階層的なスプライン補間に基づく画像データ圧縮

概要

論文の詳細を見る
Capability of the progressive reproduction is a factor desirable for embedded data compression scheme in image database systems. This paper presents an image data compression scheme based on a hierarchical bicubic spline interpolation and vector quantization. In our method, input images are decomposed hierarchically into a decimated image and some decimated residual images with various decimation ratios, e.g. 8 : 1,4 : 1 and 2 : 1. In order to remove redundancy among samples, an adaptive decimation method is introduced. And then the decimated samples are quantised by the tree search vector quantizers. Simulation experiments are done for the three-stage hierarchical decimation/interpolation and the four-dimensional eight-bit tree search vector quantization.
社団法人映像情報メディア学会の論文
1990-10-24