FI88827B - REQUIREMENTS FOR ANALYSIS OF QUALITY PAINT PAPER - Google Patents

Description

1 88627

METHOD AND APPARATUS FOR ANALYSIS OF PAPER QUALITY

The invention relates to a method for analyzing the quality of paper 5 as defined in the preamble of claim 1.

The invention also relates to an apparatus for analyzing the quality of paper as defined in the preamble of claim 6.

Methods and apparatus 10 for analyzing the quality of paper are already known, in which radiation is passed through a paper sample and the absorption of radiation is determined. Based on the measured absorption, the basis weight of the paper sample, which describes the properties of the paper, is in turn determined.

15 The measurement of the paper sample is in principle a measurement of the basis weight distribution of the paper. This can be done directly by taking samples of the paper and weighing them. The formation can also be measured indirectly by transmitting a homogeneous energy flux through the sample and measuring the absorbance. This method indirectly provides information on the basis weight of the paper. The results must be specified by means of calibrations. Particle radiation or light is mainly used as energy flux. When using light, it should be noted that the absorption correlates well with the basis weight only for uncoated paper grades. Nevertheless, light has been used to measure the formation. It has been a question of monitoring changes in quality, not so much of an absolute definition of paper quality or a comparison of paper grades.

An apparatus for analyzing a paper sample is known in which a circular paper sample is placed between two glass plates and the sample and the glass plate are rotated at a constant speed. The absorption is detected by means of a light source and a light detector in which the diameter of the light spot 35 used is 1 mm. The pair of light source detectors are moved parallel to the radius of the paper sample to measure a specific area. The format of a paper sample is characterized by the ratio of the mean and standard deviation of the light passing through the sample.

Also known is a device in which a paper sample is attached to a transparent rotating cylinder. The light source and the detector are moved parallel to the axis of the cylinder. The size of the pixel formed by the light source can be varied, but it is usually on the order of 0.1 mm in diameter. The formation of the paper sample is characterized by the frequency spectrum of the measured radiation transmissions and the RMS (Root Mean Square) value of the total variance.

A paper web formation meter for a paper machine is also known in advance. The measurement is made locally, but the measurement location can be changed with respect to the paper web ratio. The measurement is carried out by means of a light source. The light source is on one side of the web and the detector is on the opposite side of the web. The pixel size is usually 0.2 mm. The meter is used to calculate the ratio of the average transmitted light to the RMS value 20, which is considered to characterize the paper formation.

Measurement methods and devices based on β-radiation transmission, such as PIRA and STFI, are also known for the analysis of a paper sample. According to the PIRA method, a paper sample is placed between the X-ray film and the β-emitting film. The X-ray film is irradiated and developed. The X-ray film is converted to numerical form and digitized with a microdensitometer. The sample size is 50 x 50 mm2. The pixel diameter is 1 mm. Sample 30 can then be numerically characterized by a computer.

The STFI method uses X-ray film in the same way as the PIRA method. The X-ray film, i.e. the radiogram, is converted to numerical form in a manner slightly different from the PIRA method, but the result is the density value of the radiogram. The pixel diameter is 0.1 mm.

r. o r O n 3 1 o u c. /

The radiogram is obtained from an area of 100 x 150 mm2. The formation of the paper sample is characterized by calculating the power spectrum of the measured transmitted radiation divided into wavelength ranges. In addition, the formation number can be calculated on both the mik-5 ro and macro scales.

The disadvantages of the above prior art can be seen. Imaging of current devices is based on a point source radiation detector pair. The analysis of the paper sample is then relatively slow, which means that the equipment is mainly suitable for laboratory use. It is difficult to adapt the equipment to the process environment.

A further drawback is that the use of radiological methods is slow, although the results are accurate. Such methods are not suitable for the process environment.

The range of current methods and equipment is inherently diverse. The disadvantage in this case is that the quality characteristics of the paper samples determined by different measurement methods and devices are not comparable.

The object of the invention is to eliminate the above-mentioned drawbacks.

The method according to the invention is characterized by what is stated in claim 1.

The apparatus according to the invention is characterized by what is stated in claim 6.

In the method of analyzing the quality of paper according to the invention, the radiation is passed through a research sample of paper 30, the absorption of the sample is measured and based on which the quality of the paper is evaluated and the quality of the paper is evaluated. said key figures. The invention is characterized in that - the image is treated as a matrix, which is analyzed by calculating its key figures such that - the matrix is viewed area by area starting from the whole area and proceeding towards the resolution area corresponding to the smallest detection area of the matrix; 5 - generating a table of variation values for each region of the matrix, and - using as variance value the variance of the intensity of radiation entering each region through the test sample; and that 10 - the values contained in each table correspond to said parameters by which the quality of the paper is assessed.

In one embodiment of the method, the value of variation is calculated for the entire matrix and for each subregion of the matrix, which are further determined downwards by dividing the mat-15 rice to a suitable range; for each resolution range, a histogram is generated in which data on the occurrence of certain variance values are collected; and histograms are used as said key figures.

20 In one embodiment of the method, the variation value is calculated in a region of the matrix as a function of direction and distance for each resolution range; for each resolution range, a combination representing the occurrence of certain variance values or a co-occurance matrix of Figure 25 is generated; and said combinations and common histograms are used as key figures.

In one embodiment of the method, a number of paper quality classes are determined; the indicators are determined for each 30 paper grade grades; the key figures corresponding to the paper quality grades are stored; the indicators determined for the research samples are compared with the indicators of the different paper quality classes; and based on the comparison, the test sample is classified into a suitable paper grade.

35 In one application of the method, the determination of the quality grades is carried out by measuring representative samples of those grades and determining the 5 f S C 2 7 characteristics for the samples.

The apparatus for analyzing the quality of paper according to the invention comprises a radiation source and a detector for measuring the absorption of a paper test sample and a processing unit for calculating on the basis of the measured absorption indices describing the quality of the paper. at least approximately homogeneous radiation-10 flux on the surface of the test sample; a camera as a detector, by means of which the image is formed from the test sample, which camera is placed on the other side of the test sample relative to the lighting device; and the processing unit includes: an image memory in which an image of the examination sample received from the camera is stored. According to the invention, the apparatus comprises means for examining differences in the radiation intensity of the image from one region at a time to a resolution region, the resolution region comprising one or more pixels; means 20 for calculating variations, i.e., variations, in the statistical radiant intensities of each region studied; and means for storing the variation values in appropriate tables, which table values are used as indicators in assessing paper quality.

25 In one embodiment of the apparatus, the means for examining differences in the radiant intensity of the image includes a plurality of parallel processors for processing the image area by area in parallel.

30 In one application of the apparatus, the means for calculating the variations, i.e. the variation values, of the statistical radiant intensity of each area studied are arranged in parallel processors.

In one embodiment of the hardware, the concept unit includes a switch through which results are collected from each processor and stored in suitable memory units for further processing.

. p o <* Ο Γ

In one embodiment of the apparatus, the means for storing the variation values includes a classification memory in which the identification numbers of the research samples of the pre-classified papers are stored; and a measurement-5 memory in which the key figures calculated on the basis of the measurements of the actual research samples are stored; and that the processing device comprises a comparator in which the indicators of the measurement memory are compared with the indicators of the classification memory and on the basis of which the quality classification of the paper 10 is performed.

The advantage of the invention is that it is based on the formation of a matrix-shaped image of a research sample. This facilitates unambiguous characterization of the sample.

A further advantage of the invention is that it can be applied both in the process environment and in laboratories. The test sample can be photographed from a paper web moving with the camera. In a process environment, fast access to results is important for control.

It is also an advantage of the invention that the key figures 20 are unambiguously determined and relatively reliably describe the quality characteristics of the paper.

A further advantage of the invention is that, based on the method, the research samples of the paper can be classified into the desired quality classes.

A further advantage of the invention is that the quality classes can be taught to the measurement arrangement by means of paper research samples of typical representatives of those classes.

The invention makes it possible to change the quality classification if necessary; either loosening or clarifying it

The invention also makes it possible to obtain additional information and justifications for a paper classification decision made after classification.

It is also an advantage of the invention that the accuracy of the quality classification depends only on the quality of the research or measurement sample.

_ O O r o 7 '1 O o l i

Furthermore, thanks to the invention, the conversation between the printing house and the paper manufacturer, for example, is facilitated in the case of complaints. Measuring equipment can be provided for each party, which can be taught for exactly 5 specific paper grades. The taught features, i.e. key figures, can be transferred from one hardware to another by electronic or conventional mailing.

The invention also provides, with greater accuracy and resolution, the ability to measure paper research samples than currently known methods and apparatus.

Furthermore, the invention facilitates the paper manufacturer in quality standardization. For example, after a quality change, it is easy to find the paper quality level used in the previous run 15.

Furthermore, thanks to the invention, as the quality level of paper increases or decreases, the method and apparatus can be adapted to the changes through simple teaching.

A further advantage of the invention is that the method as such is suitable for both on-line and laboratory measurements.

The invention will now be described in detail with reference to the accompanying drawings, in which Figure 1 schematically shows in block diagram form 25 an apparatus according to the invention; Figure 2 illustrates an image taken from a research sample; Figure 3 illustrates a histogram; Figure 4 illustrates an image of a study sample 30 from which joint histograms are calculated; Fig. 5 shows in block diagram form a memory unit and a comparator belonging to an image processing unit; Figure 6 shows schematically in block diagram form another apparatus according to the invention.

In the following, a research sample means a paper sample or a radiogram made from it. In the case of a '"' O 'O' 7 8 '- (_- l · Z / thin paper, the test sample is preferably a paper sample. In the case of paperboard or thick paper, the test sample is a radiogram of the actual paper sample.

5 In this context, the study of paper quality refers to the measurement of formation, which is in principle a measurement of the basis weight distribution of paper. In this case, the formation is measured indirectly by transmitting a homogeneous energy flow through the sample 10 and measuring the absorption.

A variety of radiation sources can be used to measure basis weight by absorption. The most common types of radiation used are soft X-rays, particle radiation, and optical radiation. 15 The use of X-rays and particulate radiation is limited by, among other things, the size of the equipment needed to produce the radiation, its cooling, safety regulations and price. Optical radiation with a wavelength range extending from near-infrared radiation to ultraviolet radiation does not have the disadvantages of the previous radiation alternatives. Optical radiation can be produced in a small space, waste heat does not cause problems, and radiation is not dangerous to the environment at all.

The radiation transmission T, 0 <T <1, in the medium can be generally described by the formula T = e’uW (1) where u is the absorption coefficient or the corresponding radiation shear coefficient and W is the grammage.

A radiogram is made of a thick paper sample by placing the paper sample and X-ray film to be examined on top of the X-ray or particle radiation source. The radiation is allowed to act on the X-ray film 35 through the sample for a period of time. The film is developed, after which its optical density can be measured. The optical density D can be described by the formula Λ o n o '· 9 * o l / D = c t e "uW + D ° (2) where c is the proportionality constant, t is the exposure time and D ° is the optical density for an infinitely dense sample.

5 The other parameters correspond to the parameters defined in connection with formula (1).

In the method and apparatus according to the invention, the optical radiation flux is passed through a test sample and the absorption of the sample represented by the radiation trans-mission T or the corresponding optical density D is measured. As the radiation source, a lighting device 1 is preferably used, which produces and applies a homogeneous field of optical radiation to the research sample 2. The camera 3, such as a video camera or a semiconductor camera, acts as a toi-15 mix detector to form a two-dimensional image or matrix 2a (Figures 2 and 4). The camera 3 is placed relative to the lighting device 1 on the other side of the test sample 2, as illustrated in Figure 1. Arranged in this way, the measurement sample scatters and attenuates the transmitted radiation in a manner typical of the fiber structure and quality of the sample.

With the aid of the camera 3, the image formed from the examination sample is transferred to the image processing device. In the concept device, the image, i.e. the matrix, is analyzed and key figures are calculated for it, which describe the formation of the research sample.

The formation of the paper or paper sample is characterized in this context by the texture. Texture 30 can be defined as a large number of more or less ordered similar elements or figures without any special attention being paid to them. The observer perceives the texture as a large uniform impression. The texture may be a discipline of 35 matrix-like sub-characters, such as a chessboard or a weave. Such a texture is called deterministic and can be described by sub-characters ίο η ε :, 2 7 or by elements and element placement rules. The texture impression can also be random and is then called stochastic texture. Such a texture is reminiscent of noise on a television screen 5. It can also be said that between deterministic and stochastic textures there are a number of deterministic textures in which the rhythm of the element placement rule is strongly disturbed.

The formation of the paper can be characterized as a stochastic texture formed by its random fibrous structure and the random distribution of other ingredients. In this case, the paper quality indicators are in principle determined as follows. Consider one or more areas of the matrix. The amount of transmitted radiation is measured and the random variance of the transmittance is calculated for each range. A table of variation values is generated for each resolution range corresponding to the smallest detection range of the matrix. The values contained in each table represent the said 20 key figures.

The apparatus according to the invention comprises a processing device with an image memory 4 and a number of units, such as an image analysis unit 5, a calculation unit 6 and a memory unit 7, by means of which the image is examined, variation values are calculated and the results are stored for further processing.

The image from the examination sample obtained from the camera 3 is stored in the image memory 4. The radiant intensities of the image are examined one resolution range at a time in the analysis unit 5. The resolution range is defined as an area the size of one or more pixels in the image, i.e. in the matrix. The variations of the statistical radiation intensity values of each resolution range, i.e. the variation values, are calculated in the calculation unit 6. The variation values are stored in the memory unit 7 in the form of suitable tables. The table values are used in further processing as indicators to assess the quality of the paper.

a O r ο>

Il <o V- Z /

It should be noted that the above-mentioned image analysis unit 5 and calculation unit 6 can be implemented not only as concrete units but also by software means as software units of the processing device.

5 This fact also applies below, at least for the units involved in image analysis and calculation.

The key figures can be calculated in two preferred ways, which are presented below.

10 The set of variance of the intensity of the radiation coming through the research sample, ie the transmittance (or density), is used as the variation value. First, the variance is calculated for the entire image, which contains e.g. 512 x 512 pixels. The image is then divided into four parts, each of which contains e.g. 256 x 256 pixels, for which variances are calculated for each part. The sub-images thus obtained are further divided into four parts and the variances are calculated for the parts. This is continued until an image resolution range e is reached, which is e.g. 4x4 pixels in size. This process is illustrated in Figure 2. The variances calculated in this way form a separate histogram for each resolution area e of the image, i.e. a table showing how many times a certain variance has occurred. Such a histogram 25 is shown as a bar graph in Figure 3. These histograms are used as indicators to evaluate the quality of a paper measurement sample.

In another method of calculating the parameters, the set of variance of the intensity of the radiation coming through the research sample, i.e. the transmittance, is also used as the variation values. The variances are calculated in the area of the image, i.e. the matrix, as a function of direction and distance for each resolution area (cf. Figure 4). For each resolution range, a combination or co-occur matrix d (i, j, direction, distance) describing the occurrence of certain variance values 35 is generated.

opr o> 12 1 <- ‘

The joint histogram should be calculated in two directions because the paper is not an isotropic material. In practice, however, one direction is sufficient to achieve the required resolution. Distances corresponding to the resolution range should be used, as in the first calculation option. The joint histogram can also be calculated at one distance, but then variations in the size of the image area affect the calculation results. The obtained common histograms for different image areas can be used as indicators per se. To simplify the calculation method, common histograms can be characterized by their own key figures. The number of transition types, the number of transitions squared, and the inertia of the transitions can be used as such indicators. There are other possibilities, but the simple indicators presented have proven to be reliable.

In the method according to the invention, it is possible to classify the research sample into different paper quality classes. These paper quality grades are defined in advance using the same criteria as the actual study sample is classified. In practice, this is done by measuring a large number of representative research samples of the paper grades in question and determining the indicators for these samples. The indicators assigned to the research samples are compared with the indicators of different paper quality categories. Based on the comparison, the research sample is classified into a suitable quality class.

The apparatus according to the invention, in which said teaching of quality classes is utilized, comprises a classification memory 7a and a measurement memory 7b, as can be seen from Fig. 5. The classification numbers of research samples of pre-classified papers are stored in the classification memory. The key figures calculated on the basis of the measurements of the actual research samples are in turn stored in the measurement memory 7b. The apparatus also includes a comparator 8, in which the parameters of the measuring memory 7b are compared with the parameters of the creation memory, r> o ο o r '> 13 · o i. Based on the comparison, the quality classification of the paper is stated and expressed.

In the apparatus according to the invention, the image analysis unit 5 can be implemented with a number of parallel processors 9, as shown in Figure 6. By means of the processors 9, the image, i.e. the matrix, is processed area by area in parallel form. It is preferred that the computing unit 6 is also located in said parallel processors 9. The parallel processing-10 pa significantly speeds up the computation when analyzing the image, whereby the results for further processing are obtained quickly.

The parallel processors 9 are followed by a gear 10, through which the variation values of the respective regions 15 are collected from each processor and stored in a suitable memory unit 7; 7a, 7b.

The number N of parallel processors 9 is selected to be suitable. It depends on the size of the image to be processed and in particular on the number of pixels and the size of the resolution areas. The complexity of computing processes and the speed requirements set also affect the number of processors. In principle, the processors can be microprocessors.

The invention is not limited solely to the application examples presented above, but many modifications are possible while remaining within the scope of the inventive idea defined by the claims.

Claims (11)

1. A method for analyzing the quality of paper in which a radiation is passed through an examination sample of the paper, the absorbance of the sample is measured and, on the basis of this, the characteristics of the sample are described which describe the quality of the paper and by which the quality of the paper is estimated. in which method of the test sample a two-dimensional image is formed which is analyzed by an area of the thread and for this purpose the said characteristics, characterized by it, are calculated that - the image is treated as a matrix which is analyzed by calculating the characteristics of the sample. , the 15 matrix is examined with one area eating the gangway, starting with the whole range and advancing toward a distribution capability range corresponding to the smallest array's observation capability range; - a table of variance values is compiled for each region of the matrix, and - as variance values, it is used throughout the examination sample for each area of the radiation intensity variance; and that the values contained in each table correspond to said characteristics, by which the quality of the paper is estimated. 2. A method according to claim 1, characterized in that the variation value is calculated for the entire matrix and for each sub-region of the matrix, which are further determined down-grid by dividing the matrix into a suitable range; - for each distribution property area, a histogram is formed, whereby data on the occurrence of certain variance values are collected; 35. the histogram is used as said characteristic. 3. A method according to claim 1, characterized in that the 18. c Γ »2 '/' variation value Within the area of the matrix is calculated as a function of the distance and distance for each of the previous 1 n 1 times of the range. ; - for each distribution range, a descriptive combination of occurrence of variance value or the co-occurance matrix of the image is formed; - said combinatons and cohistograms are used as characters. 4. A process according to any of the preceding claims, characterized in that - a variety of paper quality classes are defined; - character is determined for each paper grade; - the characters corresponding to the paper quality classes are stored; - the characteristics determined for the examination sample are compared with the characteristics of the different paper waste classes; - because of the comparison, the test sample 1 is classified as a suitable paper swelling class. 5. A method according to claim 4, characterized in that the classification of the evening classes is done by measuring the typical paper samples of measurement questionnaires and by determining the characteristics of the sample in question. 6. An apparatus for analyzing the baldness of paper, to which a radiation source and a detector are associated, the absorbance of which is determined by the absorbance of the examination paper of the paper, and a processing unit, by which the characteristics of the absorbed absorbance are calculated, the characteristics of the samples are calculated. describing the paper's baldness, to whichever device belongs - a side illuminating device (1) as a source of radiation, with the resultant an at least almost homogeneous flow is achieved on the surface of the examination sample (2); - a camera (3) as a detector, with the help of each camera, an image of the examination sample is formed, each camera 1 IQ POrO - · 19. o Z / relation to the lighting device (1) is located on the other side of the examination sample (2); and belonging to the processing unit: - an image memory (4), in which the image (2a) of the examination sample (2) is stored from the camera (3); characterized in that the device includes means (5) for examining an area eating the thread from the entire area all the way to the distribution property area 10 (e) of the radiation intensity differences of the image, which distribution property area comprises one or more pixels; - means (6) for calculating the statistical radiation intensity variance or variance values of each investigated area; and - means (7) for storing the variation values in suitable tables, which table values are used as characteristics where the quality of the paper is estimated. 7. Apparatus according to claim 6, characterized in that the means (5) for examining the differences of the radiation intensities of the image belong to a number of parallel processors (9), by means of which the image is treated with a region eating the thread in parallel form. 8. Device according to claim 7, characterized in that the means (6) for calculating the variations or varying values of statistical radiation intensities of each investigated region are placed in parallel processors (9). Device according to claim 7 or 8, characterized in that the processing unit includes a switching device (10), through which the results coming from each processor (9) are collected and stored in suitable memory units (7) for further processing. 10. Device according to any preceding claim. 20. Examination requirements 6 - 9, characterized in that the median (7) for storing the variation values - a class memory (7a) in which the characteristics of pre-classified paper examination samples are stored; - a measurement memory (7b), in which the characteristics calculated for the measurements of the actual test samples are stored; and that the processing unit includes a reference device 10 (8), the characteristics of the measurement memory being compared with the characteristics of the class memory.