ゼロ交差と付加情報による画像信号の再構成とノイズ低減への応用 : 画像通信システム

概要

論文の詳細を見る
This paper addresses a reconstruction of a image signal with zero-crossings and additional information based on dyadic wavelet transform, and treats the application to the noise reduction. At first, for 1-dimensional signal, the reconstruction algorithm is described. The zero-crossing points are detected by using the multiscale dyadic wavelet trnsform. The signal is reconstructed with the iteration method of the projection onto convex sets. The convex sets used are obtained from the wavelet transform coefficients space and the other space. The other space is constructed with the constraints for the wavelet transform of the given input signal. This reconstruction method is extended to the 2-dimensional signal. Finally, as an application of the algorithm, the noise reduction method based on the appropriate threshold values is presented.
社団法人映像情報メディア学会の論文
1992-10-15