CN201374785Y - Image processing device for deciding motion vector of interpolated picture - Google Patents
Image processing device for deciding motion vector of interpolated picture Download PDFInfo
- Publication number
- CN201374785Y CN201374785Y CN 200920005185 CN200920005185U CN201374785Y CN 201374785 Y CN201374785 Y CN 201374785Y CN 200920005185 CN200920005185 CN 200920005185 CN 200920005185 U CN200920005185 U CN 200920005185U CN 201374785 Y CN201374785 Y CN 201374785Y
- Authority
- CN
- China
- Prior art keywords
- image
- vector
- block
- interpolation
- 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.)
- Expired - Lifetime
Links
Images
Landscapes
- Television Systems (AREA)
Abstract
The utility model discloses an image processing device for deciding the motion vector of an interpolated picture. The device is favorable for improving the image quality of performing picture renewal frequency conversion operation. The image processing device comprises a storage unit, a calculation unit, and a motion vector deciding module, wherein the storage unit is used for at least storing a former primary picture and a latter primary picture of the interpolated picture, and a start position of a former interpolated picture;; the calculation unit is coupled with the storage unit, and comprises a candidate vector generation module which is used for deciding the first candidate vectors and the second candidate vectors of the blocks of the interpolated pictures; and the motion vector deciding module decides the motion vector of a block in the image covered area or the image revealing area of the interpolated picture according to the first candidate vectors and the second candidate vectors of the blocks of the interpolated picture.
Description
Technical field
The utility model relates to a kind of image processing mechanism, refers to a kind of image processor in order to the motion-vector that determines an interpolation block especially.
Background technology
During the motion-vector of present traditional image interpolation mechanism interpolation block in determining an interpolation image, system directly decides its motion-vector with the operation result of block alignment algorithm (block matching algorithm), and produces the image of this interpolation block in view of the above.Please refer to Fig. 1, it is the operation chart of block alignment algorithm.As shown in Figure 1, image frame F2, F3 are forward and backward two picture in the input image, and picture Finter is the interpolation image that traditional image interpolation mechanism is produced, A ' ~ L ' is the background video of representing among image frame F2, the F3, the direction that dotted arrow system expression background video among the figure moves, solid arrow is then represented the direction that a foreground object moves, the background video of F ' in image frame F2 ~ I ' is covered in by this foreground object, then is that the background video of C ' ~ F ' is covered in by this foreground object in next image frame F3; Because the block alignment algorithm can all can find figure viewed from behind image A ', B ', J ', K ', L ' in forward and backward two image frame F2, F3, so, can determine the correct target motion-vector of corresponding block, and background video A ', B ', J ', K ', L ' are presented in picture Finter last (as shown in Figure 1); In addition, the block alignment algorithm also can find the image of foreground object in forward and backward two image frame F2, F3, therefore, can determine the correct target motion-vector of corresponding block, and the image of foreground object is presented on the picture Finter.
Yet, during the target motion-vector of the interpolation block in decision regional Rinter and Rinter ', the target motion-vector of the interpolation block in region R inter and the Rinter ' should be the background motion-vector so that can present background video in region R inter and the Rinter ' ideally, for example, region R inter should present background video C ' ideally, D ', and region R inter ' should present background video H ' ideally, I ', yet, in fact, because the block alignment algorithm can not find background video C ' in image frame F3, D ' (being covered by foreground object), and in image frame F2, can not find background video H ', I ' (originally being covered) by foreground object, so, the block alignment algorithm also can't determine correct motion-vector in the mode of general image comparison, cause interpolation region R inter, the actual image distortion that is presented of Rinter ', when being applied to frame updating frequency (frame rate) conversion, traditional image interpolation mechanism will significantly reduce the quality of image output.
The utility model content
Technical problem to be solved in the utility model provides a kind of image processor of motion-vector in order to the decision interpolation image, helps to promote the quality of image of carrying out the running of frame updating frequency inverted.
In order to solve above technical problem, the utility model provides following technical scheme:
The utility model provides a kind of image processor.This image processor determines when being used to carry out the motion picture interpolation that the image area of coverage/image of an interpolation image appears the motion-vector in district, wherein this interpolation image is made up of plurality of blocks, and this image processor includes a storage element and a computing unit, and this computing unit more comprises a candidate vector generation module and a motion-vector decision module.This storage element system is used for storing at least a last raw frames of this interpolation image and an initial position of back one raw frames and a last interpolation image; This candidate vector generation module system is used to determine plural number first candidate vector and plural second candidate vector of the plurality of blocks of this interpolation image; This motion-vector decision module system is according to those first candidate vector of those blocks of this interpolation image, those second candidate vector this original position with this last interpolation image, and the image area of coverage/image that determines to be positioned at this interpolation image appears the motion-vector of a block in district.
The utility model adopt in order to the image treatment method of the motion-vector of decision interpolation image and relevant apparatus thereof at belonging to the block that the image area of coverage or image appear the district in the interpolation image, can correctly calculate the target motion-vector of this block, make to obtain preferable image when carrying out the image interpolation, help to promote the quality of image of carrying out the running of frame updating frequency inverted with this target motion-vector.
Description of drawings
Fig. 1 is the operation chart of traditional block alignment algorithm.
Fig. 2 is the assembly schematic diagram of first embodiment of image processor of the present utility model.
Fig. 3 (a) is the operational flowchart of image processor shown in Figure 2.
Fig. 3 (b) is the operational flowchart of second embodiment of image processor shown in Figure 2.
The schematic diagram of the image output that Fig. 4 is produced when carrying out the frame updating frequency inverted for computing unit shown in Figure 2.
Fig. 5 (a) is the example schematic that computing unit shown in Figure 2 calculates the numerical value of representing the motion-vector degree of variation.
Fig. 5 (b) is the example schematic that includes the first numerical curve CV of a plurality of motion-vector variation values.
Fig. 6 (a) carries out the example schematic that block that the block comparison calculation calculated is respectively compared difference curve DCV for computing unit shown in Figure 2 uses candidate vector MV1.
Fig. 6 (b) carries out the example schematic that block that the block comparison calculation calculated is respectively compared difference curve DCV ' for computing unit shown in Figure 2 uses candidate vector MV2.
Fig. 7 is the original position P1 that computing unit determined ~ P4 shown in Figure 2 and the example schematic of P1 ' ~ P4 '.
[primary clustering symbol description]
200 image processors, 205 computing units
210 storage elements, 215 candidate vector generation modules
220 motion-vector decision module, 225 image generation modules
2151 image difference value generation modules, 2152 candidate vector decision module
2201 original position decision module, 2202 vectors are selected module
Embodiment
At first, be easy-to-read, below be will do not covered in last the image frame by foreground object but in next image frame by background video zone that foreground object covered, be called the image area of coverage (covered area), and will be covered by foreground object in last the image frame but the background video zone of (not covered by foreground object) in next image frame, occurs, be called image and appear district (uncovered area); For instance, background video C ', D ', E ' shown in Figure 1 is the image area of coverage, and background video G ', H ', I ' are that image appears the district; Note that above definition only in order to the operation of convenient explanation embodiment of the present utility model, is not restriction of the present utility model.
Please arrange in pairs or groups with reference to Fig. 2 and Fig. 3 (a), it illustrates the image processor 200 of the utility model first embodiment and the flow process of associative operation step thereof respectively; For convenience of description, the step that below is collocation Fig. 3 (a) illustrates the operation of the image processor 200 of Fig. 2, be noted that, if can reach identical result substantially, the sequence of steps that does not need necessarily to shine in the flow process shown in Fig. 3 (a) is carried out, and the step shown in Fig. 3 (a) not necessarily will be carried out continuously, that is other step also can be inserted wherein.As shown in Figure 2, image processor 200 includes a computing unit 205 and a storage element 210, image processor 200 determines the area of coverage/image of image at least of an interpolation image to appear the motion-vector in district when being used to carry out the motion picture interpolation, and this interpolation image system is made up of plurality of blocks; Wherein storage element 210 is to be used for storing at least the initial position (starting position/point) that last raw frames of an interpolation image and raw frames thereafter and last the interpolation image image area of coverage/image appear the district, 205 of computing units are coupled to storage element 210, and comprise a candidate vector generation module 215, an one motion-vector decision module 220 and an image generation module 225, candidate vector generation module 215 is used to determine plural number first candidate vector (also be called and move left vector) and plural second candidate vector (also being called the vector that moves right) of the plurality of blocks of this interpolation image, and motion-vector decision module 220 is to move left vector according to those of those blocks of this interpolation image, this original position vectorial and this last interpolation image moves right, decide the image area of coverage/image that is positioned at this interpolation image to appear the corresponding motion-vector of each block in district, and image generation module 225 is the image that the image area of coverage/image that the motion-vector that determined according to motion-vector decision module 220 produces this interpolation image appears each block in the district.
Specifically, candidate vector generation module 215 more comprises an image difference value generation module 2151 and a candidate vector decision module 2152, image difference value generation module 2151 is an image difference value (step 305) that decides this block according to the plural original motion-vector of a plurality of adjacent block of a block, 2152 of candidate vector decision module according to one-dimensional space plural number continuously the plural image difference values of block determine the continuous block of this plural number respectively those move left vector and those vectors (step 310) that moves right.In addition, 220 of motion-vector decision module comprise an initial determining positions module 2201 and a vector is selected module 2202, original position decision module 2201 is to utilize those of this plurality of blocks of this interpolation image to move left one of them of vector and those vectors that move right, calculate the pairing plurality of blocks comparison of this plurality of blocks difference, and decide this original position (step 315) of this interpolation image according to the numerical value change of those plurality of blocks comparison differences, and vector selection module 2202 is to produce a reference vector according to this original position of this interpolation image and this original position of this last interpolation image, and relatively this reference vector and the image area of coverage/image of this interpolation image appear this corresponding left side of each block in the district, the vector that moves right is chosen this corresponding left side, one of them more not close with this reference vector of vector that move right is used as the target motion-vector (step 320) of this each block.
Please note, original motion-vector system's last the raw frames of foundation of first embodiment and a back raw frames are carried out the block comparison and are obtained, and the step (that is step 315) that determines this original position of this interpolation image can utilize earlier those move left vector and those move right vector one of them, calculate the pairing plurality of blocks comparison of this plurality of blocks difference, then decide this original position according to the numerical value change of those blocks comparison differences of being calculated again, detail operations will be in follow-up explanation.It should be noted that, image processor 200 does not determine the image area of coverage/image of last interpolation image to appear the original position in district in first embodiment, in other words, it is under the situation of the hypothesis original position that appears the district at the image area of coverage/image of learning last interpolation image, only calculate at present that the image area of coverage/image appears the original position in district in this interpolation image, and do not need the image area of coverage/image of the last interpolation image of extra computation to appear the original position in district.
In addition, in second embodiment of the present utility model, collocation is with reference to the flow chart of Fig. 3 (b), the candidate vector generation module 215 of Fig. 2 also can be designed to determine the candidate vector of at least one first interpolation block of one first interpolation image (to be called a left side, vector moves right) MV1, MV2 (step 405), and the candidate vector MV1 ' that determines at least one second interpolation block of one second interpolation image, MV2 ' (step 410), 215 pairs first of candidate vector generation modules in the present embodiment on real the work, each interpolation block in second interpolation image all produces corresponding two candidate vector, and the size of each block for example includes 8 * 8 pixel coverage; In addition, motion-vector decision module 220 is that a candidate vector (for example MV1) that is designed to use at least respectively the first interpolation block calculates first with a candidate vector (for example MV1 ') of the second interpolation block, the pairing plurality of blocks comparison of a plurality of interpolation blocks of second interpolation image difference (step 415 and step 420), and these blocks comparison differences are stored in the storage element 210, then decide in first interpolation image at least one original position (step 425) at least one original position and second interpolation image according to the numerical value change of aforementioned these blocks comparison differences again, wherein this two original position is the marginal position that an image area of coverage or an image appear the district, motion-vector decision module 220 can decide the target motion-vector (step 430) of the second interpolation block according to this original position in this original position in first interpolation image and second interpolation image, wherein first, at least the picture that exists a non-image interpolation to be produced between second interpolation image, that is, first interpolation image can be the picture that last image interpolation of second interpolation image produced, for instance, please refer to Fig. 4, its illustrate is (picture frame or figure of 205 pairs of a plurality of input image of computing unit shown in Figure 2, only illustrate Fn-1 at this, Fn, Fn+1) carry out frame updating frequency inverted (frame rate conversion) to produce a plurality of image outputs, for example, picture frame Fn-1, Fn, Fn+1 etc. are the input picture frame of 60Hz, and computing unit 205 is that these input picture frames are carried out the frame updating frequency inverted has 120Hz with generation picture frame; The non-restriction of the present utility model of frequency 60Hz and 120Hz.Though following operation is that example describes with second embodiment, so its operation also is applicable to first embodiment, for omitting length, does not give unnecessary details in addition.
When in fact handling, image processor 200 lies in and produces an interpolation image (the first interpolation image F ') among picture Fn-1, the Fn, and in picture Fn, Fn+1, produce another interpolation image (the second interpolation image F "), the second interpolation image F " promptly be to desire the picture that interpolation goes out at present and the first interpolation image F ' is last the picture that the image interpolation is produced; In another embodiment, first interpolation image can also be the second interpolation image F " a back picture that the image interpolation is produced, this is without prejudice to technical spirit of the present utility model.Image frame Fn-1, Fn, Fn+1 with Fig. 4 are example, and the moving direction system of foreground object moves horizontally from right to left, and background video system moves horizontally from left to right; Being noted that, is only with the example that moves horizontally as an illustration at this for convenience of description, yet the image of arbitrary specific direction (vertical direction or picture to angular direction etc.) moves and all can use embodiment of the present utility model to handle.
Below narration system describes at the actual operation that each interpolation block of each interpolation image (comprise F ', F " etc.) produces two candidate vector candidate vector generation module 215 earlier; In this, picture F ' means first interpolation image, and picture F " mean second interpolation image.At first, image difference value generation module 2151 can calculate the motion-vector of all interpolation blocks according to block alignment algorithm (block matching algorithm), again with a plurality of motion-vectors of the motion-vector adjacent block a plurality of that each was calculated with it, calculate the confusion degree of motion-vector, to calculate a numerical curve, wherein, with present embodiment, this numerical curve is the formed curve of section along the motion-vector confusion degree of a certain interpolation block horizontal direction, this numerical curve is the difference value of a plurality of continuous blocks of the one-dimensional space, and this confusion degree is represented the variation value (motion vector variance) of motion-vector, that is this numerical curve includes a plurality of numerical value of the different motion-vector degrees of variation of representative on the specific direction.Please refer to Fig. 5 (a), Fig. 5 (a) is the example schematic that the utility model calculates the numerical value of representing the motion-vector degree of variation; For example, if first interpolation block adjacent block a plurality of with it are (shown in Fig. 5 (a), being positioned at 5 * 5 block scope) a plurality of motion-vectors of being calculated according to the block alignment algorithm are respectively MV00 and MV-2-2 ~ MV22, then can calculate a motion-vector variation value MV_VAR according to these motion-vectors, its algorithm is to get that the maximum horizontal component deducts the wherein absolute value of minimum level component gained in these motion-vectors, add that maximum vertical component deducts the wherein absolute value of minimum vertical component gained in these motion-vectors, MV_VAR can utilize following equation to represent it:
MV_VAR=|MAX (MV
x)-MIN (MV
x) |+| MAX (MV
y)-MIN (MV
y) | equation (1)
Wherein MVx and MVy represent horizontal component (x axle component) and vertical component (y axle component) respectively.Be noted that 5 * 5 block scope is not to be restriction of the present utility model, it also can utilize the block scope of N * N or N * M to do it in fact, and wherein parameter N and M all are that positive integer and N are not equal to M; In addition, the mode of calculating motion-vector variation value MV_VAR also can be used equation (2) instead or equation (3) is realized:
MV_VAR＝|MAX(MV
x)-MIN(MV
x)|+|MAX(MV
y)-MIN(MV
y)|+SAD
Equation (2)
MV_VAR＝α×{|MAX(MV
x)-MIN(MV
x)|+|MAX(MV
y)-MIN(MV
y)|}+β×SAD
Equation (3)
Wherein numerical value SAD is the block comparison difference that the first interpolation block is calculated according to block comparison method, and parameter alpha, β are weighting parameters; All can variation in order to arbitrary enforcement of calculating the numerical value of representing the motion-vector degree of variation all belongs to category of the present utility model.From the above mentioned, according to equation (1), equation (2) or equation (3) one of them, computing unit 205 calculates at different interpolation blocks one by one, so can draw the first numerical curve CV, shown in Fig. 5 (b).
Fig. 5 (b) illustrate is for being the first numerical curve CV that the utility model includes a plurality of motion-vector variation values.MB00 is the first interpolation block, and candidate vector decision module 2152 is the two candidate vector MV1 that determine the first interpolation block MB00 in Fig. 5 (b), MV2, at first, the direction (for example horizontal direction) that moves with background or prospect image, candidate vector decision module 2152 is spatially being prolonged the both sides of the first interpolation block MB00, and a plurality of (for example both sides are each six, the corresponding motion-vector variation value of block of MB10 ~ MB60 and MB-10 ~ MB-60), in these corresponding motion-vector variation value scopes, take out a maximum (for example VARmax shown in Fig. 5 (b)), and in these corresponding motion-vector variation value scopes, find out this maximum VARmax left side, the block that minimum corresponded to of right both sides, for example, can find block MB-40 and MB50, and this left side, right two block MB-40 and MB50 utilize motion-vector that the block alignment algorithm calculated promptly as the two candidate vector MV1 of the first interpolation block MB00 respectively, MV2, in other words, the candidate vector MV1 system of the first interpolation block MB00 corresponds to the minimum VARmin in the left side of the maximum VARmax that is positioned at the first numerical curve CV, also claim that candidate vector MV1 is the vector that moves left of the first interpolation block MB00, and its candidate vector MV2 system corresponds to the minimum VARmin ' on the right side of the maximum VARmax that is positioned at the first numerical curve CV, also claim that candidate vector MV2 is the vector that moves right of the first interpolation block MB00; And one of them of candidate vector MV1, MV2 is to correspond to background motion-vector (background motionvector), its another then corresponding to prospect motion-vector (foreground motion vector), this is because the motion-vector variation value that belongs to around the interpolation block that the image area of coverage or image appear the district will be quite big, and the minimum motion-vector variation of and arranged on left and right sides is worth pairing image block, it can be corresponding to a prospect motion-vector or a background motion-vector, looks it and is positioned at the image area of coverage or image appears Qu Erding.Therefore, if first interpolation block MB00 system is positioned at the image area of coverage and image appear the district one of them, then one of them of its candidate vector MV-1, MV2 be corresponding to the background motion-vector, and its another corresponding to the prospect motion-vector; Be noted that, two candidate vector among the embodiment of the present utility model (or claiming left and right motion-vector) are in fact respectively a prospect vector and a background vector, two candidate vector of one block may not be to be prospect, background vector just during actual operation, and right image processor 200 of the present utility model is also applicable in this situation.By above-mentioned operation, candidate vector decision module 2152 can calculate the corresponding two candidate vector MV1 and the MV2 of each interpolation block among the first interpolation image F '.
Please refer to Fig. 6 (a), the plurality of blocks corresponding to a plurality of interpolation blocks of the first interpolation image F ' that its illustrate uses for original position decision module of the present utility model 2201 that the left candidate vector MV1 of each interpolation block of the first interpolation image F ' carries out that the block comparison calculation calculated is respectively compared difference (representing with the curve D CV shown in Fig. 6 (a)); Numerical value change according to these block comparison differences DCV, original position decision module 2201 can determine a plurality of original positions of the first interpolation image F ', and these original position systems are positioned at the image area of coverage edge of (or image appears the district), therefore can be used for judging the image area of coverage actual imagery zone of (or image appears the district).On real the work, at curve D CV (including the plurality of blocks comparison difference of using candidate vector MV1 to be calculated), original position decision module 2201 is the size of its numerical value change of detecting, when from left to right a numerical value change among the curve D CV (including the plurality of blocks comparison difference of using candidate vector MV1 to be calculated) by little when changing greatly and comprise the block comparison difference that reaches a predetermined value TH, original position decision module 2201 can be used as an initial position in this first interpolation image with the pairing block of this block comparison difference position, also is called the image area of coverage starting point of (or image appears the district); For example, with Fig. 6 (a), original position decision module 2201 is that block is compared the pairing original position P2 decision of the pairing original position P1 of difference D1 and block comparison difference D2 for can judge the image area of coverage marginal position of (or image appears the district), this is because with image frame Fn-1, horizontal direction among the Fn, foreground object system moves from right to left, and background video system moves from left to right, use the candidate vector MV1 (in fact be corresponding to the background motion-vector this moment) of the block in the image capturing range R1 to calculate its block that corresponds to comparison difference, the first interpolation image F ' gets before in system, two the image frame Fn-1 in back, image capturing range R1 ' and the R1 of Fn " in image calculate; and because image capturing range R1 ' and R1 " image be identical/similar background video, so the block that calculates that image capturing range R1 is corresponded to comparison difference is all quite little; When the actual system of the block in the image capturing range that is calculated is positioned at the image area of coverage (for example block of image capturing range R2), the block comparison difference of then using candidate vector MV1 (this moment is in fact still corresponding to the background motion-vector) to be calculated can begin to become big, this is because at image capturing range R2 ' and the R2 of forward and backward two image frame F1, F2 " in can not find same or analogous image; be background video in the image capturing range R2 ' wherein, image capturing range R2 " in be the prospect image; With the pairing block comparison difference that calculates in the initial position between image capturing range R1 and the R2, its numerical value change will be from left to right, by little to big increasing suddenly and greater than predetermined value TH, therefore the pairing original position P1 of block comparison difference D1 promptly is the position, an edge of the image area of coverage (or image appears the district), please note, only know that original position P1 is the marginal position that the image area of coverage or image appear the district this moment, need unit to be calculated 205 follow-up carry out other computing beginning can learn original position P1 actual be the marginal position of the image area of coverage.
In addition, use the candidate vector MV1 (in fact be corresponding to the prospect motion-vector this moment) of the block in the image capturing range R3 to calculate its block that corresponds to comparison difference, can get image capturing range R3 ' and the R3 of image frame Fn-1, Fn " in image calculate; and because image capturing range R3 ' and R3 " image be identical/similar prospect image, so the block that calculates that image capturing range R3 is corresponded to comparison difference is all quite little; When being positioned at image, the actual system of the block in the image capturing range that is calculated appears when district (for example block of image capturing range R4), the block comparison difference of then using candidate vector MV1 (this moment is in fact still corresponding to the prospect motion-vector) to be calculated can begin to become big, this is because at image capturing range R4 ' and the R4 of image frame F1, F2 " in can not find same or analogous image; wherein be the prospect image in the image capturing range R4 ', image capturing range R4 " in be background video; With the pairing block comparison difference that calculates in the initial position between image capturing range R3 and the R4, its numerical value change will be from left to right, by little to big increasing suddenly and greater than predetermined value TH, therefore the pairing original position P2 of block comparison difference D2 promptly is the position, an edge of the image area of coverage (or image appears the district), please note, only know that original position P2 is the marginal position that the image area of coverage or image appear the district this moment, need unit to be calculated 205 follow-up carry out other computing beginning can learn original position P2 actual be the marginal position that image appears the district.
In like manner, shown in Fig. 6 (b), original position decision module 2201 can use the right candidate vector MV2 of each interpolation block of the first interpolation image F ' to carry out the block comparison calculation, with the plurality of blocks comparison difference (with the expression of the curve D CV ' shown in Fig. 6 (b)) corresponding to a plurality of interpolation blocks of the first interpolation image F ' that calculates respectively; Numerical value change according to these block comparison differences DCV ', original position decision module 2201 can determine two original position P3, P4 among the first interpolation image F ', and these original position systems are positioned at the image area of coverage edge of (or image appears the district), thereby can be used for judging the image area of coverage actual imagery zone of (or image appears the district).On real the work, at curve D CV ', original position decision module 2201 is the size of its numerical value change of detecting, when from left to right a numerical value change among the curve D CV ' (comprising the plurality of blocks comparison difference of using candidate vector MV2 to be calculated) changes from large to small and comprises the block comparison difference that reaches a predetermined value TH ', original position decision module 2201 can be used as an initial position in this first interpolation image with the pairing block of this block comparison difference position, also is called the image area of coverage starting point of (or image appears the district); For example, with Fig. 6 (b), original position decision module 2201 can be compared block the pairing original position P4 decision of pairing original position P3 of difference D3 and block comparison difference D4 for can judge the image area of coverage marginal position of (or image appears the district), this is because work as the actual system of block in the image capturing range that calculated when being positioned at the image area of coverage (for example block of image capturing range R5), use the candidate vector MV2 (in fact be corresponding to the prospect motion-vector this moment) of the block in the image capturing range R5 to calculate its block that corresponds to comparison difference, image frame Fn-1 gets in system, image capturing range R5 ' and the R5 of Fn " in image calculate; and because image capturing range R5 ' and R5 " image be different background videos, so the block that calculates that image capturing range R5 is corresponded to comparison difference is all bigger; And work as block in the image capturing range that calculated when being positioned at the foreground object imagery zone (for example block of image capturing range R6), the block comparison difference of then using candidate vector MV2 (this moment in fact still corresponding to the prospect motion-vector) to be calculated can diminish, and this is because at the image capturing range R6 ' and R6 of two image frame F1, F2 " in image be identical foreground object image; Therefore, for the pairing block comparison difference D3 that calculates of the initial position P3 between image capturing range R5 and the R6, its numerical value change will be from left to right, suddenly reduce from large to small and less than predetermined value TH ', therefore the pairing original position P3 of block comparison difference D3 promptly is the position, an edge of the image area of coverage (or image appears the district), please note, only know that original position P3 is the marginal position that the image area of coverage or image appear the district this moment, need unit to be calculated 205 follow-up carry out other computing beginning can learn original position P3 actual be the marginal position of the image area of coverage.
In addition, when being positioned at image, the actual system of the block in the image capturing range that is calculated appears when district (for example block of image capturing range R7), use the candidate vector MV2 (in fact be corresponding to the background motion-vector this moment) of the block in the image capturing range R7 to calculate its block that corresponds to comparison difference, can get image frame Fn-1, image capturing range R7 ' and the R7 of Fn " in image calculate; and because image capturing range R7 ' and R7 " image be different images, it wherein in the image capturing range R7 ' the foreground object image, and image capturing range R7 " in be background video, so the block that calculates that image capturing range R7 is corresponded to comparison difference is all quite big; Be positioned at background video when regional (for example block of image capturing range R8) and work as the actual system of block in the image capturing range that calculated, the block comparison difference of then using candidate vector MV2 (this moment in fact still corresponding to the background motion-vector) to be calculated can begin to diminish, and this is because image capturing range R8 ' and R8 " in be identical/similar background video; Therefore, for the pairing block comparison difference D4 that calculates of the initial position P4 between image capturing range R7 and the R8, its numerical value change will be from left to right, suddenly reduce from large to small and less than predetermined value TH ', therefore the pairing original position P4 of block comparison difference D4 promptly is the position, an edge of the image area of coverage (or image appears the district), please note, only know that original position P4 is the marginal position that the image area of coverage or image appear the district this moment, need unit to be calculated 205 follow-up carry out other computing beginning can learn original position P4 actual be the marginal position that image appears the district.
As mentioned above, original position decision module 2201 is to use its left and right two candidate vector MV1, MV2 to carry out the block comparison calculation to each interpolation block in first interpolation image to obtain these block comparison differences (for example curve D CV ' shown in the curve D CV shown in Fig. 6 (a) and Fig. 6 (b)); Original position decision module 2201 also can be used its left and right two candidate vector MV1 ', MV2 ' to carry out the block comparison calculation to each the interpolation block in second interpolation image and obtain other plurality of blocks comparison difference, and utilizes these other block comparison differences to obtain the image area of coverage edge original position of (or image appears the district) (or claiming starting point) P1 ' ~ P4 ' equally; Because its calculating process is similar to aforesaid calculating process, therefore do not give unnecessary details in addition at this.
After original position decision module 2201 calculates original position P1 ~ P4 and P1 ' ~ P4 ', vector selects module 2202 can learn the in esse zone of an image area of coverage (or an image appears the district) in first interpolation image and second interpolation image according to original position P1 ~ P4 and P1 ' ~ P4 ', for example, the imagery zone that vector selects module 2202 to turn right according to original position P1 (or P1 ') and find toward the left side according to original position P3 (or P3 '), promptly be that the image area of coverage or image in first interpolation image (or second interpolation image) appears the district, and the imagery zone of turning right and finding toward the left side according to original position P4 (or P4 ') according to original position P2 (or P2 '), promptly be that the image area of coverage or image in first interpolation image (or second interpolation image) appears the district, it should be noted that, the imagery zone that this place is found is not confirmed to be an image area of coverage as yet or an image appears the district, only know that it is that the image area of coverage and image appear one of them of district, need other candidate vector MV1 ' according to the block in the imagery zone that is found, the vector difference of MV2 ' judges that further this imagery zone is the image area of coverage or is that image appears the district that its detailed operation is in describing after a while; In addition, heal when big when the vector difference of candidate vector MV1 ', MV2 ', the actual imagery zone of the image area of coverage (or image appears the district) also can be healed greatly, and vice versa.
From the above, original position decision module 2201 is after calculating original position P1 ~ P4 and P1 ' ~ P4 ', the module 2202 of selecting vector can decide the target motion-vector of the above-mentioned second interpolation block according to original position P1 ~ P4 and P1 ' ~ P4 ', this target motion-vector system actual image that decides/produce the second interpolation block.Specifically, please refer to Fig. 7, it is the original position P1 ~ P4 that determined of original position decision module 2201 and the example schematic of P1 ' ~ P4 '; Vector selects module 2202 meetings respectively with reference to P1 and P1 ', P2 and P2 ', P3 and P3 ' and P4 and P4 ' calculate reference vector V1 ~ V4, with among foundation reference vector V1 ~ V4 at least one from the candidate vector MV1 ' of the second interpolation block, choose one of them among the MV2 ' and decide this target motion-vector, wherein reference vector V1 ~ V4 is all corresponding or be comparable to the prospect motion-vector, and on real the work, it is one and candidate vector MV1 ' at least among comparison reference vector V1 ~ V4 that vector is selected module 2202, MV2 ', one of them is comparable to candidate vector MV1 ' as reference vector V1 ~ V4, during one of them of MV2 ', vector selects module 2202 can choose candidate vector MV1 ', another of MV2 ' is used as this target motion-vector, therefore, if the second interpolation block is physically located in the image area of coverage or image appears the district, the target motion-vector that is then determined is understood corresponding or is comparable to the background motion-vector, for example, if the second interpolation block is MB shown in Figure 7, then candidate vector MV1 ' is a correspondence or to be comparable to background motion-vector candidate vector MV2 ' then corresponding or be comparable to the prospect motion-vector, so, the candidate vector MV1 ' that vector selection module 2202 can be chosen correspondence or be comparable to the background motion-vector is as the target motion-vector, and the target motion-vector that is determined according to this produces the image of interpolation block MB; And if the second interpolation block is MB ' shown in Figure 7, then candidate vector MV1 ' is a correspondence or to be comparable to prospect motion-vector candidate vector MV2 ' then corresponding or be comparable to the background motion-vector, the candidate vector MV2 ' that vector selection module 2202 can be chosen correspondence or be comparable to the background motion-vector is as the target motion-vector, and the target motion-vector that is determined according to this produces the image of interpolation block MB '.
In addition, if the second interpolation block is actual is not to be positioned at the image area of coverage or image appears the district, then vector selects module 2202 still can determine correct motion-vector, for example, if the second interpolation block is MB shown in Figure 7 "; its candidate vector MV1 ' then; MV2 ' all is correspondence or is comparable to the background motion-vector; therefore; no matter whether decide interpolation block MB according to reference vector V1 ~ V4 " the target motion-vector, vector selects module 2202 to select its two second candidate vector MV1 ' arbitrarily, MV2 ' one of them as its target motion-vector, in fact for interpolation block MB " all be the motion-vector that is produced preferable background video picture; In addition, if the second interpolation block is MB shown in Figure 7 " '; then its candidate vector MV1 ', MV2 ' they all are correspondence or are comparable to the prospect motion-vector; vector select module 2202 select arbitrarily two candidate vector MV1 ', MV2 ' one of them as its target motion-vector, in fact for interpolation block MB " ' all be the motion-vector that is produced preferable foreground object image frame.
In another embodiment, also can only from candidate vector MV1 ', the MV2 ' of the second interpolation block, choose one of them and decide this target motion-vector, and this also falls into category of the present utility model by single reference vector; Be noted that, in this embodiment, for example if only decide the target motion-vector by reference vector V1, then vector selects module 2202 need not calculate reference vector V2 ~ V4, yet, still need determine original position P1 ~ P4 and P1 ' ~ P4 ', appear the actual imagery zone in district so that judge the image area of coverage and image, in other words, embodiment of the present utility model only by determining two original positions (wherein a group of P1 and P1 ', P2 and P2 ', P3 and P3 ' and P4 and P4 '), can determine the target motion-vector of the second interpolation block.
The actual operation that produces the image of the second interpolation block then is specified in down.At first because embodiment of the present utility model mainly is with the example that moves horizontally image as an illustration, therefore with motion-vector direction from left to right for just, otherwise its relative direction (from right to left) is then for bearing; Yet, this non-restriction of the present utility model, all on the two dimensional surface space with a certain specific direction for just, again with its relative direction for negative example, all meet spirit of the present utility model.Computing unit 205 is that the candidate vector with the second interpolation block deducts the resulting vector difference of its another candidate vector, judge that the second interpolation block is to fall within the image area of coverage or is to fall within image to appear the district, please refer again to Fig. 7, for instance, for interpolation block MB among Fig. 7, image generation module 225 in the computing unit 205 deducts its candidate vector MV2 ' with the candidate vector MV1 ' of interpolation block MB, resulting vector difference is for just, therefore judge that interpolation block MB system belongs to the image area of coverage, the image of then obtaining corresponding block among last the non-interpolation image Fn according to target motion-vector to the second interpolation image that is determined is used as the image of interpolation block MB; In addition, for interpolation block MB ' among Fig. 7, image generation module 225 deducts its candidate vector MV2 ' with the candidate vector MV1 ' of interpolation block MB ', resulting vector difference is for negative, therefore judge that block MB system belongs to image and appears the district, then be used as the image of interpolation block MB ' according to the image of obtaining a corresponding block among non-interpolation image Fn+1 after target motion-vector to the second interpolation image that is determined; In other words, judge that the second interpolation block is to fall into an image area of coverage or is to fall into an image to appear the district owing to can utilize the vector difference of two candidate vector of the second interpolation block, so, image generation module 225 can utilize the result of judgement, chooses one as the image of the second interpolation block before second interpolation image in the image of the image of a non-interpolation image and a back non-interpolation image.
In sum, at belonging to the block that the image area of coverage or image appear the district in the interpolation image, image processor 200 among the embodiment of the present utility model can correctly calculate the target motion-vector of this block, make to obtain preferable image when carrying out the image interpolation, help to promote the quality of image of carrying out the running of frame updating frequency inverted with this target motion-vector.
The above only is preferred embodiment of the present utility model, and all equalizations of being done according to the utility model claim change and modify, and all should belong to covering scope of the present utility model.
Claims (3)
1. an image processor when being used to carry out the motion picture interpolation, determines the image area of coverage/image of an interpolation image to appear the motion-vector in district, and this interpolation image is made up of plurality of blocks, it is characterized in that, this image processor includes:
One storage element is used for storing at least a last raw frames of this interpolation image and an initial position of back one raw frames and a last interpolation image; And
One computing unit is coupled to this storage element, includes:
One candidate vector generation module is used to determine plural number first candidate vector and plural second candidate vector of the plurality of blocks of this interpolation image; And
One motion-vector decision module, according to those first candidate vector of those blocks of this interpolation image, those second candidate vector this original position with this last interpolation image, the image area of coverage/image that determines to be positioned at this interpolation image appears a motion-vector of a block in district.
2. image processor as claimed in claim 1 is characterized in that, this candidate vector generation module includes:
One image difference value generation module according to the plural original motion-vector of a plurality of adjacent block of this block, determines an image difference value of this block; And
One candidate vector decision module according to the plural image difference value of the continuous block of one-dimensional space plural number, determines plural number first candidate vector and plural second candidate vector of those continuous blocks respectively.
3. image processor as claimed in claim 1 is characterized in that, this motion-vector decision module includes:
One initial determining positions module, utilize those first candidate vector of those blocks in this interpolation image and one of them of those second candidate vector, calculate the pairing plurality of blocks comparison of those blocks difference, and decide an initial position of this interpolation image according to the numerical value change of those block comparison differences; And
One vector is selected module, produce a reference vector according to this original position of this interpolation image and this original position of this last interpolation image, and relatively the image area of coverage/image of this reference vector and this interpolation image appears this first, second candidate vector of this block in the district, chooses one of them more not close this motion-vector as this block of this first, second candidate vector and this reference vector.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 200920005185 CN201374785Y (en) | 2009-03-04 | 2009-03-04 | Image processing device for deciding motion vector of interpolated picture |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 200920005185 CN201374785Y (en) | 2009-03-04 | 2009-03-04 | Image processing device for deciding motion vector of interpolated picture |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN201374785Y true CN201374785Y (en) | 2009-12-30 |
Family
ID=41500866
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN 200920005185 Expired - Lifetime CN201374785Y (en) | 2009-03-04 | 2009-03-04 | Image processing device for deciding motion vector of interpolated picture |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN201374785Y (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101753795B (en) * | 2008-12-16 | 2012-05-16 | 晨星软件研发(深圳)有限公司 | Image processing method for determining motion vector of interpolation image and related device |
| CN109327669A (en) * | 2017-08-01 | 2019-02-12 | 联咏科技股份有限公司 | Frame per second promotes conversion equipment and method |
| CN113726980A (en) * | 2020-05-25 | 2021-11-30 | 瑞昱半导体股份有限公司 | Image processing method |
-
2009
- 2009-03-04 CN CN 200920005185 patent/CN201374785Y/en not_active Expired - Lifetime
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101753795B (en) * | 2008-12-16 | 2012-05-16 | 晨星软件研发(深圳)有限公司 | Image processing method for determining motion vector of interpolation image and related device |
| CN109327669A (en) * | 2017-08-01 | 2019-02-12 | 联咏科技股份有限公司 | Frame per second promotes conversion equipment and method |
| CN113726980A (en) * | 2020-05-25 | 2021-11-30 | 瑞昱半导体股份有限公司 | Image processing method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102088589B (en) | Frame rate conversion using bi-directional, local and global motion estimation | |
| CN100389601C (en) | Video electronic flutter-proof device | |
| KR100720722B1 (en) | Intermediate vector interpolation method and 3D display apparatus | |
| US8774276B2 (en) | Image processing method for determining motion vectors of interpolated picture and related apparatus | |
| CN103220488A (en) | Up-conversion device and method of video frame rate | |
| KR101443701B1 (en) | Adaptive motion search range determining apparatus and method | |
| CN101692692A (en) | Method and system for electronic image stabilization | |
| GB2495179A (en) | Motion estimation using weighted candidate vectors from an external source | |
| KR20070069615A (en) | Motion estimator and motion estimating method | |
| KR0182058B1 (en) | Apparatus and method of multi-resolution circulating search for motion estimation | |
| CN103402098A (en) | Video frame interpolation method based on image interpolation | |
| CN103369208A (en) | Self-adaptive de-interlacing method and device | |
| US8447126B2 (en) | Image processing method and related apparatus | |
| US20100315550A1 (en) | Image frame interpolation device, image frame interpolation method, and image frame interpolation program | |
| CN201374785Y (en) | Image processing device for deciding motion vector of interpolated picture | |
| CN102362499A (en) | Image encoding apparatus and image encoding method | |
| CN102065263A (en) | Image interpolation processing apparatus and method thereof | |
| Huang et al. | Algorithm and architecture design of multirate frame rate up-conversion for ultra-HD LCD systems | |
| CN104469381A (en) | Implement system of VLSI adopting adaptive adjustment algorithm for H.264 motion estimation search window | |
| CN107124617B (en) | Method and system for generating random vector in motion estimation motion compensation | |
| CN101753795B (en) | Image processing method for determining motion vector of interpolation image and related device | |
| CN104580978A (en) | Video detecting and processing method and video detecting and processing device | |
| US8305500B2 (en) | Method of block-based motion estimation | |
| CN201222771Y (en) | High speed edge self-adapting de-interlaced interpolation device | |
| CN201444658U (en) | Image processing device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| AV01 | Patent right actively abandoned |
Granted publication date: 20091230 Effective date of abandoning: 20090304 |