抄録
According to increase frequency in use of computer simulation, a methodology that classifies a large amount of computational results in a database and searches a certain data set has been needed. To classify and search a result of simulation, it is necessary to evaluate the similarity between a certain data and a reference. A similarity estimation method which employs "Critical Point Graph (CPG)" as an index of data is effective in evaluating. However, this method proposed in past time can not allow to transformation of the data such as rotation and scaling. In this paper, we propose a similarity estimation method which allows to affine transformation with using CPG method about two and three dimensional scalar data sets (volume data sets). In our method, we normalized data sets by principal component analysis. And we estimated a similarity between two volume data sets. From the result, one can safely state that this method allow to affine transformation.
