Image size conversion on DCT‐compressed data

Abstract

AbstractThe authors propose a method for converting the image size of DCT‐compressed coded data. This method converts the image size using the DCT coefficients as is, without decoding the compressed data (inverse DCT processing). Compared to conventional methods in which scaling is performed in the image space, image quality is equal or superior, and the calculation burden is reduced to one‐half to one‐third. In addition, the order of the DCT for the original compressed data is preserved, and the image can be decoded using common inverse DCT processing even after size conversion. This is a feature not found in conventional scaling methods using DCT coefficients. The authors' method is viable for resolution scalable coding in instances in which data compressed using JPEG or other fixed order DCTs must be compatible and be processed at high speeds. © 2003 Wiley Periodicals, Inc. Syst Comp Jpn, 34(6): 100–109, 2003; Published online in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/scj.1209