プロジェクタ・カメラシステムにおける射影変換行列と直線パターン検出に基づく複数平面の検出

概要

論文の詳細を見る
Because of their abundance and simplicity, planes are used in several computer vision tasks such as self-calibration, reconstruction, visual measurement, and obstacle detection. We present a multiple-plane detection algorithm based on homography and projected line detection for use in an uncalibrated projector-camera system. Detection of the projected line from dense corresponding points roughly segments a camera image into multiple planar regions. The homography is calculated using the robust estimator in each plane and outliers are recursively segmented into new planar regions. Lastly, coplanar regions are integrated using re-porjection errors in the recovered planar homographies. Results with real images are shown to illustrate the efficiency of the proposed method.
社団法人映像情報メディア学会の論文
2007-01-01