@article{oai:ous.repo.nii.ac.jp:00001046,
 author = {谷口, 香苗 and Taniguchi, Kanae and 澤見, 英男 and Sawami, Hideo},
 journal = {岡山理科大学紀要. A, 自然科学, Bulletin of Okayama University of Science. A, Natural Sciences},
 month = {Mar},
 note = {P(論文), カラー自然画像に関する縮小・拡大画像を得るために変換符号化法を用いた。基準となる方法としては離散コサイン変換(DCT), 提案する方法としては重複ブロック離散コサイン変換(OBDCT)を適用し, 符号化特性と画質の比較, および演算量の比較を行った。符号化特性と画質の比較では, 総合的にOBDCT法を用いる方が優れており, また, 演算量による比較においても, OBDCT法はDCT法に対して演算量が削減されていることがわかった。, It is well known that the discrete cosine tansform (DCT) successfully transforms correlated image data into uncorrelated coefficients. Appling quantization to this DCT coefficients and encoding the quantized coefficients, we are able to obtain compressed image data. We are also able to obtain reconstructed image through decoding, dequantization and inverse transform to compressed image data. Although image quality becomes higher as decreasing compression ratio, it is possible to obtain high quality image at high compression ratio in use. These process is referred as transform coding. Coding efficiency, measure for getting high quality image with high compression ratio, tend to increase as transform length becomes longer in the transform coding. Computational complexity, measure for implementing the method on computer, also becomes larger as transform length becomes longer. Some transform codings are usually considered to be selectable depending on coding efficiency and computational complexity. On the other hand, the transform coding for still image data compression is considered as an approach to image resolution conversion, since it has superiority to another methods on image quality. So, the transform coding is not only used as image data compression but also as resolution conversion, the transform coding must be selected from this viewpoint. In this paper, we provide comparison between two methods, the DCT and the overlapped block discrete cosine transform (OBDCT) methods for coding efficiency, computational complexity and image quality on resolution conversion. The OBDCT is equivalent to the DCT with two-times transform length in coding efficiency, and to the DCT with same transform length in computational complexity. For reconstructed image quality, the OBDCT is used with weighting function suppressing block artifact, the OBDCT method show superiority to the DCT method. We also illustrate that the OBDCT gives significantly good result for enlargement of image.},
 pages = {9--20},
 title = {カラー静止画像の解像度変換について},
 volume = {29},
 year = {1994},
 yomi = {タニグチ, カナエ and サワミ, ヒデオ}
}