Subjective Image Quality for Multi-Level Error Diffusion and Its Objective Evaluation Method

Kimiyoshi Miyata, Norimichi Tsumura, Hideaki Haneishi, Yoichi Miyake · Journal of Imaging Science and Technology · 1999

The quality of images produced by multi-level error diffusion was evaluated both subjectively and objectively. The images were recorded on color transparencies by a digital film recorder with a resolution of 1600 dpi. Thirty-six sample images with four different resolutions and nine different quantization levels were prepared, and the tone reproduction, color reproduction, graininess and sharpness of the images were analyzed and evaluated. We found that the graininess gives the strongest influence on the quality of the image. The relationship between observer rating values and measured quality criteria are also discussed.

Read the paper · More papers on PaperTik