主観的評価尺度に基づく画像の復元法

概要

論文の詳細を見る
This paper describes a new method for the restoration of a noise-blurred image. Firstly, a given image, which is blurred by noise, is estimated by person according to certain estimate terms using the fuzzy measure. A set of the estimated values is regarded as a matrix. A membership function is obtained by eigenvectors, corresponding to the maximum eigenvalue of the matrix. A structure of the total estimation for the image is characterised by the membership function. Secondly, power spectrum of the noise in the blurred image is obtained by using fuzzy inference and the membership function. The power spectrum is applied to the Parametric Wiener Filter and the restoration image is obtained from the noise-blurred image. The restored image is estimated by using histogram and the similarity of the image objectively, and a good result is obtained.
東海大学の論文