Topic of your interest

Basis for standardization of digital production printing using toner based and large format inkjet printers

Short name: Image quality

Fogra-No. 35.003
Project leader: Dr A. Kraushaar
Funding: BMWi (IGF) via AiF
 

Timescale: 1.10.2013 - 30.09.2015

Objective and relevance

In this research project, prints in digital large format as well as in high-volume small format were systematically examined with regard to image quality. The main focus of this project was the process-independent evaluation of inhomogeneities and the detail sharpness. Two methods were developed which allow an objective evaluation of the respective image quality attribute with a high correlation to the visual results. 

On the basis of extensive visual experiments, an index (M-Score) including the evaluation matrix was derived in the first method, which ranges from 0 (worst homogeneity) to 100 (perfect homogeneity). In the second method the L-Score was developed. It describes the detail sharpness.

In addition, further evaluation methods were developed. These are the number of visually recognisable tonal value levels (P-Score), the graininess (Z-Score) and printer noise.

Dr Andreas Kraushaar

Head of Prepress Technology

+49 89 431 82 - 335

get in touch

Your contact

I would be pleased to answer your questions about the project and look forward to exchanging ideas with you.

Solution steps

A plausibility test has been developed for print samples with pronounced directional dependence (anisotropy), specifically for non-paper-like substrates, to help determine the suitability of conventional colour measurement technology for capturing the colour appearance. Read more in the PSD handbook.

The main focus of this project was the process-independent evaluation of inhomogeneities and the sharpness of detail. Two methods were developed which allow an objective evaluation of the respective image quality attribute with a high correlation to the visual results. In the first method, homogeneously applied colour areas are recorded in spatial resolution using standard test chart readers and evaluated by image analysis. On the basis of extensive visual experiments, an index (M-Score) including the evaluation matrix is derived, which ranges from 0 (worst homogeneity) to 100 (perfect homogeneity). Here a 2D cross-correlation between the ideal-typical (data-technical) reference and the scanned image is carried out.

By complying with the requirements for image acquisition and processing as well as for the simulation of the human visual system by means of blur filters dependent on the viewing distance, an index (L-score) was developed which enables a high correlation with the results of the visual tests. 

Achieved results

The test parameters developed were incorporated into the ongoing standardisation work, specifically the 18621 series of standards.

The current status:

  • ISO/TS 18621-21: M-Score (Image quality evaluation methods for printed matter – Part 21: Measurement of 1D distortions of macroscopic uniformity utilizing scanning spectrophotometers)

  • ISO/TS 18621-22: Z-Score  (Measurement of visual attributes of printed materials – Part 22: G-Score test method for evaluation of graininess)

  • ISO/TS 18621-31: L-Score (Image quality evaluation methods for printed matter – Part 31: Evaluation of the perceived resolution of printing systems with the contrast–resolution chart)

In the course of the project a comprehensive collection of image quality elements was compiled. It currently comprises 25 pages, is continuously developed and expanded with new test elements. The Image Quality test forms are available free of charge.

Download test forms

Downloads

Investigation of homogeneity

What do the M-Score values mean?

M-Score 

Category

Practical meaning

≥ 95

Perfect

Inkjet test print on high-quality test print paper. No inhomogeneities visible.

≥ 80

Very good

Print with slightly visible inhomogeneities (e.g. some mottling). No visible stripes.

≥ 70

Good

Print with visible inhomogeneities (mottling) but almost no visible stripes. Typical limit for offset quality.

≥ 60

Satisfactory 

Mottling and stripes visible, still accepted by most observers.

≥ 50

Adequate

Print with clearly visible mottling and/or stripes. Acceptance is highly dependent on the printed image.

< 50

Poor

Clearly visible mottling and stripes. Not accepted as high quality print. The acceptance depends strongly on the application.

Investigation of detail sharpness

What do the L-Score values mean?

L-Score 

Category

Practical meaning

≥ 90

Perfect

Inkjet test print on high-quality test print paper. Excellent contrast and high detail sharpness.

≥ 80

Very good

Very high detail sharpness.

≥ 60

Good

Medium level of contrast and spatial resolution. Finest details are no longer resolved.

≥ 40

Satisfactory 

Fine structures are no longer visible. For most applications still saleable print quality.

≥ 20

Adequate

Contrast and fine structures are usually not adequately reproduced. For simple applications ok.

< 20

Poor

Contrast and fine structures are reproduced very weakly.

Image Quality report

In the context of the investigated image quality attributes, the focus is on a compact summary of the results. The aim is a clear compilation of the essential image quality criteria. This can form the basis for standard or customised assessments of printing systems.  

Image Quality Report