-mentioned mode, can obtain the allocation result such as the picture to be estimated 202 of Fig. 3 (a) to production units 120.
In brief, in the present embodiment, when distributing set motion-vector value for each territory element block separately to production units 120 respectively based on the seed region unit, with a plurality of the first subregions of seed region dividing elements, and each territory element except the seed region unit is divided into a plurality of the second subregions, and each the second subregion block separately corresponds to the first subregion block separately.Then, respectively based on the first subregion distribution aspect separately, and determine that for the second corresponding in each territory element subregion it distributes aspect respectively to production units 120.All distribute to the condition of at least one block in the seed region unit 400 because the distribution aspect of seed region unit 400 satisfies each in worthwhile of set motion-vector, so all the other can satisfy above-mentioned condition equally by the territory element 400 that aforesaid way produces.As a result, can guarantee in fixing or Variable Area, to produce the motion-vector of all directions, and then converge to the required time of correct vector and can significantly shorten.Although it should be noted that in this method of salary distribution that only comprises two sub regions take each subregion as example, so class is pushed into the method for salary distribution of the subregion of other numbers easily.
Under the above-mentioned method of salary distribution of foundation, the one to multiple block of the corresponding subregion of wantonly two territory elements can concern to be assigned identical set motion-vector value by a rotation map.It should be noted, the existing rotation map relation clockwise of the corresponding subregion of wantonly two territory elements only in order to illustrate, the invention is not restricted to this.For example in other embodiment, can change with anti-clockwise rotation map relation, or any mapping relations (such as mapping etc. is shone upon, reflected in translation) or its combination.Yet, no matter which kind of mapping relations producing allocation result with, each adjacent block is preferably and corresponds to different motion-vector values in the picture to be estimated, to guarantee in the corresponding block to be estimated the motion-vector that adjacent block is corresponding different.
Fig. 5 is the flow chart of steps of the mobile estimating method of first embodiment of the invention.Please refer to Fig. 1 to Fig. 5, the mobile estimating method of the present embodiment comprises the steps.
At first, in step S600, one first picture 202 to be estimated is divided into a plurality of territory elements 400, each territory element comprises a plurality of blocks.Then, in step S602, one group of motion-vector value is distributed to each territory element.Afterwards, in step S604, according to the allocation result of step S602, produce the block motion-vector separately of each territory element of the first picture 202 to be estimated.
It should be noted that in step 602, the motion-vector value comprises a plurality of set motion-vector values, and each set motion-vector value is assigned at least one block in each territory element.In addition, preferably, in each territory element, the set motion-vector value that wantonly two adjacent blocks are assigned with is not identical.All the other can obtain enough teachings, suggestion and implementation in the narration by the exemplary embodiment of Fig. 1～Fig. 5 about the more details of the mobile estimating method of the first embodiment, therefore repeat no more.
The second embodiment
Except the turnover rate (refresh rate) with panel TV is increased to the 120Hz by 60 hertz (Hz), its turnover rate more can increase to 240Hz.At this moment, between two reference pictures, then must insert at least one picture to be estimated, to promote the image quality of video image.
In the first embodiment, the illustrated person of mobile estimating method is for inserting a picture to be estimated between two reference pictures.In the present embodiment, mobile estimating method will be illustrated in the situation of inserting two pictures to be estimated between two reference pictures.
Fig. 6 is the handled video image of mobile estimating device of in the present embodiment Fig. 1.Please refer to Fig. 1 and Fig. 6, in the present embodiment, video image comprises reference picture 701,704 and picture to be estimated 702,703.
At first, the mobile estimating of image estimation unit 110 pictures 702 to be estimated for example is according to reference picture 701,704 image information and the motion-vector that provides to production units 120.For example can adopt the described mode of the first embodiment at this mobile estimating method that adopts.In other words, can produce each territory element distribution aspect separately in the first picture 702 to be estimated according to the mode that produces each territory element distribution aspect separately in the first picture 202 to be estimated among the first embodiment first to production units 120.I.e. one of them territory element 802 of the first picture 702 to be estimated shown in Fig. 7 (a).
Then, image estimation unit 110 is the mobile estimating of picture 703 to be estimated then, and can distribute the motion-vector of aspect according to reference picture 701,704 image information and to the difference that production units 120 provides equally, and mobile estimating method can adopt the described mode of the first embodiment equally.
Preferably, image estimation unit 110 can decide distribution aspect in the second picture 703 to be estimated based on each territory element distribution aspect separately in the first picture 702 to be estimated to production units 120 when the mobile estimating of picture 703 to be estimated.
Specifically, can be based on each territory element distribution aspect separately in the first picture 702 to be estimated to production units 120, and respectively for territory element corresponding in the second picture 703 to be estimated, distribute set motion-vector value.The territory element 803 of the territory element 802 of corresponding the first picture 702 to be estimated in i.e. the second picture 703 to be estimated shown in Fig. 7 (b).
In a preferred embodiment, the block of the territory element 802 of the first picture 702 to be estimated with the block of territory element 803 corresponding in the second picture 703 to be estimated, is that the rotation map with clockwise concerns to be assigned identical set motion-vector value.Similarly, except territory element 802 and 803, the block of other each territory elements of the first picture 702 to be estimated, the block with each territory element of the second picture 703 to be estimated also can concern to be assigned identical set motion-vector value by identical rotation map.
By the above-mentioned method of salary distribution, owing to having the rotation map relation between the distribution aspect of the first picture 702 to be estimated and the distribution aspect of the second picture 703 to be estimated, therefore the territory element of corresponding the first picture 702 to be estimated in the second picture 703 to be estimated, the set motion-vector value that its corresponding block is assigned with is not identical.As a result, can avoid image estimation unit 110 for the block of two pictures 702 to be estimated, 703 corresponding same positions, use identical motion-vector to carry out mobile estimating and affect the situation generation of convergence rate.
It should be noted that the present invention is not limited to concern to distribute with the rotation map of clockwise.In other embodiment, can be by any way to production units 120, for example be various mapping relations (such as translation, reflection) or its combination, and based on each territory element distribution aspect separately in the first picture 702 to be estimated, and respectively for territory element corresponding in the second picture 703 to be estimated, distribute set motion-vector value.On better, as long as satisfy in picture 703 to be estimated, the condition that the block of correspondence picture 702 same positions to be estimated corresponds to different motion-vectors gets final product.
In addition, as previously mentioned, each territory element in the first picture 702 to be estimated, the set motion-vector value that its wantonly two adjacent block is assigned with is preferably not identical.By the above-mentioned method of salary distribution, owing to being based on each territory element distribution aspect separately in the first picture 702 to be estimated, come respectively to distribute set motion-vector value for territory element corresponding in the second picture 703 to be estimated.Therefore allocation result also can be so that each territory element in the second picture 703 to be estimated, and the set motion-vector value that its wantonly two adjacent block is assigned with is not identical.
Fig. 8 is the flow chart of steps of the mobile estimating method of second embodiment of the invention.Please refer to Fig. 1 to Fig. 8, the mobile estimating method of the present embodiment comprises the steps.
At first, in step S900, one first picture 702 to be estimated is divided into a plurality of territory elements 801, each territory element comprises a plurality of blocks.Then, in step S902, one group of motion-vector value is distributed to each territory element.Afterwards, in step S904, according to the allocation result of step S902, produce the block motion-vector separately of each territory element of the first picture 702 to be estimated.
Then, in step S906, based on each territory element distribution aspect separately in the first picture 702 to be estimated, and respectively for territory element corresponding in the second picture 703 to be estimated, distribute set motion-vector value.Then, in step S908, according to the allocation result of step S906, produce the block motion-vector separately of each territory element of the second picture 703 to be estimated.It should be noted that in the present embodiment, the block of each territory element of the first picture 702 to be estimated, the block with each territory element of the second picture 703 to be estimated can concern to be assigned identical set motion-vector value by rotation map.
In addition, the mobile estimating method of the second embodiment of the present invention can be obtained enough teachings, suggestion and implementation in the narration by the exemplary embodiment of Fig. 1～Fig. 7, therefore repeats no more.
In sum, in the various embodiments described above of the present invention, distribute at least one block in each territory element by during set motion-vector is worthwhile each, can guarantee in fixing or Variable Area, produce the motion-vector of all directions, and can the situation that the motion-vector of some direction fails to rotate not occur such as prior art.Thus, compared to prior art, it is significantly to shorten that the mobile estimating method of the various embodiments described above converges to the required time of correct vector.In addition, block adjacent in the picture to be estimated more can arrange different motion-vector values, therefore when carrying out mobile estimating, can further shorten mobile object again and converge to the required time of correct vector.
Although the present invention discloses as above with embodiment, so it is not to limit the present invention, any person of an ordinary skill in the technical field, but do a little change and retouching, and do not break away from the spirit and scope of the present invention.
Claims (11)
1. mobile estimating method, this mobile estimating method comprises:
One first picture to be estimated is divided into a plurality of territory elements, and each this territory element comprises a plurality of blocks; And
One group of motion-vector value is distributed to each this territory element, wherein
This group motion-vector value comprises a plurality of set motion-vector values, and
Each was assigned at least one block during these a plurality of set motion-vectors were worthwhile in each this territory element.
2. mobile estimating method according to claim 1, wherein in each this territory element, the set motion-vector value that wantonly two these adjacent a plurality of blocks are assigned with is not identical.
3. mobile estimating method according to claim 1, wherein each this territory element has variable or the fixing number of blocks of sum.
4. mobile estimating method according to claim 1 wherein also comprises providing a question blank, is used for the set motion-vector value that these a plurality of blocks of each territory element of this first picture to be estimated of record are assigned to separately.
5. mobile estimating method according to claim 4 wherein also comprises with reference to this question blank, produces these a plurality of blocks motion-vector separately of each this territory element of this first picture to be estimated.
6. mobile estimating method according to claim 1, wherein each this territory element of this first picture to be estimated has one or more subregion, each this subregion comprises respectively one or more block, and this one to multiple block of the corresponding subregion of wantonly two territory elements, be to concern to be assigned identical set motion-vector value with a rotation map.
7. mobile estimating method according to claim 1, wherein will organize the motion-vector value and distribute to the step of each this territory element and comprise:
One seed region unit is provided, and it comprises a plurality of blocks, is assigned respectively this a plurality of set motion-vector values; And
Based on the distribution aspect of this seed region unit, and respectively for each this territory element these a plurality of blocks separately, distribute this a plurality of set motion-vector values.
8. mobile estimating method according to claim 7, distribute the step of these a plurality of set motion-vector values to comprise for each this territory element these a plurality of blocks separately respectively based on this seed region unit:
Be one or more first subregion with this seed region dividing elements, each this first subregion comprises one or more block;
Each this territory element is divided into one or more the second subregion, each this second subregion comprises one or more block, and wherein each this second subregion this one or more block separately corresponds to this one or more the first subregion this one or more block separately; And
Respectively based on this one or more the first subregion distribution aspect separately, and determine that for this one or more the second subregion corresponding in each this territory element it distributes aspect respectively.
9. mobile estimating method according to claim 8, this one or more the first subregion this one or more block separately wherein, with this one or more the second subregion this one or more block separately in each this territory element, be to concern to be assigned identical set motion-vector value with a rotation map.
10. mobile estimating method according to claim 1 wherein also comprises:
Based on each this territory element distribution aspect separately in this first picture to be estimated, and respectively for this corresponding in one second picture to be estimated territory element, distribute this a plurality of set motion-vector values.
11. mobile estimating method according to claim 10, these a plurality of blocks of each this territory element of this first picture to be estimated wherein, with these a plurality of blocks of each this territory element of this second picture to be estimated, be to concern to be assigned identical set motion-vector value with a rotation map.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2011102153991A CN102905124A (en) | 2011-07-29 | 2011-07-29 | Motion estimation method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2011102153991A CN102905124A (en) | 2011-07-29 | 2011-07-29 | Motion estimation method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN102905124A true CN102905124A (en) | 2013-01-30 |
Family
ID=47577141
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2011102153991A Pending CN102905124A (en) | 2011-07-29 | 2011-07-29 | Motion estimation method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102905124A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111836055A (en) * | 2020-07-17 | 2020-10-27 | 上海顺久电子科技有限公司 | Image processing device and image block matching method based on image content for MEMC |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1925614A (en) * | 2005-08-30 | 2007-03-07 | 深圳清华大学研究院 | Rapid movement estimating method |
| CN1992789A (en) * | 2005-12-28 | 2007-07-04 | 三星电子株式会社 | Motion estimator and motion method |
| CN101288310A (en) * | 2005-06-23 | 2008-10-15 | Nxp股份有限公司 | Motion estimation |
| CN101610410A (en) * | 2008-06-19 | 2009-12-23 | 联发科技股份有限公司 | The method for processing video frequency and the device thereof of the target travel vector of decision target image block |
-
2011
- 2011-07-29 CN CN2011102153991A patent/CN102905124A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101288310A (en) * | 2005-06-23 | 2008-10-15 | Nxp股份有限公司 | Motion estimation |
| CN1925614A (en) * | 2005-08-30 | 2007-03-07 | 深圳清华大学研究院 | Rapid movement estimating method |
| CN1992789A (en) * | 2005-12-28 | 2007-07-04 | 三星电子株式会社 | Motion estimator and motion method |
| CN101610410A (en) * | 2008-06-19 | 2009-12-23 | 联发科技股份有限公司 | The method for processing video frequency and the device thereof of the target travel vector of decision target image block |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111836055A (en) * | 2020-07-17 | 2020-10-27 | 上海顺久电子科技有限公司 | Image processing device and image block matching method based on image content for MEMC |
| CN111836055B (en) * | 2020-07-17 | 2023-01-10 | 上海顺久电子科技有限公司 | Image processing device and image block matching method based on image content for MEMC |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105049831B (en) | Use the super-resolution display of cascade panel | |
| CN104471941B (en) | The method and apparatus of son segmentation prediction between view in 3D Video codings | |
| CN107240384B (en) | Display brightness compensation method and device | |
| CN108781288A (en) | Independent multiresolution decoding | |
| CN102144394B (en) | Stereoscopic video delivery | |
| CN108063976A (en) | A kind of method for processing video frequency and device | |
| CN105629620B (en) | Refractor and its driving method, display device | |
| KR100815486B1 (en) | A method and apparatus for encoding multiview video and a storage medium using the same | |
| WO2010037120A1 (en) | Demultiplexing for stereoplexed film and video applications | |
| JP2002523932A (en) | Automatic 3D display method | |
| JP2014531807A (en) | Image processing system and method | |
| US20170201757A1 (en) | Video data encoding and decoding | |
| WO2016119396A1 (en) | Pixel structure, array substrate and control method thereof, and display device | |
| JP2012075030A (en) | Video distribution server and video reproduction terminal | |
| KR101625910B1 (en) | Method and device for image processing by image division | |
| CN102137268A (en) | Line-staggered and tessellated rendering method and device for three-dimensional video | |
| CN102792703A (en) | Video processing device and video processing method | |
| CN102905124A (en) | Motion estimation method | |
| US20070008562A1 (en) | Image processing apparatus, image processing method and program | |
| EP3270234B1 (en) | Method, computer program and apparatus for processing three-dimensional image | |
| CN101626517B (en) | Method for synthesizing stereo image from parallax image in a real-time manner | |
| CN103188554A (en) | Image coding method and image coding system | |
| US20100322535A1 (en) | Image transformation method adapted to computer program product and image display device | |
| CN104796706A (en) | Video coding method and device | |
| CN105227866B (en) | A kind of multi-channel video display methods and device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C02 | Deemed withdrawal of patent application after publication (patent law 2001) | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20130130 |