Jun Jiang


Fine arts refer to a broad spectrum of art formats, ie~painting, calligraphy, photography, architecture, and so forth. Fine art reproductions are to create surrogates of the original artwork that are able to faithfully deliver the aesthetics and feelings of the original. Traditionally, reproductions of fine art are made in the form of catalogs, postcards or books by museums, libraries, archives, and so on (hereafter called museums for simplicity). With the widespread adoption of digital archiving in museums, more and more artwork is reproduced to be viewed on a display. For example, artwork collections are made available through museum websites and Google Art Project for art lovers to view on their own displays. In the thesis, we study the fine art reproduction of paintings in the form of soft copy viewed on displays by answering four questions: (1) what is the impact of the viewing condition and original on image quality evaluation? (2) can image quality be improved by avoiding visual editing in current workflows of fine art reproduction? (3) can lightweight spectral imaging be used for fine art reproduction? and (4) what is the performance of spectral reproductions compared with reproductions by current workflows? We started with evaluating the perceived image quality of fine art reproduction created by representative museums in the United States under controlled and uncontrolled environments with and without the presence of the original artwork. The experimental results suggest that the image quality is highly correlated with the color accuracy of the reproduction only when the original is present and the reproduction is evaluated on a characterized display. We then examined the workflows to create these reproductions, and found that current workflows rely heavily on visual editing and retouching (global and local color adjustments on the digital reproduction) to improve the color accuracy of the reproduction. Visual editing and retouching can be both time-consuming and subjective in nature (depending on experts' own experience and understanding of the artwork) lowering the efficiency of artwork digitization considerably. We therefore propose to improve the workflow of fine art reproduction by (1) automating the process of visual editing and retouching in current workflows based on RGB acquisition systems and by (2) recovering the spectral reflectance of the painting with off-the-shelf equipment under commonly available lighting conditions. Finally, we studied the perceived image quality of reproductions created by current three-channel (RGB) workflows with those by spectral imaging and those based on an exemplar-based method.

Library of Congress Subject Headings

Art--Reproduction--Quality control; Imaging systems--Image quality; Reflectance--Measurement; Workflow

Publication Date


Document Type


Student Type


Department, Program, or Center

Chester F. Carlson Center for Imaging Science (COS)


Voelkel, Joseph


Note: imported from RIT’s Digital Media Library running on DSpace to RIT Scholar Works. Physical copy available through RIT's The Wallace Library at: N8580 .J42 2013


RIT – Main Campus

Plan Codes