能動的制御による三次元パターン認識の一方法

概要

論文の詳細を見る
The 3-D vision system is important for robot control and other machine control. Previous research has generally been based on passive methods, by making assumptions that the edges and other image features can easily be obtained by employing these image operators. However, such is generally not the case, so the aim of this paper is to propose a method of 3D recognition by the active information acquisition of objects. The image information necessary to the pattern recognition and understanding of objects is obtained by a TV camera actively controlling the attitude of objects. Keeping in mind that images are fuzzy themselves. Then, based on these results, the image interpretation processing is carried out more easily by the expert system than by the conventional passive method.
一般社団法人日本機械学会の論文
1988-06-25