CN102930565A - Construction method for discrete degradation image of turn-back motion target in static background - Google Patents

Abstract

The invention relates to a construction method for a discrete degradation image of a turn-back motion target in a static background, belonging to one part using more than one image in the field of image data processing or generation, in particular to a construction method for a discrete motion blur image. The method comprises the following steps of dividing a turn-back motion process of a target image into m sections of one-way motion processes; constructing ni sub-images figi j (j is equal to 1, 2,..., ni) according to the distances among ni pixels of the target image motion in the ith section of one-way motion process in the premise that the edge of the target image does not exceed the edge of a static background image all the time; and performing weighted linear superposition according to the following formula: in the formula, wi and j are weighting coefficients, and fig is the constructed discrete degradation image. The construction method for the discrete degradation image has the advantages that the operation time is short; a degradation process is intuitive and convenient to understand; the image is not required to be readjusted; the motion blur image which is formed by motion targets occupying partial view fields can be simulated, and the phenomenon that information on both sides of the image is superposed is prevented.

Description

The discrete degraded image building method of fold return motion target in a kind of static background

Technical field

In a kind of static background the discrete degraded image building method of fold return motion target belong to that general view data is processed or the generation field in by using the part more than piece image, relate in particular to a kind of discrete motion blurred picture building method.

Background technology

If have relative motion between imageing sensor and the target in imaging process, the gained image will produce the motion blur phenomenon.At daily life, commercial production, aerospace field, this phenomenon is very general.Although motion blur image can present the aesthetic feeling of art in some special dimension, but in most fields such as communications and transportation, commercial production, motion blur image but can only make troubles to the identification of target in the image and to obtaining of target detail information to us.Such as the electronic eyes in the traffic and transport field, bring difficulty to the identification of license plate number if photograph the fog-level of image, so just be difficult to vehicles peccancy is pressed chapter punishment, be unfavorable for the conventional maintenance of traffic order, cause potential safety hazard for people's life.

Avoid the generation of motion blur phenomenon in the image, generally adopt anti-shake technology, anti-shake technology comprises the anti-shake and electronic flutter-proof of optics, optics is anti-shake to be divided into again that camera lens is anti-shake and imaging is anti-shake, anti-shake the referring to of camera lens arranges special anti-shake compensating glass group in camera lens, according to jitter direction and the degree of camera, and compensating glass group corresponding adjustment position and angle, make light path keep stable, such as the EF IS of Canon series camera lens, Nikon VR series camera lens, suitable horse OS series camera lens; The anti-shake image device that refers to of imaging is behind the perception camera shake, and the position or the angle that change image device are held in the stable of picture, and this technology was widely used in the digital camera epoch.Electronic flutter-proof refers to by one-tenth's image is analyzed, and the anti-shake technology of then utilizing algorithm that image is compensated, this technology are actually by reducing image quality comes compensate for jitter, attempts to obtain between image quality and float an equilibrium point.Electronic flutter-proof is compared with optics is anti-shake, has a cost low, the characteristics of weak effect, so electronic flutter-proof only is used in the low side camera.But, but more be subjected to the concern of academia than the anti-shake technology of optics about the algorithm of electronic flutter-proof.

Algorithm for electronic flutter-proof, it is exactly the restoration algorithm of corresponding academia motion blur image, present stage, restoration algorithm was very many, traditional liftering algorithm, Wiener filtering algorithm are arranged, also have numerous blind restoration algorithms such as Kalman filtering algorithm and convex set projection method, up to now, still there is improved new algorithm to continue to bring out out.In order to verify the adaptability of these new restoration algorithms, need to restore the only different degraded image of degradation parameter, and compare with original image.In each collection image, although we can artificially set the degradation parameter of image as required, but can't avoid the impact of random noise, so that the image sequence that actual acquisition arrives is except the degradation parameter difference, will inevitably be subject to the impact of random noise, therefore can't actual acquisition arrive the only different degraded image of degradation parameter.

The way that overcomes this problem is very simple, with the mode of software simulation the different degenrate functions of original non degenerate imagery exploitation is manually degenerated exactly.By people's works such as Gonzalez, the people such as Ruan Qiuqi translation, and summed up the method for two kinds of artificial degraded images that prior art adopts in " Digital Image Processing " book of being published by China Machine Press:

The first is spatial domain convolution degeneration method, if original image is f (x, y), degenrate function is h (x, y), and then degraded image g (x, y) is expressed as:

g(x,y)=f(x,y)*h(x,y)

In the formula, " * " represents convolution algorithm; For the discrete picture of M * N, the first spatial domain convolution way of degeneration obtains the process of degraded image and can further be write as:

$g(x,y)=\underset{m=1}{\overset{M}{\mathrm{\Σ}}}\underset{n=1}{\overset{N}{\mathrm{\Σ}}}f(m,n)h(x-m,y-n)$

In the formula, x=1,2 ..., M; Y=1,2 ..., N.Can know according to top formula, calculate discrete degraded image g (x, y), need to be to x, y, m, n finish the quadruple loop computation and could realize, the quadruple loop computation is so that the computation process of discrete degraded image g (x, y) is very consuming time, and this is the shortcoming of spatial domain convolution way of degeneration.

The second is the frequency domain Fourier method of degenerating, if the frequency spectrum of original image f (x, y) is F (u, v), the frequency spectrum of degenrate function h (x, y) is H (u, v), and then the frequency spectrum designation of degraded image g (x, y) is:

G(u,v)=F(u,v)H(u,v)

In the formula, u=1,2 ..., M; V=1,2 ..., N.Because the existence of Fast Fourier Transform (FFT) method, so that frequency domain Fourier degeneration method is compared spatial domain convolution degeneration method significantly lifting was arranged in operation time, yet, this method also has the shortcoming of himself: at first, whole degenerative process is finished in frequency domain, degenerative process is neither directly perceived, is difficult for again understanding; Secondly, carry out inverse Fourier transform by the frequency spectrum G (u, v) of degraded image g (x, y) and obtain in the process of degraded image g (x, y), also need the image border is moved to the center, otherwise do not go up with real image is corresponding.

In addition, can know according to the basic theories of Fourier transform, for the finite image of a width of cloth length and width, in the edge of its image, owing to have very abundant high-frequency information, so that the frequency spectrum of this image is without several.Yet, before carrying out Fourier transform, capital image that length and width are finite is respectively take its row and column as the cycle, carry out periodic extension, continuation becomes the infinite image with cyclophysis of length and width, image after the continuation just has finite frequency spectrum, and the quantity of frequency spectrum is identical with the element number that original image ranks direction has.This method can realize description to original image so that adopt with the frequency spectrum sequence that the ranks element number is identical in the original image.

Yet this method has the shortcoming of himself, after image carried out periodic extension, one side of image will be bordered with opposite side, for the motion blur image degenerate problem, on direction of motion, the information of image one side will add in the opposite side information of bordering on it, two superimposed phenomenons of marginal information of image appear, this phenomenon is that the algorithm principle of discrete Fourier transformation causes, and is inevitable, is not inconsistent with real motion blur image.

And, because frequency domain Fourier degeneration method is to utilize Fourier transform formula directly to obtain on the basis of spatial domain convolution degeneration method, therefore two kinds of methods is essential identical, that is to say that two marginal informations of image also can appear in spatial domain convolution degeneration method superimposed, with the phenomenon that real motion blur image is not inconsistent, this is spatial domain convolution degeneration method and the frequency domain Fourier common shortcoming of method of degenerating.

In addition, spatial domain convolution degeneration method and the frequency domain Fourier method of degenerating all is that image is carried out integral operation, and integral image is degenerated.Yet in actual application problem, more common but is that the target that occupies imaging system part visual field is kept in motion, and the situation that the part except target remains static, for this situation, spatial domain convolution degeneration method no matter, or the frequency domain Fourier method of degenerating all is to simulate the effect identical with true degraded image, and this also is spatial domain convolution degeneration method and the frequency domain Fourier common shortcoming of method of degenerating.

Summary of the invention

There is following shortcoming in art methods: 1) spatial domain convolution degeneration method is long operation time; 2) frequency domain Fourier transform degeneration method is directly perceived, difficult understands, and also needs image is carried out move operation after inverse Fourier transform; 3) the superimposed phenomenon of image both sides information that is not inconsistent with truth all appears in spatial domain method and frequency domain method; 4) spatial domain method and frequency domain method all can only be degenerated for entire image.For above-mentioned shortcoming, the present invention proposes the discrete degraded image building method of fold return motion target in a kind of static background; The method not only operation time short, and degenerative process is directly perceived, is convenient to understand, and need not image is adjusted again; Corresponding real scene can be simulated the motion blur image of the moving target formation that occupies the part visual field, and the superimposed phenomenon of image both sides information can not occurred simultaneously.

The object of the present invention is achieved like this:

The discrete degraded image building method of fold return motion target may further comprise the steps in a kind of static background:

A. be M in resolution ₁* N ₁The static background image in, having resolution is M ₂* N ₂(M ₂＜M ₁, N ₂＜N ₁) target image, and the row of described target image and row respectively row and the row of parallel static background image; All the time be no more than static background image edge take the target image edge as prerequisite, target image fold return motion process is divided into m section one-way movement process, i (i=1,2 ..., in the section one-way movement process, be M according to resolution m) ₂* N ₂Target image along its row or column direction n that moves _iThe distance of individual pixel constructs n _iNumber of sub images fig _{I, j}(j=1,2 ..., n _i);

B. the n that step a is obtained ₁+ n ₂+ ... + n _mNumber of sub images is weighted linear superposition according to following formula:

$\mathrm{fig}=\underset{i=1}{\overset{m}{\mathrm{\Σ}}}\underset{j=1}{\overset{n_i}{\mathrm{\Σ}}}{w}_{i,j}\·{\mathrm{fig}}_{i,j}$

In the formula, w _{I, j}Be weighting coefficient, fig is the discrete degraded image that constructs.

The discrete degraded image building method of fold return motion target in above-mentioned a kind of static background, among the described step a, the static background image is fig _Background, target image is fig _ObjectWherein, fig _BackgroundBe M ₁* N ₁Matrix, fig _ObjectBe M ₂* N ₂Matrix; The 1st number of sub images fig of corresponding i one-way movement process _{I, 1}, target image fig _ObjectThe 1st row the 1st row pixel b _1,1Be positioned at static background image fig _BackgroundThe capable yi1 row of xi1 pixel a _{Xi1, yi1}Position, then the 1st number of sub images fig of corresponding i one-way movement process _{I, 1}Be M1 * N ₁Matrix, and fig _{I, 1}(xi1:xi1+M ₂-1, yi1:yi1+N ₂-1)=fig _Object, fig _{I, 1}All the other elements and fig _BackgroundThe correspondence position element is identical; Wherein, xi1:xi1+M ₂-1 expression xi1 walks to xi1+M ₂-1 row, yi1:yi1+N ₂-1 expression yi1 is listed as to yi1+N ₂-1 row.

The discrete degraded image building method of fold return motion target in above-mentioned a kind of static background, among the described step a, the static background image is fig _Background, target image is fig _ObjectWherein, fig _BackgroundBe M ₁* N ₁Matrix, fig _ObjectBe M ₂* N ₂Matrix; The j of corresponding i one-way movement process (the number of sub images fig of 1≤j≤ni-1) _{I, j}, fig is arranged _{I, j}(xij:xij+M ₂-1, yij:yij+N ₂-1)=fig _Object, fig _{I, j}All the other elements and fig _BackgroundThe correspondence position element is identical; Wherein, xij:xij+M ₂-1 expression xij walks to xij+M ₂-1 row, yij:yij+N ₂-1 expression yij is listed as to yij+N ₂-1 row; If target image fig _ObjectMove upward, then the j+1 number of sub images fig of corresponding i one-way movement process _{I, j+1}Be M ₁* N ₁Matrix, and fig _{I, j+1}(xij-1:xij+M ₂-2, yij:yij+N ₂-1)=fig _Object, fig _{I, j+1}All the other elements and fig _BackgroundThe correspondence position element is identical; Wherein, xij-1:xij+M ₂-2 expression xij-1 walk to xij+M ₂-2 row, yij:yij+N ₂-1 expression yij is listed as to yij+N ₂-1 row.

The discrete degraded image building method of fold return motion target in above-mentioned a kind of static background, among the described step a, the static background image is fig _Background, target image is fig _ObjectWherein, fig _BackgroundBe M ₁* N ₁Matrix, fig _ObjectBe M ₂* N ₂Matrix; J (1≤the j≤n of corresponding i one-way movement process _i-1) number of sub images fig _{I, j}, fig is arranged _{I, j}(xij:xij+M ₂-1, yij:yij+N ₂-1)=fig _Object, fig _{I, j}All the other elements and fig _BackgroundThe correspondence position element is identical; Wherein, xij:xij+M ₂-1 expression xij walks to xij+M ₂-1 row, yij:yij+N ₂-1 expression yij is listed as to yij+N ₂-1 row; If target image fig _ObjectMove downward, then the j+1 number of sub images fig of corresponding i one-way movement process _{I, j+1}Be M ₁* N ₁Matrix, and fig _{I, j+1}(xij+1:xij+M ₂, yij:yij+N ₂-1)=fig _Object, fig _{I, j+1}All the other elements and fig _BackgroundThe correspondence position element is identical; Wherein, xij+1:xij+M ₂Represent that xij+1 walks to xij+M ₂OK, yij:yij+N ₂-1 expression yij is listed as to yij+N ₂-1 row.

The discrete degraded image building method of fold return motion target in above-mentioned a kind of static background, among the described step a, the static background image is fig _Background, target image is fig _ObjectWherein, fig _BackgroundBe M ₁* N ₁Matrix, fig _ObjectBe M ₂* N ₂Matrix; J (1≤the j≤n of corresponding i one-way movement process _i-1) number of sub images fig _{I, j}, fig is arranged _{I, j}(xij:xij+M ₂-1, yij:yij+N ₂-1)=fig _Object, fig _{I, j}All the other elements and fig _BackgroundThe correspondence position element is identical; Wherein, xij:xij+M ₂-1 expression xij walks to xij+M ₂-1 row, yij:yij+N ₂-1 expression yij is listed as to yij+N ₂-1 row; If target image fig _ObjectTo left movement, the j+1 number of sub images fig of corresponding i one-way movement process then _{I, j+1}Be M ₁* N ₁Matrix, and fig _{I, j+1}(xij:xij+M ₂-1, yij-1:yij+N ₂-2)=fig _Object, fig _{I, j+1}All the other elements and fig _BackgroundThe correspondence position element is identical; Wherein, xij:xij+M ₂-1 expression xij walks to xij+M ₂-1 row, yij-1:yij+N ₂-2 expression yij-1 are listed as to yij+N ₂-2 row.

The discrete degraded image building method of fold return motion target in above-mentioned a kind of static background, among the described step a, the static background image is fig _Background, target image is fig _ObjectWherein, fig _BackgroundBe M ₁* N ₁Matrix, fig _ObjectBe M ₂* N ₂Matrix; J (1≤the j≤n of corresponding i one-way movement process _i-1) number of sub images fig _{I, j}, fig is arranged _{I, j}(xij:xij+M ₂-1, yij:yij+N ₂-1)=fig _Object, fig _{I, j}All the other elements and fig _BackgroundThe correspondence position element is identical; Wherein, xij:xij+M ₂-1 expression xij walks to xij+M ₂-1 row, yij:yij+N ₂-1 expression yij is listed as to yij+N ₂-1 row; If target image fig _ObjectMove right, then the j+1 number of sub images fig of corresponding i one-way movement process _{I, j+1}Be M ₁* N ₁Matrix, and fig _{I, j+1}(xij:xij+M ₂-1, yij+1:yij+N ₂)=fig _Object, fig _{I, j+1}All the other elements and fig _BackgroundThe correspondence position element is identical; Wherein, xij:xij+M ₂-1 expression xij walks to xij+M ₂-1 row, yij+1:yij+N ₂Represent that yij+1 is listed as to yij+N ₂Row.

The discrete degraded image building method of fold return motion target in above-mentioned a kind of static background, when i 〉=2, the 1st number of sub images fig of described corresponding i section one-way movement process _{I, 1}N with corresponding i-1 section one-way movement process _I-1Number of sub images Equate.

The discrete degraded image building method of fold return motion target in above-mentioned a kind of static background, among the described step b, weighting coefficient w _{I, j}Ratio be expressed as:

$w_{\mathrm{1,1}}:w_{\mathrm{1,2}}:\·\·\·:w_{i,j}:\·\·\·w_{m,n_m-1}:w_{{m,n}_m}=\frac{1}{v_{\mathrm{1,1}}}:\frac{1}{v_{\mathrm{1,2}}}:\·\·\·:\frac{1}{v_{i,j}}:\·\·\·:\frac{1}{v_{{m,n}_m-1}}:\frac{1}{v_{{m,n}_m}}$

In the formula, v _{I, j}The j number of sub images fig of corresponding i section one-way movement process _{I, j}Corresponding target image movement velocity.

The discrete degraded image building method of fold return motion target in above-mentioned a kind of static background, carry out the gray-scale value adjustment with the resulting discrete degraded image fig of step b according to following formula:

fig_improve=k·fig

In the formula, k is for adjusting coefficient, and fig_improve is the discrete degraded image after adjusting.

The discrete degraded image building method of fold return motion target in above-mentioned a kind of static background, described adjustment coefficient

$k=1/\underset{i=1}{\overset{m}{\mathrm{\Σ}}}\underset{j=1}{\overset{n_i}{\mathrm{\Σ}}}{w}_{i,j}.$

The invention has the beneficial effects as follows:

1) discrete degraded image building method of the present invention is according to the n that moves altogether in the one-way movement of m section of the target image in the static background image ₁+ n ₂+ ... + n _mThe distance of individual pixel constructs n ₁+ n ₂+ ... + n _mNumber of sub images, and these subimages are weighted the linear, additive computing.Be comprised of four parts the operation time of the method, and first is n ₁+ n ₂+ ... + n _mThe structure time of number of sub images, second portion is n ₁+ n ₂+ ... + n _mThe computing time of individual weight, third part is n ₁+ n ₂+ ... + n _mThe numeral that individual respective weights and image multiply each other and the computing time of matrix multiple, the 4th part is to represent n ₁+ n ₂+ ... + n _mThe n of number of sub images ₁+ n ₂+ ... + n _mThe computing time of individual matrix addition.Because method of the present invention is avoided the quadruple circulation of spatial domain convolution degeneration method, and each step computing all is the simplest arithmetic, so the method has short beneficial effect operation time;

2) because the n that the present invention constructs ₁+ n ₂+ ... + n _mIn the number of sub images, every number of sub images all represents the actual position corresponding relation of corresponding constantly target image and static background image, with this n ₁+ n ₂+ ... + n _mIt is exactly instantaneous picture accumulative process in time in the true imaging process that number of sub images is weighted the linear, additive computing, therefore this method has the motion blur image that can simulate the moving target formation that occupies the part visual field, and the beneficial effect of the superimposed phenomenon of image both sides information can not occur.

3) since the whole degenerative process of discrete degraded image building method of the present invention in the spatial domain, finish, the direct corresponding imaging process of calculating process, thus the method to have degenerative process directly perceived, be convenient to understand, and need not beneficial effect that image is adjusted again.

Description of drawings

Fig. 1 is the static background image.

Fig. 2 is target image.

Fig. 3 is the 1st number of sub images of corresponding the 1st section one-way movement process.

Fig. 4 is discrete degraded image.

Embodiment

Below in conjunction with accompanying drawing the specific embodiment of the invention is done and to be described in further detail.

The resolution of static background image is 256 * 256, as shown in Figure 1; The resolution of target image is 64 * 64, as shown in Figure 2, in the present embodiment, the 1st row the 1st row pixel of target image is from the 150th row the 100th row location of pixels of static background image, with speed v along static background image line direction to the right uniform motion 10 pixels, again with speed 2v along static background image line direction left uniform motion 5 pixels.

The discrete degraded image building method of fold return motion target may further comprise the steps in a kind of static background:

A. in resolution is 256 * 256 static background image, having resolution is 64 * 64 target image, and the row of described target image and row respectively row and the row of parallel static background image; All the time be no more than static background image edge take the target image edge as prerequisite, target image fold return motion process is divided into 2 sections one-way movement processes, in the 1st section one-way movement process, be 64 * 64 target image according to resolution along the move distance of 10 pixels of its line direction, construct 10 number of sub images fig _{1, j}(j=1,2 ..., 10); In the 2nd section one-way movement process, be 64 * 64 target image according to resolution along the move distance of 5 pixels of its line direction, construct 5 number of sub images fig _{2, j}(j=1,2 ..., 5);

The 1st number of sub images fig of corresponding the 1st one-way movement process _1,1, target image fig _ObjectThe 1st row the 1st row pixel be positioned at static background image fig _BackgroundThe 150th row the 100th row location of pixels, the 1st number of sub images fig of corresponding the 1st one-way movement process then _1,1Be 256 * 256 matrix, and fig _1,1(150:213,100:163)=fig _Object, fig _1,1All the other elements and fig _BackgroundThe correspondence position element is identical; Wherein, 150:213 represents that the 150th walks to the 213rd row, and 100:163 represents that the 100th row are to the 163rd row;

The 1st number of sub images fig from the 1st one-way movement process _1,1Beginning, j (1≤j≤9) the number of sub images fig of corresponding the 1st one-way movement process _{1, j}, fig is arranged _{1, j}(x1j:x1j+63, y1j:y1j+63)=fig _Object, fig _{1, j}All the other elements and fig _BackgroundThe correspondence position element is identical; Wherein, it is capable that x1j:x1j+63 represents that x1j walks to x1j+63, and y1j:y1j+63 represents that y1j is listed as to y1j+63 and is listed as; The j+1 number of sub images fig of corresponding the 1st one-way movement process then _{1, j+1}Be 256 * 256 matrix, and fig _{1, j+1}(x1j:x1j+63, y1j+1:yij+64)=fig _Object, fig _{I, j+1}All the other elements and fig _BackgroundThe correspondence position element is identical; Wherein, it is capable that x1j:x1j+63 represents that x1j walks to x1j+63, and y1j+1:y1j+64 represents that y1j+1 is listed as to y1j+64 and is listed as;

The 1st number of sub images fig from the 2nd one-way movement process _2,1Beginning, j (1≤j≤4) the number of sub images fig of corresponding the 2nd one-way movement process _{2, j}, fig is arranged _{2, j}(x2j:x2j+63, y2j:y2j+63)=fig _Object, fig _{2, j}All the other elements and fig _BackgroundThe correspondence position element is identical; Wherein, it is capable that x2j:x2j+63 represents that x2j walks to x2j+63, and y2j:y2j+63 represents that y2j is listed as to y2j+63 and is listed as; The j+1 number of sub images fig of corresponding the 2nd one-way movement process then _{2, j+1}Be 256 * 256 matrix, and fig _{2, j+1}(x2j:x2j+63, y2j-1:y2j+62)=fig _Object, fig _{2, j+1}All the other elements and fig _BackgroundThe correspondence position element is identical; Wherein, it is capable that x2j:x2j+63 represents that x2j walks to x2j+63, and y2j-1:y2j+62 represents that y2j-1 is listed as to y2j+62 and is listed as;

And, the 1st number of sub images fig of described corresponding the 2nd section one-way movement process _2,1The 10th number of sub images fig with corresponding the 1st section one-way movement process _1,10Equate.

B. the 10+5=15 number of sub images that step a is obtained is weighted linear superposition according to following formula:

$\mathrm{fig}=\underset{i=1}{\overset{2}{\mathrm{\Σ}}}\underset{j=1}{\overset{n_i}{\mathrm{\Σ}}}{w}_{i,j}\·{\mathrm{fig}}_{i,j}$

In the formula, w _{I, j}Be weighting coefficient, fig is the discrete degraded image that constructs, and has: n ₁=10, n ₂=5.

According to: v _1,1=v _1,2=...=v _1,10=v, and v _2,1=v _2,2=...=v _2,5=2v can obtain weighting coefficient w _{I, j}Ratio be expressed as:

$w_{\mathrm{1,1}}:\·\·\·:w_{\mathrm{1,10}}:w_{\mathrm{2,1}}:\·\·\·:w_{\mathrm{2,5}}=\frac{1}{v_{\mathrm{1,1}}}:\·\·\·:\frac{1}{v_{\mathrm{1,10}}}:\frac{1}{v_{\mathrm{2,1}}}:\·\·\·:\frac{1}{v_{\mathrm{2,5}}}=1:\·\·\·:1:\frac{1}{2}:\·\·\·:\frac{1}{2}$

In the formula, v _{I, j}The j number of sub images fig of corresponding i section one-way movement process _{I, j}Corresponding target image movement velocity, and have: v _{I, j}≠ 0.

The discrete degraded image building method of fold return motion target in above-mentioned a kind of static background, carry out the gray-scale value adjustment with the resulting discrete degraded image fig of step b according to following formula:

fig_improve=k·fig

In the formula, k is for adjusting coefficient, and fig_improve is the discrete degraded image after adjusting.Described adjustment coefficient

$k=1/\underset{i=1}{\overset{m}{\mathrm{\Σ}}}\underset{j=1}{\overset{n_i}{\mathrm{\Σ}}}{w}_{i,j}$

The discrete degraded image that obtains after the adjustment as shown in Figure 4.

Claims (10)

1. the discrete degraded image building method of fold return motion target in the static background is characterized in that said method comprising the steps of:

A. be M in resolution ₁* N ₁The static background image in, having resolution is M ₂* N ₂(M ₂＜M ₁, N ₂＜N ₁) target image, and the row of described target image and row respectively row and the row of parallel static background image; All the time be no more than static background image edge take the target image edge as prerequisite, target image fold return motion process is divided into m section one-way movement process, i (i=1,2 ..., in the section one-way movement process, be M according to resolution m) ₂* N ₂Target image along its row or column direction n that moves _iThe distance of individual pixel constructs n _iNumber of sub images fig _{I, j}(j=1,2 ..., n _i);

B. the n that step a is obtained ₁+ n ₂+ ... + n _mNumber of sub images is weighted linear superposition according to following formula:

$\mathrm{fig}=\underset{i=1}{\overset{2}{\mathrm{\Σ}}}\underset{j=1}{\overset{n_i}{\mathrm{\Σ}}}{w}_{i,j}\·{\mathrm{fig}}_{i,j}$

In the formula, w _{I, j}Be weighting coefficient, fig is the discrete degraded image that constructs.

2. according to the discrete degraded image building method of fold return motion target in claim 1 described a kind of static background, it is characterized in that among the described step a, the static background image is fig _Background, target image is fig _ObjectWherein, fig _BackgroundBe M ₁* N ₁Matrix, fig _ObjectBe M ₂* N ₂Matrix; The 1st number of sub images fig of corresponding i one-way movement process _{I, 1}, target image fig _ObjectThe 1st row the 1st row pixel b _1,1Be positioned at static background image fig _BackgroundThe capable yi1 row of xi1 pixel a _{Xi1, yi1}Position, then the 1st number of sub images fig of corresponding i one-way movement process _{I, 1}Be M ₁* N ₁Matrix, and fig _{I, 1}(xi1:xi1+M ₂-1, yi1:yi1+N ₂-1)=fig _Object, fig _{I, 1}All the other elements and fig _BackgroundThe correspondence position element is identical; Wherein, xi1:xi1+M ₂-1 expression xi1 walks to xi1+M ₂-1 row, yi1:yi1+N ₂-1 expression yi1 is listed as to yi1+N ₂-1 row.

3. according to the discrete degraded image building method of fold return motion target in claim 1 described a kind of static background, it is characterized in that among the described step a, the static background image is fig _Background, target image is fig _ObjectWherein, fig _BackgroundBe M ₁* N ₁Matrix, fig _ObjectBe M ₂* N ₂Matrix; J (1≤the j≤n of corresponding i one-way movement process _i-1) number of sub images fig _{I, j}, fig is arranged _{I, j}(xij:xij+M ₂-1, yij:yij+N ₂-1)=fig _Object, fig _{I, j}All the other elements and fig _BackgroundThe correspondence position element is identical; Wherein, xij:xij+M ₂-1 expression xij walks to xij+M ₂-1 row, yij:yij+N ₂-1 expression yij is listed as to yij+N ₂-1 row; If target image fig _ObjectMove upward, then the j+1 number of sub images fig of corresponding i one-way movement process _{I, j+1}Be M ₁* N ₁Matrix, and fig _{I, j+1}(xij-1:xij+M ₂-2, yij:yij+N ₂-1)=fig _Object, fig _{I, j+1}All the other elements and fig _BackgroundThe correspondence position element is identical; Wherein, xij-1:xij+M ₂-2 expression xij-1 walk to xij+M ₂-2 row, yij:yij+N ₂-1 expression yij is listed as to yij+N ₂-1 row.

4. according to the discrete degraded image building method of fold return motion target in claim 1 described a kind of static background, it is characterized in that among the described step a, the static background image is fig _Background, target image is fig _ObjectWherein, fig _BackgroundBe M ₁* N ₁Matrix, fig _ObjectBe M ₂* N ₂Matrix; J (1≤the j≤n of corresponding i one-way movement process _i-1) number of sub images fig _{I, j}, fig is arranged _{I, j}(xij:xij+M ₂-1, yij:yij+N ₂-1)=fig _Object, fig _{I, j}All the other elements and fig _BackgroundThe correspondence position element is identical; Wherein, xij:xij+M ₂-1 expression xij walks to xij+M ₂-1 row, yij:yij+N ₂-1 expression yij is listed as to yij+N ₂-1 row; If target image fig _ObjectMove downward, then the j+1 number of sub images fig of corresponding i one-way movement process _{I, j+1}Be M ₁* N ₁Matrix, and fig _{I, j+1}(xij+1:xij+M ₂, yij:yij+N ₂-1)=fig _Object, fig _{I, j+1}All the other elements and fig _BackgroundThe correspondence position element is identical; Wherein, xij+1:xij+M ₂Represent that xij+1 walks to xij+M ₂OK, yij:yij+N ₂-1 expression yij is listed as to yij+N ₂-1 row.

5. according to the discrete degraded image building method of fold return motion target in claim 1 described a kind of static background, it is characterized in that among the described step a, the static background image is fig _Background, target image is fig _ObjectWherein, fig _BackgroundBe M ₁* N ₁Matrix, fig _ObjectBe M ₂* N ₂Matrix; J (1≤the j≤n of corresponding i one-way movement process _i-1) number of sub images fig _{I, j}, fig is arranged _{I, j}(xij:xij+M ₂-1, yij:yij+N ₂-1)=fig _Object, fig _{I, j}All the other elements and fig _BackgroundThe correspondence position element is identical; Wherein, xij:xij+M ₂-1 expression xij walks to xij+M ₂-1 row, yij:yij+N ₂-1 expression yij is listed as to yij+N ₂-1 row; If target image fig _ObjectTo left movement, the j+1 number of sub images fig of corresponding i one-way movement process then _{I, j+1}Be M ₁* N ₁Matrix, and fig _{I, j+1}(xij:xij+M ₂-1, yij-1:yij+N ₂-2)=fig _Object, fig _{I, j+1}All the other elements and fig _BackgroundThe correspondence position element is identical; Wherein, xij:xij+M ₂-1 expression xij walks to xij+M ₂-1 row, yij-1:yij+N ₂-2 expression yij-1 are listed as to yij+N ₂-2 row.

6. according to the discrete degraded image building method of fold return motion target in claim 1 described a kind of static background, it is characterized in that among the described step a, the static background image is fig _Background, target image is fig _ObjectWherein, fig _BackgroundBe M ₁* N ₁Matrix, fig _ObjectBe M ₂* N ₂Matrix; J (1≤the j≤n of corresponding i one-way movement process _i-1) number of sub images fig _{I, j}, fig is arranged _{I, j}(xij:xij+M ₂-1, yij:yij+N ₂-1)=fig _Object, fig _{I, j}All the other elements and fig _BackgroundThe correspondence position element is identical; Wherein, xij:xij+M ₂-1 expression xij walks to xij+M ₂-1 row, yij:yij+N ₂-1 expression yij is listed as to yij+N ₂-1 row; If target image fig _ObjectMove right, then the j+1 number of sub images fig of corresponding i one-way movement process _{I, j+1}Be M ₁* N ₁Matrix, and fig _{I, j+1}(xij:xij+M ₂-1, yij+1:yij+N ₂)=fig _Object, fig _{I, j+1}All the other elements and fig _BackgroundThe correspondence position element is identical; Wherein, xij:xij+M ₂-1 expression xij walks to xij+M ₂-1 row, yij+1:yij+N ₂Represent that yij+1 is listed as to yij+N ₂Row.

7. according to the discrete degraded image building method of fold return motion target in claim 1 described a kind of static background, it is characterized in that when i 〉=2 the 1st number of sub images fig of described corresponding i section one-way movement process _{I, 1}N with corresponding i-1 section one-way movement process _I-1Number of sub images

Equate.

8. the discrete degraded image building method of fold return motion target in a kind of static background according to claim 1 is characterized in that among the described step b weighting coefficient w _{I, j}Ratio be expressed as:

$w_{\mathrm{1,1}}:w_{\mathrm{1,2}}:\·\·\·:w_{i,j}:\·\·\·w_{m,n_m-1}:w_{{m,n}_m}=\frac{1}{v_{\mathrm{1,1}}}:\frac{1}{v_{\mathrm{1,2}}}:\·\·\·:\frac{1}{v_{i,j}}:\·\·\·:\frac{1}{v_{{m,n}_m-1}}:\frac{1}{v_{{m,n}_m}}$

In the formula, v _{I, j}The j number of sub images fig of corresponding i section one-way movement process _{I, j}Corresponding target image movement velocity.

9. the discrete degraded image building method of fold return motion target in a kind of static background according to claim 1 is characterized in that the resulting discrete degraded image fig of step b is carried out the gray-scale value adjustment according to following formula:

fig_improve=k·fig

In the formula, k is for adjusting coefficient, and fig_improve is the discrete degraded image after adjusting.

10. according to the discrete degraded image building method of fold return motion target in claim 9 described a kind of static background, it is characterized in that described adjustment coefficient

Images