Counterfeiting ID cards and negotiable securities such as bills and gift certificates are always repeated. To improve performance of detection for authenticity, development of new technology that combines sensor technology and encoding techniques for security information are required. There are many reports to confirm printing regions for security documents, however, methods to detect features of paper media are rarely studied We have developed a method for detecting the authenticity of paper by optically sensing the characteristics of uniformity, brightness, color values, gloss and near-infrared absorption properties. Using multiple features, the separation boundary in a multidimensional feature space is determined. We report results of separation abilities of twenty kinds of paper media based on a multidimensional analysis.