by clicking the arrows at the side of the page, or by using the toolbar.
by clicking anywhere on the page.
by dragging the page around when zoomed in.
by clicking anywhere on the page when zoomed in.
web sites or send emails by clicking on hyperlinks.
Email this page to a friend
Search this issue
Index - jump to page or section
Archive - view past issues
FLEXO Magazine : February 2009
TECHNOLOGIES & TECHNIQUES Press sheet # 3 I . I I . I I I I . I I . . I . . . I I I I I - True CIELAB values -( Camera CIELA values r How well do these agree? Press sheet # 7 > True CIELAB values True AE > Camera CIELAB Camera AE values FIGURE 5. Test of how well the camera can measure AE. Thus, I had 1,296 different pairs of AE values to compare for each combination of two sheets. A little cogitation will show that there are 36 distinct pairings between the nine sheets. This results in 46,656 comparisons. I generated these for each of the four cameras. Table 2 summarizes how well each of the RGB devices would be able to estimate AE. Recalling from before that an ideal color measurement device should be able to measure with accuracy of 0.2sAE, and that 0.7sAE is still acceptable, any of these devices are acceptable for the production run requirements of ISO 12647-6. The table also lists what I call the "false conclusion rate." A measurement of the color of a production sheet results in a sheet either being in tolerance or not. The false conclusion rate is the percentage of time that a spectrophotometer and an RGB camera would disagree on whether a sheet is in tolerance. Clearly we TABLE 2. Accuracy of color difference measurement with an RGB - would want the two to never disagree, but is 3 percent an accept- able error rate? These pass/fail numbers are used to assess the run. The ulti- mate useful measurement is whether 68 percent of the sheets are passing. The tolerance range (for assessing the whole run) is thus o percent to 68 percent. By the previous discussion, a device to measure the number of sheets that are within tolerance should be accurate to within 6.8 percentage points, and must be accurate to within 20 percentage points. An RGB camera is clearly acceptable for measuring a color change during the run. . ABOUT THE AUTHOR:]ohn Seymour is a principal engineer for QuadTech, where he has been doing research in printing, color theory, and imaging for the past 16 years. Seymour was instrumental in the development of QuadTech's CCS (Color Control System). He currently holds 11 patents, has authored 11 technical papers, and is a member of CGATS. Prior to working with QuadTech, Seymour worked in medi- cal image processing, wrote software for alignment of digital imagery from weather satellites, worked on a team that designed an electron microscope and an accompanying image processing system, and wrote software for spectroscopy. He holds Bachelor's degrees in mathematics and in computer sciencefrom the University of Wisconsin-Madison. FEBRUARY 2009 www.flexography.org FLEXO
Sustainable Winter 2009