著者
辻 正文 鎌田 清一郎 イースン・リチャード 河口 英二
雑誌
情報処理学会研究報告コンピュータビジョンとイメージメディア(CVIM)
巻号頁・発行日
vol.1992, no.7, pp.173-180, 1992-01-23

A camera calibration technique plays an important role in the field of robot vision. There are a lot of techniques to this field for several decades. In this paper we present a method for determining the position of a camera using a few target-points. We use 3 rotation angles and translation vector to describe the position of the camera for a pinhole model. In general for solving the 6 unknown parameters a minimum of six target-points in a world coordinate system is required to uniquely define the matrix. However we show that by using the properties of the matrix we can reduce this number to four. Some experimental results of this method using synthetic data and real image data are also shown.A camera calibration technique plays an important role in the field of robot vision. There are a lot of techniques to this field for several decades. In this paper, we present a method for determining the position of a camera using a few target-points. We use 3 rotation angles and translation vector to describe the position of the camera for a pinhole model. In general, for solving the 6 unknown parameters, a minimum of six target-points in a world coordinate system is required to uniquely define the matrix. However, we show that by using the properties of the matrix we can reduce this number to four. Some experimental results of this method using synthetic data and real image data are also shown.