| Abstrakt: |
Printing quality control has been one of the challenges in the printing industry. Therefore, it is indispensable to find evaluation methods with high correlations with the human visual system to assess the quality of the printed images. Standard ISO/IEC 24790 is one of the recommended standards to achieve this goal. An evaluation procedure for graininess, a common defect in printing, has been presented in this standard; however, this procedure is suggested only for monochrome printed images. Hence, in this research, the graininess of coloured samples is evaluated based on the proposed method in this standard with considering the effect of the paper. For this purpose, fictitious graininess defects are designed and applied on coloured samples and are printed on different papers. A visual assessment is performed to examine the effect of colour and paper differences on graininess. Six different methods, including the method suggested in the standard and the proposed modified procedures, are used to evaluate graininess, and their results are compared to subjective results. Moreover, three other wavelets are studied besides the suggested one in the standard. To eliminate the effect of paper texture, a Gaussian filter is utilised. The results show that applying the Gaussian filter to images increases the correlation between computational procedure results and visual assessment results. Also, the results indicate that utilising haar wavelet slightly increases the correlation. The results demonstrate that besides assessing graininess in greyscale mode, evaluating graininess in RGB (red, green, blue) colour space results in a high correlation with the visual assessment results. [ABSTRACT FROM AUTHOR] |
Plný text