画像処理と3次元CAD図形情報による形状・面情報を用いた認識手法 : 移動可能なカメラを用いた場合(機械力学,計測,自動制御)

概要

論文の詳細を見る
In this paper, we propose a recognition method combining three-dimensional CAD information with image processing. In this recognition method, images are made from three-dimensional CAD figures by rendering, which is processed using the practical position of a light source and reflectance of an object. Then the object is recognized by comparing input images taken from a CCD camera with the images, which are made from three-dimensional CAD information. In particular, this recognition method can recognize similar objects, which have different shapes of holes in the same position and different surface curvatures, because it compares the surface information of the object with that of the three-dimensional CAD figure.
社団法人日本機械学会の論文
2005-07-25