A unique image analysis technique for the determination of the wicking properties in paper produced for inkjet application

Inkjet print technologies have become common in our lives. Beginning with the original home and office applications, today they are used in the production of many printed pieces ranging from direct mail and banners to large wide-format posters.  Inkjet printing is used increasingly in offset print technology. At DRUPA 2012 printing machine manufacturers displayed inkjet heads mounted on an offset press producing personalized printed pieces in a single step.

Due to inkjet’s speed and rapid set-up, application growth is certain.  The inkjet print process is non-contact as such it differs from most other printing processes; To create the image, inks are propelledthrough the air as fine droplets to strike the medium. The electronic pump creating these droplets requiresthem to be very low in viscosity, and as such, the droplets are at least partially absorbed by the paper. Inksorption directly affects the image sharpness or visual clarity, which is often referred to as wicking or raggedness. Wicking occurs between printed characters, lines, and in half-tone images between the dots.

In the case of inkjet print, dot sharpness and internal uniformity are affected not only by the formulation of the ink but also the properties of the paper, its surface treatment and the internal size, both of whichaffect the absorptive properties of the paper.  A requirement for glossy inkjet paper coating is limiting the spread of the ink, i.e. to fix the inkjet colorants to the coating.  In this study papers that have been specifically made for inkjet printing and regular offset papers will be tested with water-based and UV inkjet inks. It will be noted if the commercial inkjet printer applies a clear sealant prior to printing coloredinkjet ink.

Although the differences between two papers may be minimal in terms of wicking (i.e. raggedness and feathering), it can make the difference between a sharp, crisp looking print and a fuzzy looking print. This applies to text appearing fuzzy or too bold and half-tone areas will darken as more wicking occurs.

The common test inkjet pattern contains lines but because scanners were used to acquire the test images,the authors chose to use dots as the primary test pattern, contrary to common practice. Using dots minimizes the possibility of the Moiré effect in the analysis: e. g. Misalignment of the rectilinear printed line with the rectilinear scanner camera and the subsequent rectilinear image analysis.  In addition, dots are the basis half-tone images and the analysis program used in these tests was initially developed to analyze the quality of the offset/flexographic/gravure half-tone print; it was enhanced somewhat to handle non-contact inkjet tests as well.  Practice has determined there is no difference between the perimeter wicking of a line and a dot. In this analysis the authors have chosen to use dots with a diameter of 2mm printed in a range of colors (CMYKRGB).

To measure ink wicking, the authors have chosen this pattern of dots:

[IMAGE REMOVED]

These dots are uniformly 2 mm in diameter and may be printed in any color,within fields of any color.  Dots offer many advantages to the image analyst using, as we have, a high quality scanner, because even at a low resolution of 600 ppi, dots cannot be misaligned.

[IMAGE REMOVED] 

Immediately below the dots is an unprinted area outlined in green.  To prevent any contamination from ink over-spray on the original print, this area, with the dots above, is the final image on the page. In this unprinted area, defined by the green box, the analysis system measures the paper reflectivity and topography as shown to the right.  The panel shows the actual topographic surface image of the paper on which the dots were printed. This area was imaged, and then measured for background reflectivity and topography before the dots were measured.

To measure the dots the software uses a concept known as “Thresholding”.  To understand the workings of a threshold one must first understand the image in which the measurement will be made is an 8-bit derivative of a full color 24- bit image. In an 8-bit image the pixel luminance (brightness) can have a value between 0, (pure black), to 255 (pure white). To see and to measure an object within the image it must contrast with its immediate surroundings. The degree of contrast is determined by the pixel luminance value ( i e. pixel or picture element brightness).  The IA program extracts or identifies the object of interest based upon the threshold value; those pixels having a value less than or equal to the threshold value are identified as being of interest.  The IA program measurement algorithm then associates the identified pixels to form the objects that are then measured and reported.  The associated pixels forming the object of interest, in our case a dot image, are further analyzed to compute the mean luminance value of all the pixels within its perimeter, i.e. its brightness. The series of dots shown below illustrate how the IA computes the mean luminance at five (5) progressively higher threshold values. These multiple thresholds are set using a rigid mathematic progression that measures the dot from its core or darkest value through to its fully wicked condition. 

The pictures below show the same single black inkjet dot printed on uncoated white paper measured by the software using the progressive threshold.  It is the same dot in all the pictures taken from the IA system in an actual test.

[IMAGE REMOVED]

Below each of the images are data and the pixel luminance histogram derived from the application of progressively higher thresholds.  The green line traces the perimeter of the dot. 

With the lowest threshold in picture 1, the core or darkest portion of the dot is measured, as indicated by the equivalent diameter (Eq. Diam) measurement of 2.017 mm. As the threshold is changed the measured equivalent diameter becomes larger and the other property measurements change as well.  Of primary interest is the variation in the pixel luminance value within the perimeter of the dot. The degree of variance in the dot mean pixel luminance measurement from low (core) to high threshold (maximum wicking) is the prime determinant of inkjet wicking.  By computing the standard deviation of the mean pixel luminance value within each dot perimeter the IA program presents us a measure of ink wicking and image quality.  In the pictures above:

​The mean pixel luminance within the perimeter:

​1. 77.29 (The minimum or core of the dot)

​2. 80.26

​3. 83.78

​4. 89.23

​5. 108.4   (The maximum wicking into the inter-space)

​The Mean Pixel Luminance Standard Deviation:       = 12.35

The Mean Pixel Luminance Standard Deviation is proposed as the basis of comparative measurement of substrate quality for inkjet print.

Inter-color bleed is also an inkjet quality problem that is similar to wicking. This inter-color bleed depends on the colors that are being printed and can have an adverse effect on small colored text printed on a colored background. Line width and raggedness are defined in ISO 13660 and parts of this method will be used to evaluate the printed samples by substituting dots for lines.

The porosity of the specimen paper will also be evaluated using a Cobb tester to see if there is a correlation between the measured absorptivity and the observed and measured wicking. The Cobb Test measures surface water absorption over 60 seconds, expressed in g/m2.  The procedural standards for the Cobb test are explained in the TAPPI method T 441.

One of the authors of this paper (Rosenberger) has done analysis of inkjet wicking and feathering in the past. Based on his analysis the new software was developed. Using the new Verity IA Micro Dot software presents a clear advantage in combining all analysis methods into one step and exports the results directly to an EXCEL spreadsheet for further evaluation.

Apart from doing an evaluation with the help of image analysis software the printed samples will also be shown to a panel of human observers. Before the human observer can participate he/she will have to undergo an Ishihara test for color blindness. Once they have passed this test they can proceed with the evaluation of the printed samples. Each observer will have a sheet that contains the judging criteria. Since the evaluation will only be of a qualitative nature the observers will choose, in their opinion, the best sheet and compare the other printed samples to their reference sheet. It will be noted which sheet was chosen by the observer as the reference sheet.

The human component will show if results gained from the image analysis software correlate with the human perception of print quality.