33. A Method of Color Inverse Halftoning Image Quality Assessment Based on Image Structural Property

The traditional image quality assessment method and index cannot meet with the inverse halftoning image quality assessment. We proposed a color inverse halftoning image quality assessment method based on image structural property. First, we transformed printing image into inverse halftoning image with wavelet transforming inverse halftoning algorithm. Then, taking quaternion matrix as the carrier and combining with details of image, brightness and color information, we can convert halftoning image into quaternion matrix and obtain feature vector through singular value decomposition. Finally, the inverse halftoning image quality assessment is accomplished by calculating the quaternion singular feature vector angle to indicate the similarity of the original image and the inverse halftoning image. The result of experiment shows that the assessment can excellently exhibit the reduction of the inverse halftone image quality. Meanwhile, it is also in line with subjective visual perception and provides a reference for constructing quantitative assessment index of the inverse halftone color image quality.