Objective quality assessment of lossy image compression codecs have becomean important part of the recent call of the JPEG committee for AdvancedImage Coding. The aim of this work is twofold: First, we presentstate-of-the-art still image codecs, JPEG in two variations, a visual andPSNR optimal JPEG2000 version, H.264/AVC and therecently proposed HDPhoto format introduced by Micrsoft. Wemeasure the performance of these codecs by subjective, full-referenceordering tests and two visual image quality metrics on a carefully selectedtest set. Second, we evaluate the prediction quality of the metrics bycomparing them with the outcome of the subjective tests.
Index Terms:
Image Metrics, Objective Evaluation, Subjective Evaluation, JPEG2000, Still Image Compression
Citation:
Thomas Richter, Chaker Larabi, "Subjective and Objective Assesment of Visual Image Quality Metrics and Still Image Codecs," dcc, pp.541, Data Compression Conference (dcc 2008), 2008