Jun 04, 2015
Commercial systems dedicated to assess print quality already exist. They usually implement the ISO 13660 standard with a focus on character and line attributes (raggedness, line width, darkness, character, contrast, fill, extraneous marks, character field, background haze, character field)
However, they are expensive, lack of flexibility and the access to their algorithms is not possible.
Fiji/ImageJ and its environment enables to develop an open source program dedicated to print quality assessment› free license› Interface based› Development using Java language and pre-coded functions› Enhance collaboration
Registration mark
Features extraction
Statistical analysis
Print target design
Choice of relevant features
High resolution & low field of view vs low resolution & high field of view
Printing sequences
Inter-samples
Intra-samples
Dot
Lines
TextLighting conditions
Image acquisition
Printing
Fiji
Printing parameters(drop speed,…) Substrate
Ink
Printing configuration
System model
Dithering techniques
Fiji is an image processing package. It can be described as a distribution of ImageJ together with Java, Java 3D and a lot of plugins.
User interface
Macro/plugings editor
Individual ‘bricks’-Color conversion-Binary analysis-Illumination correction-…
Increase productivityAutomationShorter development timeProprietary routines and development of standard
Software development MacroJava based program
Image creation to design test charts
Dithering techniques
Dots and satellites
Line width/raggedness
Series of parallel lines
By example, test of optical density regularity along the swathe › Periodic pattern
› Random pattern
Comparison and impact of different dithering techniques
Implementation of algorithms specific to greyscale printhead
Background subtractionRGB – one layerBinarization
Particles extraction Particles counting
Histogram for a given parameter
Individual object’s properties
› Edges extraction + angular correction + average and standard deviation on both sides of the line
Idea
di
%
di
Histogram
REFERENCE
Original image
The centre of each line is computed to obtain the distance between two successive lines
The histogram is then plotted
implementation The red circles represent the middle of the line
Average and standard deviation are then calculated
Colour deconvolution
Results tolerance
Text recognition and limit of “readability”
Objective: to quantify colour variations caused by due to undesirable drops overlap,...
Colour clustering
in RGB space
Colour split
Placement accuracy
% of coverage
Printing direction
Dithering techniques
Objective: to evaluate the interval of confidence for results like dot diameter, (x,y) positioning, roundness,…
Pixel size: 2 mDot diameter: 60 m
Distribution of error centred of the theoretical dot area
Algorithm based
on dot permutation
Image analysis is an approach, but not the end of the story
PQ can be described along several others physical dimensions (Optical density, colour gamut, full tone uniformity…)
Some levels of overlapping exist between them
If you:› want to know more about ImageJ / Fiji› are interested in developing an ISO13660
open source software› have a specific need regarding print
quality assessment, beyond image analysis
I would be please to heard from you.