CN101587589A - Inkjet printing quality fuzzy evaluation method - Google Patents

Abstract

A kind of new inkjet printing quality fuzzy evaluation method is characterized in that: utilize acquisition technology, realize the conversion of inkjet print image and digital picture, based on the color evaluation of S-CIELAB color space simulation human visual system to inkjet printing; By the blotting amount of mass distribution theoretical modeling Alhue paper, estimate the definition of spray ink printing quality by measure theory to single pixel.Based on color and definition quality, in conjunction with human-eye visual characteristic, set up the inkjet printing quality fuzzy evaluation algorithms, and this algorithm is composed with numerical value Image-QA _IJPTest data of the present invention is few, speed is fast, overcome effectively the print image quality subjective evaluation method consuming time, expensive and the traditional objective evaluation method is single or with drawbacks such as the real image quality does not match, for the analysis of inkjet printing quality provides a kind of effective means.And required support equipment is simple, and cost is low, and is easy to use.

Description

Inkjet printing quality fuzzy evaluation method

Technical field

The present invention relates to a kind of inkjet printing quality fuzzy evaluation method,, be applicable to that printing and field of papermaking detect the method for inkjet printing quality with mathematics metering formula objective quantification observer inkjet print image quality evaluation result in the heart.

Background technology

Color inkjet printer relates to growing field, the corresponding increase of ink-jet printer and Alhue paper consumption, make people more and more higher to the requirement of print image quality, therefore, must the quality information of image be detected and analyze, thus control and improve paper, printing ink in conjunction with effect.In recent years, scientist is obtaining many progress aspect the print image quality evaluation.

Density value in the conventional measurement techniques characterizes color can not reach the eye evaluation result, is difficult for and printing customer communication color information.At present, industrial test printing image color and gradation reproduction method for quality are that the CIELAB aberration is measured.Based on match colors and discrimination test, the CIELAB aberration that calculates is under fixing visual adaptation condition, is used to test the difference that paper is printed large tracts of land homogeneous area color and level.The CIELAB colour difference formula partly reflects the adaptive influence of vision system, but can not this phenomenon of good explanation.The development and application of each colour difference formula only be applicable to it at that time developing period fixed certain environmental conditions, judge that the factor of aberration has a lot of aspects and influence human eye, also comprise light source, illumination, test color to relative distance, test color to size and presentation content structure etc.

In numerous inkjet printing quality analytical approachs, the computer image analysis method is a kind of of development abroad comparative maturity, this analytic system is with good man computer interface, measure accurate and visual performance rapidly, visual, the microcosmicization of paper inkjet printing quality and the quality analysis of paper own are attained a new height again.Be subject to processing the restriction of speed and instrument automatic control aspect, major part is confined to the applied analysis of laboratory scope now.Two-dimensional CCD camera can be realized the collection of image preferably.But ccd video camera and framework can only be taken the regional area of high graphics, and the ccd video camera light source can not be consistent on time and homogeneity fully, and the test result of some print performance is inaccurate.The scanner cost is low, can carry out the image scanning of folk prescription to, large area region, avoids problems such as the repeated acquisition of pixel and the restriction of light source homogeneity.But the image acquisition effect to tiny characteristic is relatively poor, and as ink dot or tiny lines, so scanner is unsuitable for micro-analysis.Also there are defectives such as the incomplete and measurement index disunity of price height, performance at present in paper inkjet printing quality analytic system.

The ISO13660 standard only defines large tracts of land and fills quality standard with lines on the spot, but the performance index of most imaging systems and the visual importance of measuring difference do not obtain embodying in standard.Therefore the ISO13660 standard can not well be used for the evaluation of Alhue paper print quality.

People can estimate inkjet printing quality respectively on microcosmic and macro-scale by computer image analysis method and digital image analysis method, but also have very people's gap with picture quality and observer's image quality evaluation result in the heart of mathematics metering formula tolerance.Combining of objective measurement and subjective determination how, physical index combines with psychological indicator, with the evaluation result of mathematics metering formula objective quantification observer image quality in the heart, is the matter of utmost importance that the inkjet printing quality evaluation will solve.

Summary of the invention

The present invention combines objective measurement and subjective determination, physical index combines with psychological indicator, mathematical theory and S-CIELAB color space methods such as employing is estimated, mass distribution and Triangular Fuzzy Number, make up the inkjet printing quality fuzzy evaluation algorithms, with mathematics metering formula objective quantification inkjet printing quality.These test of heuristics data are few, speed is fast, overcome effectively the print image quality subjective evaluation method consuming time, expensive and the traditional objective evaluation method is single or with drawbacks such as the real image quality does not match, for the analysis of inkjet printing quality provides a kind of effective means.And required support equipment is simple, and cost is low, and is easy to use.

Technical solution of the present invention is:

A kind of new inkjet printing quality fuzzy evaluation method is characterized in that: utilize acquisition technology, realize the conversion of inkjet print image and digital picture, based on the color evaluation of S-CIELAB color space simulation human visual system to inkjet printing; By the blotting amount of mass distribution theoretical modeling Alhue paper, estimate the definition of spray ink printing quality by measure theory to single pixel.Based on color and definition quality, in conjunction with human-eye visual characteristic, set up the inkjet printing quality fuzzy evaluation algorithms, and this algorithm is composed with numerical value Image-QA _IJP

To described inkjet printing quality fuzzy evaluation algorithms, determine by color rendering quality and two quality factors of definition quality.

To described color rendering evaluation procedure, adopt six quality factors to determine, promptly shadow, to a high-profile, neutral ash, the colour of skin, heavy shade and memory colour; To described sharpness evaluation procedure, adopt four quality factors to determine, be respectively black, cyan, magenta and yellow pixel point.

To described inkjet printing quality fuzzy evaluation of estimate, obtain the vision weight of inkjet printing quality factor by the Triangular Fuzzy Number method.

Vision weight to described color rendering quality factor and sharpness quality factor can be calculated by analytical hierarchy process.

Above-mentioned color rendering quality factor comprise shadow, to a high-profile, the color rendering effect of neutral ash, the colour of skin, heavy shade and memory colour image, its weight is respectively w ₁=0.1515, w ₂=0.1554, w ₃=0.1731, w ₄=0.1724, w ₅=0.1666, w ₆=0.1810.Color evaluating Δ E _ICComputing formula is as follows:

ΔE _IC＝0.1515·ΔE(L)+0.1554·ΔE(H)+0.1731·ΔE(G)

+0.1724·ΔE(S)+0.16666·ΔE(C)+0.181·ΔE(M) (1)

In the formula Δ E (L), Δ E (H), Δ E (G), Δ E (S), Δ E (C) and Δ E (M) represent respectively print shadow, to a high-profile, the heterochromia value between neutral ash, the colour of skin, heavy shade, memory colour image and the original image.

Above-mentioned sharpness quality factor is made up of four parts, is the print quality of black, cyan, magenta and yellow pixel point successively.Four quality factors are to the influence degree of definition quality, and promptly shared weight is followed successively by w ₁=0.2701, w ₂=0.2421, w ₃=0.2414, w ₄=0.2464.Sharpness evaluating S _PMSDComputing formula is as follows:

S _PMSD＝0.2701·P _K+0.2421·P _C+0.2414·P _M+0.2464·P _Y (2)

P in the formula _K, P _C, P _MAnd P _YBe black, cyan, magenta and yellow pixel point PMSD value successively.

Above-mentioned inkjet printing quality fuzzy evaluating is described by color rendering and definition quality, and by the Triangular Fuzzy Number method, the weight that can obtain the color and the sharpness factor is respectively w ₁=0.5234, w ₂=0.4766.Inkjet printing quality fuzzy evaluating Image-QA _IJPComputing formula is as follows:

Image-QA _IJP＝0.5234·k ₁·ΔE _IC+0.4766·k ₂·S _PMSD (3)

K in the formula ₁And k ₂Be respectively normalized factor.

Accompanying drawing is briefly described

Fig. 1 is the process flow diagram of inkjet printing quality fuzzy evaluation method of the present invention

The quality image that Fig. 2 estimates for definition quality

Embodiment:

1. inkjet printing color rendering evaluation method

As shown in Figure 1, color scanner will be printed six kinds of quality factor images (shadow, to a high-profile, neutral ash, the colour of skin, heavy shade and memory colour) and be converted to digital picture, and the RGB image transitions that scans is become the XYZ pattern.Because the spectrum sensitivity inductor and the XYZ color matching function of scanner are non-linear conversion, so the RGB image that scanner obtains can not convert the XYZ image directly and accurately to.In order to address this problem, in this experiment, adopt LaserSoft ImagingTM SilverFast software to generate the scanner characteristics file, scanner is calibrated.

The XYZ image and the original image of scanning are transformed to three opposite color passages, i.e. a luminance channel and two chrominance channels.Be converted to three-channel image and can carry out spatial filtering, with the space mixed effect of simulating human vision system.The wave filter that is adopted is near human visual system's contrast influence function.Both can in spatial domain, carry out convolution, also can carry out filtering by the product of frequency field.Spatial filtering carries out the two-dimensional convolution computing respectively to each opposite color component, and convolution kernel is by the roomage response characteristic decision of human eye to image.At last, tristimulus values is transformed in the CIE LAB opposite color coordinate system, transformation for mula is as follows:

$L^{*}=\left\{\begin{array}{cc}116{(Y/Y_n)}^{1/3}-16& (Y/Y_n>0.008856)\\ 903.3(Y/Y_n)& (Y/Y_n\≤0.008856)\end{array}\right.---\left(4\right)$

$\left\{\begin{array}{c}{a}^{*}=500[f(X/X_n)-f(Y/Y_n)]\\ b^{*}=200[f(Y/Y_n)-f(Z/Z_n)]\end{array}\right.---\left(5\right)$

$f\left(q\right)=\left\{\begin{array}{cc}{q}^{1/3}& (q>0.008856)\\ 7.787q+16/116& (q\≤0.008856)\end{array}\right.---\left(6\right)$

Formula (4)～(6) can make the XYZ pixel value of original copy and test pattern filtering be transformed into CIE LAB space.X _n, Y _n, Z _nIt is the tristimulus values of CIE standard illuminants; X, Y, Z are the tinctorial pattern tristimulus values.Q is respectively X/X _n, Y/Y _n, Z/Z _nX _n, Y _n, Z _nBe the tristimulus values of display white point, can be by the device characteristics acquisition.L ^*, a ^*, b ^*Represent the lightness axis of the CIELAB color space, red-greenness axle and Huang-blue chrominance axis respectively.

In case obtain the CIE LAB coordinate of filtering image, six kinds of color rendering quality images carried out the point-to-point aberration of individual element calculate.Adopt up-to-date colour difference formula CIE DE2000 to be used for the calculating of this value of chromatism at present.By the vision weight of color rendering quality factor, can try to achieve color evaluating value Δ E _IC

2. definition of spray ink printing evaluation method

As shown in Figure 1, adopt 3-D videomicroscopy system air exercise printing vegetarian refreshments image to carry out digital image acquisition.

For R ⁿIn each Borel collection A, suppose that μ (A) is collection the estimating of A.Claim R ⁿThe bounded Borel subclass A μ that go up to satisfy 0＜μ (A)＜∞ be mass distribution, can think the quality of μ (A) for collection A.

F is defined in R ²Go up and satisfy following condition:

(1) f is at R ²Last bounded;

(2) f is at R ²On definitely can amass, promptly Convergence;

(3) exist limited not hand over R ²Borel subclass A ₁, A ₂..., A _nSatisfy A ₁∪ A ₂∪ ... ∪ A _n=R ²And f|A _i∈ C (A _i), i=1,2 ..., n.

For f ∈ PC (R ²), if $\∀A\∈B_2$ (A is two-dimentional Borel collection), integration ∫ _A| f|d σ exists, definable μ _fFor:

μ _f(A)＝∫ _A|f|σ (7)

μ then _fBe a mass distribution, f wherein (x) can think " density " under certain meaning.

In the present invention, for f ∈ PC (R ²), the definition mass distribution || μ _f|| norm be

$\left|\right|{\mathrm{\μ}}_f\left|\right|=\underset{A\⋐R^2}{\sup }{\mathrm{\μ}}_f\left(A\right).---\left(8\right)$

For f, g ∈ PC (R ²), can defined function f and g apart from d (f g) is:

d(f，g)＝||μ _f-g|| (9)

The pixel image available functions f (x) that one width of cloth is printed is expressed as follows,

If paper surface has China ink, then f (x) is 1, otherwise is 0.Wherein, B _LBe the intensity level of RGB, k is a normalizing factor, in this experiment, gets k=1/255.

For ink-jet printer, the ideal form of printing minimum pixel point is circular.Therefore, setting the actual ink quality that is adsorbed on paper surface is μ (A), the paper surface desirable pixel image that can reproduce so, and its respective function g (x) is expressed as follows,

$g\left(x\right)=\left\{\begin{array}{cc}1,& x\∈D;\\ 0,& x\∉D.\end{array}\right.---\left(11\right)$

If x ∈ is D, then g (x) is 1, otherwise is 0.Wherein, D is to be the circle at center with (0,0), and satisfies μ _f(R ²)=μ _g(R ²).

Pixel image of printing and desirable pixel image, difference respective function f (x) and g (x), obvious f, g is R ²On the bounded sectional-continuous function.The present invention proposes one and new measures the difference value that PMSD (Pixel Micro-structure Difference) describes two width of cloth image print sharpness, promptly utilizes the given f of formula (9), and the distance of g is as the size of its difference:

PMSD＝d(f，g)＝||μ _f-g|| (12)

The sharpness quality factor is made up of four parts, is the print quality of black, cyan, magenta and yellow pixel point successively.By the vision weight of sharpness quality factor, can try to achieve sharpness evaluating value S _PMSD

3. inkjet printing quality fuzzy evaluation method

Determine the vision weight of color rendering and sharpness quality factor can obtain inkjet printing quality fuzzy evaluating value Image-QA by the Triangular Fuzzy Number method _IJP

The present invention can be applicable to the quality testing of various inkjet print images.

Claims (5)

1. a new inkjet printing quality fuzzy evaluation method is characterized in that: utilize acquisition technology, realize the conversion of inkjet print image and digital picture, simulate the human visual system based on the S-CIELAB color space and obtain inkjet printing color rendering quality; By the blotting amount of mass distribution theoretical modeling Alhue paper, estimate the definition of spray ink printing quality by measure theory to single pixel.Based on color and definition quality, in conjunction with human-eye visual characteristic, set up the inkjet printing quality fuzzy evaluation algorithms, and this algorithm is composed with numerical value Image-QA _IJP

2. according to the inkjet printing quality fuzzy evaluation method described in the claim 1, it is characterized in that:, determine by color rendering quality and two quality factors of definition quality to described inkjet printing quality fuzzy evaluation algorithms.

3. according to the inkjet printing quality fuzzy evaluation method described in the claim 1, it is characterized in that: to described color rendering evaluation procedure, adopt six quality factors to determine, promptly shadow, to a high-profile, neutral ash, the colour of skin, heavy shade and memory colour; To described sharpness evaluation procedure, adopt four quality factors to determine, be respectively black, cyan, magenta and yellow pixel point.

4. according to the inkjet printing quality fuzzy evaluation method described in the claim 2, it is characterized in that:, obtain the vision weight of inkjet printing quality factor by the Triangular Fuzzy Number method described inkjet printing quality fuzzy evaluation of estimate.

5. according to the inkjet printing quality fuzzy evaluation method described in the claim 3, it is characterized in that: the vision weight to described color rendering quality factor and sharpness quality factor can be calculated by analytical hierarchy process.

Images