CN102521794B - Image interpolation method and device based on spline surface - Google Patents
Image interpolation method and device based on spline surface Download PDFInfo
- Publication number
- CN102521794B CN102521794B CN201110354217.9A CN201110354217A CN102521794B CN 102521794 B CN102521794 B CN 102521794B CN 201110354217 A CN201110354217 A CN 201110354217A CN 102521794 B CN102521794 B CN 102521794B
- Authority
- CN
- China
- Prior art keywords
- pixel
- matrix
- interpolation
- image
- node
- 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.)
- Active
Links
Abstract
The invention discloses an image interpolation method and an image interpolation device based on a spline surface. The image interpolation method comprises the following steps that: a control point matrix containing pixels to be interpolated in an original image is selected; and the pixel value of each pixel of an interpolation matrix containing the pixels to be interpolated in an enlarged image through a spline surface function and the control point matrix. With the adoption of the image interpolation method based on the spline surface and the device, the spline surface function can be adopted to construct a curved surface so as to realize the interpolation of the image, serration generated by the conventional interpolation method can be well avoided, the values of reference points in the image enlarging process are moderately adjusted, so that the contrast ratio and the effect of the edge of the enlarged image are better improved, and the efficiency of image enlargement reconstruction is ensured through simplification of the algorithm.
Description
Technical field
The present invention relates to image processing techniques, particularly relate to the image interpolation method based on spline surface and device.
Background technology
In traditional images interpolation algorithm, arest neighbors interpolation is comparatively simple, and easily realize, time early stage, Application comparison is general.But the method can produce obvious jagged edges and mosaic phenomenon in new images.Bilinear interpolation has smoothing function, effectively can overcome the deficiency of nearest neighbor method, but the HFS of meeting degraded image, image detail is fogged.When enlargement factor is higher, high-order interpolation, as bicubic and cubic spline interpolation geometric ratio low order interpolation good.The continuity of the grey scale pixel value continuity original image grey scale change that these interpolation algorithms can make interpolation generate, thus make the deep or light change of enlarged image naturally level and smooth.But in the picture, between some pixel and neighbor there is sudden change in gray-scale value, namely there is gray scale uncontinuity.These pixels with gray-scale value sudden change are exactly the profile of description object or the edge pixel of texture image in image.In Nonlinear magnify, these are had to the pixel of discontinuous gamma characteristic, if adopt conventional interpolation algorithm to generate the pixel newly increased, will certainly make the profile of enlarged image and texture fuzzy, reduction picture quality.
In order to overcome the deficiency of classic method; propose many edge-protected interpolation methods; certain enhancing is had to the edge of interpolation image; make the visual effect of image better, edge-protected interpolation method can be divided into two classes: based on the method at original low resolution image edge and the method based on high-definition picture edge after interpolation.Method based on original low-resolution image edge: (1) first detects the edge of low-resolution image, then according to the edge detected by pixel classifications process, for the pixel of flat site, adopt classic method interpolation; For the pixel of fringe region, design special interpolation method, to reach the object keeping edge details.(2) based on this kind of interpolation method of method at high-definition picture edge after interpolation: first adopt classic method interpolation low-resolution image, then the edge of high-definition picture is detected, last edge and neighbouring pixel carry out special processing, fuzzy to remove, strengthen the edge of image, this convergent-divergent algorithm, better treatment effect can be obtained in edge, details etc., obtain high-precision result, but can only sacrifice speed, adopt the algorithm that complexity is higher.
Consider above analysis, when carrying out image scaling with image interpolation algorithm, always there is image procossing precision and Image Reconstruction speed two key elements of runing counter to are difficult to take into account.
Summary of the invention
In order to solve the problems of the technologies described above, the invention provides the image interpolation method based on spline surface and device, the algorithm of simplification can be utilized to improve Image Reconstruction speed and promote picture processing precision.
The invention provides the image interpolation method based on spline surface, comprising:
The interpolating matrix comprising interpolation pixel is selected in original image;
Obtain each node in interpolating matrix described in enlarged image relative to the horizontal stroke of described interpolation pixel, advance from;
According to the horizontal stroke of each node, advance from, calculated the off-set value of each node in described interpolating matrix by spline surface function and described interpolating matrix;
After the difference of the described pixel value of interpolation pixel and the off-set value of each node is multiplied by default scaling coefficient, sue for peace with the pixel value of this interpolation pixel again, obtain the pixel value of each node in described interpolating matrix: Q=P+a × (P-Q '), wherein Q is the pixel value of each node in described interpolating matrix, P is the pixel value of interpolation pixel, a is default scaling coefficient, the off-set value that Q ' is each node.
Correspondingly, present invention also offers the image interpolation device based on spline surface, comprising:
Gating matrix chooses unit, for selecting the interpolating matrix comprising interpolation pixel in original image;
The interpolating matrix computing unit that unit is connected is chosen, for obtaining by spline surface function and described interpolating matrix the pixel value comprising each node in the interpolating matrix of described interpolation pixel in enlarged image with described gating matrix; Described interpolating matrix computing unit comprises:
Relative distance acquiring unit, for obtain each node in interpolating matrix described in enlarged image relative to the horizontal stroke of described interpolation pixel, advance from;
Splines unit, for selecting spline surface function;
The offset computation unit that unit is connected respectively is chosen with described relative distance acquiring unit, splines unit and described gating matrix, for the horizontal stroke according to each node, advance from, calculated the off-set value of each node in described interpolating matrix by spline surface function and described interpolating matrix;
The node pixel value determining unit be connected with described offset computation unit, after the difference of the described pixel value of interpolation pixel and the off-set value of each node is multiplied by default scaling coefficient, sue for peace with the pixel value of this interpolation pixel again, obtain the pixel value of each node in described interpolating matrix: Q=P+a × (P-Q '), wherein Q is the pixel value of each node in described interpolating matrix, P is the pixel value of interpolation pixel, a is default scaling coefficient, the off-set value that Q ' is each node.
Implement the present invention, there is following beneficial effect:
Because the present invention adopts Construction of Operator Spline Interpolation curved surface to realize the interpolation of image, the sawtooth that traditional interpolation method produces can be evaded well, the value of reference point is suitably adjusted in the process of Nonlinear magnify, the effect of the contrast and edge of amplifying rear image is improved preferably, and by shortcut calculation, ensure that the efficiency that Nonlinear magnify reconstructs.
Accompanying drawing explanation
Fig. 1 is the process flow diagram of the image interpolation method that the present invention is based on spline surface;
Fig. 2 is the embodiment process flow diagram of the image interpolation method that the present invention is based on spline surface;
Fig. 3 is the embodiment schematic diagram of the image interpolation method that the present invention is based on spline surface;
Fig. 4 is the schematic diagram of the image interpolation device that the present invention is based on spline surface;
Fig. 5 is one of embodiment schematic diagram of the image interpolation device that the present invention is based on spline surface;
Fig. 6 is the embodiment schematic diagram two of the image interpolation device that the present invention is based on spline surface.
Embodiment
For making the object, technical solutions and advantages of the present invention clearly, below in conjunction with accompanying drawing, the present invention is described in further detail.
Fig. 1 is the process flow diagram of the image interpolation method that the present invention is based on spline surface, comprising:
S101: select the reference mark matrix comprising interpolation pixel in original image;
S102: obtain the pixel value comprising each pixel in the interpolating matrix of described interpolation pixel in enlarged image by spline surface function and described reference mark matrix.
Different from traditional image interpolation mode, the present invention creatively by the building method of spline surface, applies to image interpolation.By shortcut calculation, greatly improve the reconstruct efficiency of image, the large multiplying power of image is amplified, Be very effective.
In interpolation problem, spline interpolation is usually handy than polynomial interpolation.Can produce by the spline interpolation of low order and effect that the polynomial interpolation of high-order is similar, and the appearance of the numerical value instability being called as imperial lattice phenomenon can be avoided.And the critical nature that the spline interpolation of low order also has " protecting convex ".In the computer-aided design (CAD) and computer graphics of computer science, batten typically refers to the polynomial parametric curves of segmentation definition.Due to batten simple structure, easy to use, matching is accurate, and shape that can be complicated in curve of approximation matching and interactive Curve Design, batten is the conventional method for expressing of curve in these fields.
In order to apply splines in image interpolation, need in original image, to select a reference mark matrix, for coordinating the calculating of splines.Based on different spliness, the shape of reference mark matrix and size can be different, are generally square formation, such as 3 × 3 matrixes or 4 × 4 matrixes.In original image, depending on the interpolation pixel that each pixel is relatively independent, in original image, select the reference mark matrix comprising interpolation pixel, original image becomes enlarged image through process, and namely this enlarged image is made up of the interpolating matrix comprising each interpolation pixel.And in enlarged image, the pixel value comprising interpolating matrix each pixel interior of each interpolation pixel is obtained by spline surface function and described reference mark matrix computations.
Fig. 2 is the embodiment process flow diagram of the image interpolation method that the present invention is based on spline surface, comprising:
S201: select the reference mark matrix comprising interpolation pixel in original image;
S202: obtain each node in interpolating matrix described in enlarged image relative to the horizontal stroke of described interpolation pixel, advance from;
According to the enlargement factor of length and width, obtain described interpolation matrix column number, line number respectively;
According to the normalized cumulant of interpolation knot relative to interpolation pixel, obtain the horizontal stroke of this interpolation knot, advance from.
S203: according to the horizontal stroke of each node, advance from, by the off-set value of each node in interpolating matrix described in spline surface function and described reference mark matrix computations;
Each pixel in the matrix of described reference mark, as interpolation pixel, regains pixel value.
S204: after the difference of the pixel value of described interpolation pixel and the off-set value of interpolation node is multiplied by default scaling coefficient, then sue for peace with the pixel value of this interpolation node, obtain the pixel value of each interpolation node in described interpolating matrix.
Compared to Figure 1, Fig. 2 is the particular flow sheet of embodiment.Fig. 3 is the embodiment schematic diagram of the image interpolation method that the present invention is based on spline surface.Below in conjunction with Fig. 3, the present invention is described further.
Be illustrated in figure 3 the pixel distribution plan of enlarged image, wherein " ⊙ " represents the pixel of original image.Suppose P
1,1for interpolation pixel, in original image, select the reference mark matrix comprising interpolation pixel.Preferably, in the middle of the present embodiment, the size of this reference mark matrix is 4 × 4 square formations, and interpolation pixel is positioned at the first row first row of reference mark matrix.Similarly, this reference mark matrix also 3 × 3 square formations or adopt other matrix form, the number row of presetting that interpolation pixel can be positioned at reference mark matrix presets ordered series of numbers, only needs to do corresponding conversion, then can solve technical matters of the present invention equally in the middle of follow-up calculating.
It should be noted that, as interpolation point P
1,1after deciding, comprise described interpolation pixel P
1,1interpolating matrix can also decide.U × v matrix is as shown in Figure 3 and P
1,1corresponding interpolating matrix, obviously, the interpolating matrix that namely enlarged image comprises interpolation pixel by each formed.Next need to calculate in the middle of each interpolating matrix, the pixel value of the new pixel inserted.
At this, each pixel in interpolating matrix is referred to as node, newly to insert node Q (" ■ " as shown in Figure 3), now, can acquisition point Q relative to the horizontal stroke of interpolation pixel, advance from being respectively uQ=1; VQ=2.UQ, uV illustrate in enlarged image, the distance between the interpolation pixel of new insertion point and original image, and uQ, uV parameter can be used to calculate Q point relative to P
1,1pixel value, embody P
1,1to between a Q similar/degree of correlation.
Owing to comprising described interpolation pixel in the middle of each pixel in described interpolating matrix, according to step S202, described interpolation pixel relative to self horizontal stroke, advance is from being all zero; According to step S203, the off-set value of described interpolation pixel can be calculated by spline surface function and described reference mark matrix.According to step S204, readjust the pixel value of described interpolation pixel.So analogize, each pixel in the matrix of described reference mark, as interpolation pixel, regains pixel value, adjustment reference mark matrix.In the middle of step S203, calculate with the reference mark matrix after adjusting, the contrast of enlarged image can be strengthened.It should be noted that, work as P
1,2as interpolation pixel, regain pixel value NewP
1,2time, for calculating P
1,2reference mark matrix be different from and P
1,1corresponding reference mark matrix, only provides NewP below
1,2account form, the account form of all the other each pixels is similar in the matrix of described reference mark.
NewP
1,2=P
1,2+a×(P
1,2-P
1,2(u,v))
According to the enlargement factor of length and width, obtain described interpolation matrix column number, line number respectively.As shown in Figure 3, the present embodiment amplifies long four times, and amplify wide four times, namely original image amplifies 16 times.Preferably, each interpolation matrix is 4 × 4 matrixes.The present invention is not limited thereto, amplify length and width if can adopt not doubly, also can the enlargement factor of length and width different.Obviously, when the enlargement factor of length and width is different, image, except amplifying, also has the effect of " elongation ".In order to embody Q and interpolation pixel P
1,1and with interpolation pixel P
1,1adjacent pixel similar/degree of correlation, according to the normalized cumulant of interpolation knot relative to interpolation pixel, obtain the horizontal stroke of this interpolation knot, advance from.According to the concept of existing normalized cumulant, when described interpolation matrix column number, line number be four determine after, can by above-mentioned horizontal stroke, advance from being modified to uQ=1/4; VQ=2/4=1/2.
Described spline surface function comprises B-spline surface function; B-spline surface is one curved surface very flexibly, and the local shape of curved surface is very directly perceived by the control of respective vertices.If these summit control technologys are used well, whole B-spline surface can be made to meet some special technical requirements at some position.It can thus be appreciated that the characteristic of B-spline surface meets in the inventive method well to the operational requirements that pixel adjusts accordingly.In addition, described spline surface function can also comprise NURBS(Non-Uniform Rational B-Splines) i.e. non-uniform rational B-spline method, it and B-spline method both unifications mutually of Freeform Surface are described, again can Precise Representation conic arc and quadric surface.Its treatment effect is slightly better than B-spline surface.
When described spline surface is Bicubic B-Spline Surfaces, described reference mark matrix is 4 × 4 matrixes; When described spline surface is Biquadratic B-spline curved surface, described reference mark matrix is 3 × 3 matrixes.In the middle of the present embodiment, adopt Bicubic B-Spline Surfaces, preferably, in order to simplified operation process, its function adopts Q (u, v)=UBPB
-1v.Wherein, U, V correspond to the horizontal stroke of each pixel, advance from; P corresponds to described reference mark matrix; B is parameter matrix, can rule of thumb or the different parameter matrix of Setup Experiments; Q (u, v) is off-set value.
Preferably, the spline surface function that the present embodiment adopts is:
Carry out amplification interpolation at utilization spline surface function, corresponding adjustment has been done to original pixels, can figure have been added
The contrast of picture.Certainly, can still calculate with original pixels, but acquisition is secondary good effect,
Now, the computing formula of above-mentioned spline surface function becomes:
As shown in Figure 3, the P in above-mentioned formula
1,1~ P
4,4corresponding with the original image pixel in reference mark matrix in Fig. 3.For a Q, u=uQ=1/4 in above-mentioned formula; V=vQ=1/2.Accordingly, off-set value Q ' can be obtained.Certainly, when described spline surface is Biquadratic B-spline curved surface, described reference mark matrix is 3 × 3 matrixes.Also for Fig. 3, will corresponding to P
1,1~ P
3,3reference mark matrix, now, according to the function of Biquadratic B-spline curved surface, the matrix of U, V, B is all different.
Finally, the pixel value Q=P+a of the new node Q inserted × (P-Q '), wherein, P is the pixel value of interpolation pixel, and a is default scaling coefficient, and Q '=Q (u, v) is above-mentioned off-set value.Connect described in example, that corresponding with the node Q that Fig. 3 newly inserts is interpolation pixel P
1,1pixel value.
Adopt in the same way, calculate P
1,1the pixel value of all the other each points in corresponding interpolating matrix, the final pixel value obtaining enlarged image each point.
When obtaining the pixel value of all the other each points of enlarged image, may run into following problem: in original image, the reference mark matrix chosen according to certain rule is beyond the pixel coverage of original image.Such as, connect described in example, 4 × 4 matrixes being the first row first row with interpolation pixel, for the bottom-right pixel of original image, this matrix will exceed the pixel coverage of original image.Certainly, also can not choose reference mark matrix with unified rule to avoid the generation of this problem, such as, for the pixel in the original image lower right corner, 4 × 4 matrixes arranged for fourth line the 4th with interpolation pixel can be selected; For the pixel in the original image lower left corner, 4 × 4 matrixes that can to select with interpolation pixel be fourth line first row; For the pixel in the original image upper right corner, 4 × 4 matrixes arranged for the first row the 4th with interpolation pixel can be selected.But obviously, this needs to increase some determining steps, at least need to know whether interpolation pixel is positioned at the edge of original image, so, be unfavorable for simplified operation.In the middle of the present embodiment, preferably, adopt image overflow processing technique conventional in the industry, when the reference mark matrix comprising interpolation pixel selected exceeds the pixel coverage of original image, for the pixel exceeding part, fill into the pixel of the original image adjacent with its position.
Fig. 4 is the schematic diagram of the image interpolation device that the present invention is based on spline surface, comprising:
Gating matrix chooses unit, for selecting the reference mark matrix comprising interpolation pixel in original image;
The interpolating matrix computing unit that unit is connected is chosen, for obtaining by spline surface function and described reference mark matrix the pixel value comprising each pixel in the interpolating matrix of described interpolation pixel in enlarged image with described gating matrix.
Fig. 4 and Fig. 1 is corresponding, the method for operation of each unit and identical in method in Fig. 4.
Fig. 5 is one of embodiment schematic diagram of the image interpolation device that the present invention is based on spline surface;
As shown in Figure 5, described interpolating matrix computing unit comprises:
Relative distance acquiring unit, for obtain each node in interpolating matrix described in enlarged image relative to the horizontal stroke of described interpolation pixel, advance from;
Splines unit, for selecting spline surface function;
The offset computation unit that unit is connected respectively is chosen with described relative distance acquiring unit, splines unit and described gating matrix, for the horizontal stroke according to each node, advance from, by the off-set value of each node in interpolating matrix described in spline surface function and described reference mark matrix computations;
The node pixel value determining unit be connected with described offset computation unit, after the difference of the pixel value of described interpolation pixel and the off-set value of interpolation node is multiplied by default scaling coefficient, sue for peace with the pixel value of this interpolation node again, obtain the pixel value of each interpolation node in described interpolating matrix.
Fig. 5 and Fig. 2 is corresponding, the method for operation of each unit and identical in method in Fig. 5.
Fig. 6 is the embodiment schematic diagram two of the image interpolation device that the present invention is based on spline surface.
As shown in Figure 6, described splines unit comprises:
The B-spline function unit that unit is connected is chosen, for selecting B-spline surface function with described gating matrix; Described gating matrix chooses unit, also for when described spline surface is Bicubic B-Spline Surfaces, determines that described reference mark matrix is 4 × 4 matrixes; When described spline surface is Biquadratic B-spline curved surface, determine that described reference mark matrix is 3 × 3 matrixes.
As shown in Figure 6, the shown image interpolation device based on spline surface also, comprising:
The interpolating matrix tectonic element be connected with described relative distance acquiring unit, for the enlargement factor according to length and width, obtains described interpolation matrix column number, line number respectively;
Described relative distance acquiring unit, also for according to the normalized cumulant of interpolation knot relative to interpolation pixel, obtain the horizontal stroke of this interpolation knot, advance from.
Choosing with described gating matrix the edge treated unit that unit is connected, during for exceeding the pixel coverage of original image when the reference mark matrix comprising interpolation pixel selected, for the pixel exceeding part, filling into the pixel of the original image adjacent with its position.
Fig. 6 is the preferred implementation in the middle of embodiments of the invention, the method for operation of unit and corresponding in method in Fig. 6.
Through the above description of the embodiments, those skilled in the art can be well understood to the mode that the present invention can add required hardware platform by software and realize, and can certainly all be implemented by hardware.Based on such understanding, what technical scheme of the present invention contributed to background technology can embody with the form of software product in whole or in part, this computer software product can be stored in storage medium, as ROM/RAM, magnetic disc, CD etc., comprising some instructions in order to make a computer equipment (can be personal computer, server, or the network equipment etc.) perform the method described in some part of each embodiment of the present invention or embodiment.
Above-described embodiment of the present invention, does not form limiting the scope of the present invention.Any amendment done within the spirit and principles in the present invention, equivalent replacement and improvement etc., all should be included within claims of the present invention.
Claims (10)
1. based on an image interpolation method for spline surface, it is characterized in that, comprising:
The reference mark matrix comprising interpolation pixel is selected in original image;
Described original image becomes enlarged image through process, described enlarged image is made up of the interpolating matrix comprising each interpolation pixel, obtain each node in interpolating matrix described in enlarged image relative to the horizontal stroke of described interpolation pixel, advance from, in described interpolating matrix, each pixel is referred to as node;
According to the horizontal stroke of each node, advance from, by the off-set value of each node in interpolating matrix described in spline surface function and described reference mark matrix computations, the function of off-set value adopts Q (u, v)=UBPB
-1v, wherein, U, V correspond to the horizontal stroke of each pixel, advance from, P corresponds to described reference mark matrix, and B is parameter matrix, and Q (u, v) is off-set value, and Q '=Q (u, v) is above-mentioned off-set value;
After the difference of the described pixel value of interpolation pixel and the off-set value of each node is multiplied by default scaling coefficient, sue for peace with the pixel value of this interpolation pixel again, obtain the pixel value of each node in described interpolating matrix: Q=P+a × (P-Q '), wherein Q is the pixel value of each node in described interpolating matrix, P is the pixel value of interpolation pixel, a is default scaling coefficient, the off-set value that Q ' is each node.
2. the image interpolation method based on spline surface according to claim 1, it is characterized in that: the described horizontal stroke according to each node, advance from, in step by the off-set value of each node in interpolating matrix described in spline surface function and described reference mark matrix computations, each node in described interpolating matrix, as interpolation pixel, regains pixel value.
3. the image interpolation method based on spline surface according to claim 1 and 2, is characterized in that, in interpolating matrix described in described acquisition enlarged image each node relative to the horizontal stroke of described interpolation pixel, advance from step, comprising:
According to the enlargement factor of length and width, obtain described interpolation matrix column number, line number respectively;
According to the normalized cumulant of interpolation knot relative to interpolation pixel, obtain the horizontal stroke of this interpolation knot, advance from.
4. the image interpolation method based on spline surface according to claim 1 and 2, is characterized in that: described spline surface function comprises B-spline surface function and non-uniform rational B-spline;
When described spline surface is Bicubic B-Spline Surfaces, described reference mark matrix is 4 × 4 matrixes;
When described spline surface is Biquadratic B-spline curved surface, described reference mark matrix is 3 × 3 matrixes.
5. the image interpolation method based on spline surface according to claim 3, is characterized in that: described spline surface function comprises B-spline surface function and non-uniform rational B-spline;
When described spline surface is Bicubic B-Spline Surfaces, described reference mark matrix is 4 × 4 matrixes;
When described spline surface is Biquadratic B-spline curved surface, described reference mark matrix is 3 × 3 matrixes.
6. the image interpolation method based on spline surface according to claim 4, is characterized in that, describedly in original image, selects to comprise the step of the reference mark matrix of interpolation pixel, comprising:
When the reference mark matrix comprising interpolation pixel selected exceeds the pixel coverage of original image, for the pixel exceeding part, fill into the pixel of the original image adjacent with its position.
7. the image interpolation method based on spline surface according to claim 5, is characterized in that, describedly in original image, selects to comprise the step of the reference mark matrix of interpolation pixel, comprising:
When the reference mark matrix comprising interpolation pixel selected exceeds the pixel coverage of original image, for the pixel exceeding part, fill into the pixel of the original image adjacent with its position.
8., based on an image interpolation device for spline surface, it is characterized in that, comprising:
Gating matrix chooses unit, for selecting the reference mark matrix comprising interpolation pixel in original image;
The interpolating matrix computing unit that unit is connected is chosen, for obtaining by spline surface function and described reference mark matrix the pixel value comprising each node in the interpolating matrix of described interpolation pixel in enlarged image with described gating matrix; Described interpolating matrix computing unit comprises:
Relative distance acquiring unit, for obtain each node in interpolating matrix described in enlarged image relative to the horizontal stroke of described interpolation pixel, advance from, described original image becomes enlarged image through process, described enlarged image is made up of the interpolating matrix comprising each interpolation pixel, and in described interpolating matrix, each pixel is referred to as node;
Splines unit, for selecting spline surface function;
The offset computation unit that unit is connected respectively is chosen with described relative distance acquiring unit, splines unit and described gating matrix, for the horizontal stroke according to each node, advance from, by the off-set value of each node in interpolating matrix described in spline surface function and described reference mark matrix computations, the function of off-set value adopts Q (u, v)=UBPB
-1v, wherein, U, V correspond to the horizontal stroke of each pixel, advance from, P corresponds to described reference mark matrix, and B is parameter matrix, and Q (u, v) is off-set value, and Q '=Q (u, v) is above-mentioned off-set value;
The node pixel value determining unit be connected with described offset computation unit, after the difference of the described pixel value of interpolation pixel and the off-set value of each node is multiplied by default scaling coefficient, sue for peace with the pixel value of this interpolation pixel again, obtain the pixel value of each node in described interpolating matrix: Q=P+a × (P-Q '), wherein Q is the pixel value of each node in described interpolating matrix, P is the pixel value of interpolation pixel, a is default scaling coefficient, the off-set value that Q ' is each node.
9. the image interpolation device based on spline surface according to claim 8, is characterized in that, comprising:
The interpolating matrix tectonic element be connected with described relative distance acquiring unit, for the enlargement factor according to length and width, obtains described interpolation matrix column number, line number respectively;
Described relative distance acquiring unit, also for according to the normalized cumulant of interpolation knot relative to interpolation pixel, obtain the horizontal stroke of this interpolation knot, advance from.
10. the image interpolation device based on spline surface according to claim 8 or claim 9, is characterized in that, comprising: B-spline function unit and/or edge treated unit;
Described B-spline function unit is connected to described splines unit and described gating matrix is chosen between unit, for selecting B-spline surface function; Described gating matrix chooses unit, also for when described spline surface is Bicubic B-Spline Surfaces, determines that described reference mark matrix is 4 × 4 matrixes; When described spline surface is Biquadratic B-spline curved surface, determine that described reference mark matrix is 3 × 3 matrixes;
Described edge treated unit is chosen unit with described gating matrix and is connected, during for exceeding the pixel coverage of original image when the reference mark matrix comprising interpolation pixel selected, for the pixel exceeding part, fill into the pixel of the original image adjacent with its position.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201110354217.9A CN102521794B (en) | 2011-11-10 | 2011-11-10 | Image interpolation method and device based on spline surface |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201110354217.9A CN102521794B (en) | 2011-11-10 | 2011-11-10 | Image interpolation method and device based on spline surface |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN102521794A CN102521794A (en) | 2012-06-27 |
| CN102521794B true CN102521794B (en) | 2015-01-21 |
Family
ID=46292699
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201110354217.9A Active CN102521794B (en) | 2011-11-10 | 2011-11-10 | Image interpolation method and device based on spline surface |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102521794B (en) |
Families Citing this family (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104346770A (en) | 2013-07-24 | 2015-02-11 | 联咏科技股份有限公司 | Data interpolation method and data interpolation system |
| CN103646422B (en) * | 2013-12-19 | 2016-03-09 | 哈尔滨工程大学 | Based on the 3 D displaying method of hereditary Multi-B Spline interpolation algorithm |
| CN104318621B (en) * | 2014-10-23 | 2017-12-19 | 中国船舶工业集团公司第七〇八研究所 | Ship hull surface reconstructing method based on non-uniform rational B-spline surface interpolation |
| CN104933679B (en) * | 2015-07-06 | 2018-07-24 | 福州瑞芯微电子股份有限公司 | A kind of method and its correspondence system of enlarged drawing |
| CN106204444B (en) * | 2016-06-29 | 2019-06-25 | 青岛海信移动通信技术股份有限公司 | A kind of method and apparatus of image amplification |
| CN108596835B (en) * | 2018-05-04 | 2022-04-26 | 嘉兴南湖学院 | Image amplification method for circular area |
| CN109816590B (en) * | 2018-12-26 | 2023-03-14 | 呈像科技(北京)有限公司 | Image extrapolation processing method |
| CN109727196B (en) * | 2018-12-26 | 2023-06-02 | 呈像科技(北京)有限公司 | Image interpolation processing method |
| CN110264401A (en) * | 2019-05-16 | 2019-09-20 | 平安科技（深圳）有限公司 | Continuous type image magnification method, device and storage medium based on radial basis function |
| CN111462006B (en) * | 2020-03-31 | 2023-06-20 | 华南理工大学 | Multi-target image complement method |
| CN113706625A (en) * | 2021-07-28 | 2021-11-26 | 昆山丘钛微电子科技股份有限公司 | Lens distortion correction method and device |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101833787B (en) * | 2010-04-13 | 2011-09-07 | 北京农业信息技术研究中心 | Plant leaf modeling method based on ball B-spline |
| CN101908235A (en) * | 2010-06-23 | 2010-12-08 | 苏州科技学院 | Novel method for modeling B spline surface |
-
2011
- 2011-11-10 CN CN201110354217.9A patent/CN102521794B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN102521794A (en) | 2012-06-27 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102521794B (en) | Image interpolation method and device based on spline surface | |
| Li et al. | Cubic mean value coordinates. | |
| Chen et al. | Computing the minimum distance between a point and a NURBS curve | |
| Lin et al. | An extended iterative format for the progressive-iteration approximation | |
| EP1453007B1 (en) | Color gradient paths | |
| EP1750229B1 (en) | Process for creating from a mesh an isotopologic set of parameterized surfaces | |
| EP1750228A2 (en) | Process for creating a parametric surface having a required geometrical continuity | |
| CN103927727B (en) | Method for converting scalar image into vector image | |
| WO2007083602A1 (en) | Interpolation processing method, interpolation processing device, shape evaluation method, and shape evaluation device | |
| CN104732479B (en) | Resizing an image | |
| Loseille et al. | Vizir: High-order mesh and solution visualization using OpenGL 4.0 graphic pipeline | |
| CN105139338B (en) | The generation method and device of multidimensional lookup table, image zoom processing method and device | |
| Šír et al. | On de Casteljau-type algorithms for rational Bézier curves | |
| Ruprecht et al. | Deformed cross‐dissolves for image interpolation in scientific visualization | |
| Concheiro et al. | Interactive rendering of NURBS surfaces | |
| US20040156556A1 (en) | Image processing method | |
| CN103456031A (en) | Novel method for area image interpolation | |
| Amanatiadis et al. | Design and implementation of a fuzzy area-based image-scaling technique | |
| Yang | Matrix weighted rational curves and surfaces | |
| Rossignac et al. | J-splines | |
| Piegl et al. | Circle approximation using integral B-splines | |
| Borouchaki et al. | Surface mesh enhancement with geometric singularities identification | |
| Tai et al. | Interpolation with slackness and continuity control and convexity-preservation using singular blending | |
| Costantini et al. | A local shape-preserving interpolation scheme for scattered data | |
| Liu et al. | Constructing n-sided toric surface patches from boundary curves |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CP03 | Change of name, title or address |
Address after: 510670 Guangdong city of Guangzhou province Kezhu Guangzhou high tech Industrial Development Zone, Road No. 233 Patentee after: Wei Chong group Limited by Share Ltd Address before: 510663 Guangzhou province high tech Industrial Development Zone, Guangdong, Cai road, No. 6, No. Patentee before: Guangdong Weichuangshixun Science and Technology Co., Ltd. |
|
| CP03 | Change of name, title or address |