WO2008033090A1

WO2008033090A1 - A method of predicting the printability of a paper, or paper board

Info

Publication number: WO2008033090A1
Application number: PCT/SE2007/050632
Authority: WO
Inventors: Mats Fredlund; Vesna Rakic; Isabel KNÖÖS
Original assignee: Stora Enso Ab
Priority date: 2006-09-11
Filing date: 2007-09-07
Publication date: 2008-03-20
Also published as: SE0601860L; SE530070C2

Abstract

The invention refers to a method of prediction the printability of an unprinted paper, paper board or board surface. The method comprises the steps of acquiring a topographic image of the surface; identifying depressions regions in the topographical image having a depression exceeding a predefined depression threshold; establishing a relationship between the relative amount of identified depressions regions and the areas of said regions calculating a value on the basis of said relationship; and using the value as measurement of the prediction of the printability of the paper, paper board or board surface.

Description

A method of predicting the printability of a paper or paper board .

Technical field

The invention refers to a method of predicting the print quality with regard to uncovered areas of unprinted surfaces .

The invention is especially advantageous, but not limited to, the prediction of the print quality of paper and board surfaces .

Background art

Local unprinted areas of a printed substrate, which in the following will be referred to as uncovered areas, occur when incomplete ink transfer to the substrate takes place during the printing process. The occurrence of uncovered areas effects the print quality negatively. For the board and paper producer, the occurrence of uncovered areas leads to complaints and reclaims from the customers.

The most common way for measuring print quality is to evaluate the printed surface. This means that the producer of the paper or board has to make regular tests by printing test strips at a laboratory and then measuring the print quality of the test strips.

One such method is described in the paper "Topographic Distribution of Uncovered Areas (UCA) in Full Tone Flexographic Prints", Gustavo Gil Barros et al, (56th annual Technical conference of the Technical Association on the Graphic Arts, TAGA 2004), where printed samples was analysed using a grazing light topography characterizing method known as OptiTopo. This grazing light topography characterizing method is further described in WO 00/45125 Al. In said paper "Topographic Distribution of Uncovered Areas (UCA) in Full Tone Flexographic Prints" it was concluded that surface depressions are the driving cause of uncovered areas and that a spectral analysis of the topography can provide an accurate estimator of the magnitude of uncovered areas.

However, the method disclosed in the above-mentioned paper does not involve the prediction of uncovered areas on unprinted samples. Furthermore, the method is only based on the depth and the total area of the defects. It does not consider the area of each defect or the distribution of the defects on the surface.

Also, a major drawback of the above-described methods is that the test strips need to be printed before the analysis of uncovered areas can be made. This is very time consuming work .

Thus, there is a need for a method to predict the print quality of an unprinted surface.

The object of the present invention is to provide such a method. Another object of the present invention is to provide a method to predict the print quality of a surface with regard to uncovered areas .

These, as well as other objects, are achieved by the present invention.

Disclosure of the invention

It has now been surprisingly shown that a value characterizing the printability of an unprinted surface, e.g. a paper, paper board or board surface, can be calculated on the basis of the areas of individual defects on the unprinted surface. The method according to the invention is characterized by the steps of:

- acquiring a topographic image of the surface;

identifying depressions regions in the topographical image having a depression exceeding a predefined depression threshold;

- establishing a relationship between the relative amount of identified depressions regions and the areas of said regions

- calculating a value on the basis of said relationship; and

- using said value as measurement for the prediction of the printability of the unprinted paper, paper board or board surface. The method according to the invention makes it possible to predict the printability of an unprinted surface. This facilitates the control of the print quality, since it eliminates the time consuming work of removing test strips from the production, printing said test strips at a laboratory and then measuring the print quality of the test strips .

Brief description of the drawings

Fig. 1 shows a topographical image of an unprinted board sample .

Fig. 2 shows the image of Fig. 1 after filtering and binarization.

Fig. 3 shows in a diagram the relative amount of white spots in the image of Fig. 2 plotted against the area size and also a Weibull distribution adapted to the relative amount of white spots.

Fig. 4 shows in a diagram Weibull mean values for seven board samples plotted against the visually ranking scores of the samples.

Fig. 5-7 show surface defects in three different samples. Detailed description of the invention

In the following the invention will be described in more detail with reference to the above-mentioned drawings.

The first step of the method according to the invention is to acquire an image of the unprinted surface disclosing the topography of the surface. The topographic image is preferably acquired using the acquisition method described in the above-mentioned document WO 00/045125 Al. As such, this method is well described in WO 00/045125 Al and it will only be described schematically in the following.

The surface of the sample is illuminated by incident light. The intensity of the light reflected from the surface is recorded as a first image. Then, using a different angle of illumination, the intensity of the light reflected from the surface is recorded as a second image. Only the diffusely reflected light is recorded. The difference between the recorded intensities of the first and the second images is determined to obtain a representation that emphasizes variations in gradient of the surface. This representation is further processed by signal-adapted integration to a topographic description, that is, a height function of the surface. It is, however, understood that other methods for acquiring the topographic image, e.g. by means of a stylus or a laser beam, may alternatively be used.

Following the acquisition of the topographic image, depression regions of the image are identified. Depression regions represent defect regions of the sample where uncovered areas may appear when the sample is printed. Preferably the identification is done by filtering and binarizing the topographic image in such a way that regions having a depression exceeding a predefined depression threshold are identified. Preferably, the threshold is set so that depression regions under a certain percentile, e.g. the 10% percentile, are identified.

After the step of identifying depression regions, there are a number of identified regions having different sizes, i.e. areas, in the filtered and binarized image. Next, a relationship between the relative amount of identified depressions regions and region size, i.e. the relative amount of identified regions as a function of region size, is established. Preferably, this is done by calculating the area of each identified region and also the total area A of all of the identified regions, i.e. the sum of the area of all identified regions. Thereafter, the identified regions are arranged in groups according to size, i.e. each identified region is placed in a group 1, 2, 3, ..., i where the identified region has an area within a predefined area interval. Preferably the area interval of each group increases as the area values increase. For example, the area interval can increase with a factor Λ/2 as the area value increases. The total area a_±, a∑, a3, ..., a_± of the identified regions in each group is calculated. Here a_± denotes the total area of the identified regions in group i. Then, the total area of the identified regions in each group is divided by the total area of all identified regions so that the normalized values ai/A, a2/A, a3/A, ..., a_x/A are obtained. These normalized values represent the relative amount of identified regions in each group.

Alternatively, the proportion of the surface area that is covered by the identified regions is calculated by dividing the total area A of all of the identified regions with the total area T of the surface. Also, within each group, the proportion of the surface area that is covered by the identified regions belonging to that group is calculated by dividing the total area a_± of the identified regions in each group with the total area T of the surface. The relative amount of identified regions in each group is then obtained by dividing the proportion of the surface area that is covered by the identified regions within each group with the proportion of the surface area that is covered by identified regions such that the value Ia₁Zi) / (A/T) = ai/A is obtained for each group. This will yield the same result as the first method. These values then represent the relative amount of identified regions in each group.

Next, a mathematical function is adapted to the relationship between the relative amount of identified depressions regions and region size. In other words, the relative amount of identified regions as a function of region size is approximated by a mathematical function. Preferably, the mathematical function is a statistical distribution, most preferably the Weibull distribution. From the mathematical function, a value is calculated. This value is then used as a measurement of the prediction of the printability of the surface. Preferably, the value is a statistical value. The value can, for example, be the mean, the variance, a parameter, a constant or an independent variable of the mathematical function. For example, from the adapted Weibull distribution, either one of the mean or the variance of the distribution is calculated. It has been found that this value can be used to predict the occurrence of uncovered areas in a printed sample. Preferably, the mean is used as it gives a slightly better correlation against a visual ranking of the printing quality.

It is understood that other statistical distributions than the Weibull distribution can be used, e.g. Lognormal distribution, Gamma distribution, Chisquare distribution, even though these distributions gave less reliable results than the Weibull distribution in a screening study Alternatively, a mathematical function, e.g. an exponential distribution or a power function, can be used.

In the following, a test performed to predict the printability of seven unprinted board samples using the method according to the invention will be described.

For each of the seven unprinted board samples a topographical image was provided by the method described in WO 00/045125 Al. The topographical image from sample 1 is shown in Fig. 1. In the image surface depressions appear as dark regions. The image was transformed into a binary image so that depression regions under the 10% percentile were visualized as white spots. Fig. 2 shows the image of Fig. 1 after filtering and binarization . From the binarized picture, the area of each white spot was obtained using conventional image analysing techniques. The white spots were grouped according to size, i.e. each spot was placed in a group where the spots had areas within a predefined area interval. Table 1 shows the predefined area intervals that were used.

Table 1

It is to be noted that for groups i = 1, the lower and upper interval limit was chosen to be 0 and 0,08 respectively, and for groups i = 2, 3, 4, ... the lower and upper interval limit was chosen to be 0,08 (V2)^1"2 and 0,08 (Λ/2)^1"1, respectively. In other words, the area interval was increased by a factor of Λ/2 for each group in concurrence with what was previously exemplified.

The total proportion of the measured surface area that was covered by white spots was calculated by dividing the total area A that was covered by white spots with the total area T of the surface. Also, within each group, 1-8, the proportion of white spots belonging to that group was calculated by dividing the total area a_x that was covered by white spots in that group with the total area T of the surface. Table 2 shows the obtained proportion values A/T and ai/T-as/T for each sample.

Table 2

For each sample, the proportion ai/T of white spots in each size interval were normalized to the total proportion of white spots, A/T by calculating (a₁/1) / (A/T) = a_x/A for each group i, whereby a measurement of the relative amount of white spots within each size interval was obtained. Table 3 shows the normalized values, which represent a relationship between the relative amount of white spots and spot size.

Table 3

For each sample, a Weibull distribution according to formula 1 above was adapted, i.e. fitted, to the relative amount of white spots in each size interval. For sample 1 this is illustrated in Fig. 3, which shows the relative amount of white spots in each size interval as a function of the white spot area size. Fig. 3 also shows a Weibull distribution adapted to the set of relative amount vs. white spot area size pairs. After having adapted the Weibull distribution to the relative amount of white spots a mean value of the Weibull distribution was calculated. This procedure was repeated for all seven samples, whereby a Weibull mean value was obtained for each sample.

In order to validate the method, the seven board samples were evaluated after converting and printing in a commercial full-scale process. After the converting and printing the print quality was visually judged and a ranking socre from 0-10, wherein 10 is the best, was put to the samples respectively. Table 4 shows the ranking scores and the Weilbull mean values obtained for the seven samples .

Table 4

Fig. 4 shows the Weibull mean values as a function of the ranking score. From table 4 and Fig. 4 it is clear that a better printing quality was obtained for samples having a low Weibull mean value than for samples having a high Weibull mean value. Thus, by using the above-described method on an unprinted sample in order to obtain the Weibull mean value, a good prediction of the print quality of the sample after full-scale converting and printing operation can be made.

The above-described method is very suited to be used on line, i.e. during production of paper, paper board or board in a paper, paper board or board producing machine. The steps of the method can be automated using a computerized image acquisition system and the computational power of a central processing unit of a computer. The method and the apparatus implementing the method can be set to monitor the surface of the produced paper, paper board or board continuously, recurrently or intermittently, issuing an alarm if a statistical value indicating a surface unsuitable for printing is encountered. This value may be a predefined threshold value.

Fig. 5-7 illustrates the importance of the size distribution of the defects for the print quality. The dark areas in Fig. 5-7 are defects in the surface of the board where it is a risk that unprinted areas will occur. The calculated Weibull mean value for the board in Fig. 5 is 0.43. The calculated Weibull mean value for the board in Fig. 6 is 0.58 and for the one in Fig. 7 it is 0.78. Thus, from the Figs. 5-7 it is apparent that as the Weibull mean value increases, i.e. the print quality decreases, the defects are fewer but larger.

The invention has been described in connection with what is presently considered to be the most practical and preferred embodiment, but it is to be understood that the invention is not to be limited to this embodiment. The invention is therefore intended to cover various modifications and equivalent arrangements included within the spirit and scope of the appended claims.

Claims

1. A method of predicting the printability of an unprinted paper, paper board or board surface, characterized by the steps of:

- acquiring a topographic image of the surface;

- establishing a relationship between the relative amount of identified depressions regions and the areas of said regions;

- calculating a value on the basis of said relationship; and

- Using said value as measurement for the prediction of the printability of the paper, paper board or board surface .

2. A method according to claim 1, characterized in the further steps of;

- adapting a mathematical function to the relationship;

- calculating the value from the mathematical function.

3. The method according to claim 2, characterized in that said mathematical function is a statistical distribution.

4 The method according to claim 3, characterized in that said statistical distribution is the Weibull distribution.

5. The method according to any one of claims 2-3, characterized in that said value is the mean or the variance of the mathematical function.

6. The method according to any one of claims 1-5, characterized in that said threshold is set so that depression regions under a certain percentile are identified.

7. The method according to claim 6, characterized in that said percentile is the 10% percentile.

8. The method according to any one of claims 1-7, characterized in that said step of establishing a relationship between the relative amount of identified depressions regions and region size comprises the steps of

- calculating the total area (A) of all of the identified depressions regions; - arranging the depressions regions in groups according to size;

- calculating the total area Ia₁) of the identified depressions regions within each group; and - calculating a measurement of the relative amount of the identified depressions regions within each group by normalizing the total area of the depressions regions within each group to the total area of all of the identified depressions regions.

9. The method according to any one of claims 1-7, characterized in that said step of establishing a relationship between the relative amount of identified depressions regions and region size comprises the steps of:

- calculating the total area (T) of the surface;

- calculating the total area (A) of all of the identified depressions regions;

- arranging the depressions regions in groups according to size;

- calculating the total area (a_±) of the identified depressions regions within each group;

- calculating the total proportion of the surface area that is covered by identified depressions regions by dividing the total area (A) of all of the identified depressions regions with the total area (T) of the surface;

- calculating the proportion of the surface area that is covered by identified depressions regions within each group by dividing the total area Ia₁) of the identified depressions regions within each group with the total area (T) of the surface; and

- calculating a measurement of the relative amount of the identified depressions regions within each group by normalizing the proportion of the surface area that is covered by the identified depressions regions within each group to the total proportion of the surface area that is covered by the identified depressions regions.