PROBLEM TO BE SOLVED: To provide a system for analyzing and evaluating the degree of attention of published information on a printed article to be an object for evaluation or a commodity on a showcase by estimating and pursuing the eye line of a subject to be tested.SOLUTION: This eye-line analyzer includes: eye-line pursuing eyeglasses for transmitting a scene image seen through the eyeglasses and coordinates of a point of attention on the scene image to the eye-line analyzer a storing means for storing the coordinates of characteristic points and characteristic quantities of a reference image and divided images of the reference image a means for receiving the scene image and the coordinates of the point of attention from the eye-line pursuing eyeglasses a means for combining the characteristic points with the highest evaluation on the degree of similarity based on the characteristic quantity of the characteristic point of the scene image and the characteristic quantity of the characteristic point of the divided image a means for calculating a coordinate transforming matrix based on the coordinates of the characteristic points of the divided image having the largest number of combination and the coordinates of the characteristic point of the scene image and a means for synthesizing a mark for the point of attention by transforming the coordinates of the point of attention on the divided image calculated from the coordinate transforming matrix and the coordinate of point of attention on the scene image to the reference image. A eye-line measuring system characterized by the eye-line analyzer is also provided.COPYRIGHT: (C)2013,JPO&INPIT【課題】本発明の課題は、被験者の視線を推定・追跡することで、評価対象印刷物の掲載情報または陳列棚の商品の注目度を分析評価するシステムを提供することである。【解決手段】メガネを通して見るシーン画像とシーン画像上の注視点座標を視線分析装置に送信する視線追跡メガネと、基準画像および基準画像の分割画像の、特徴点の座標と特徴量を記憶する手段と、視線追跡メガネからシーン画像と注視点座標を受信する手段と、シーン画像の特徴点の特徴量と分割画像の特徴点の特徴量とに基づき最も類似度評価の高い特徴点を組み合せる手段と、最も多い組合せ個数を持つ分割画像の特徴点の座標とシーン画像の特徴点の座標に基づいて座標変換行列を算出する手段と、座標変換行列とシーン画像上の注視点座標から算出した分割画像
