照片三维几何参数估量

Estimating 3D Geometric Parameter in the Picture

针对照片景物参数测定受拍摄角度以及像素影响的不足,提出基于照片景物由三维向二维映射的关系,利用平行交比不变的性质,运行图形辅助工具对照片二维图形进行有效处理,通过MATLAB标定二维长度,得到待测景物的几何参数,对物体轮廓进行拟合。将相机进行自标定,获得照片景物的像素坐标,修正几何参数估量误差。提出SIFT算法模型,匹配照片间的特征,实现三维重建,获得特征点的空间坐标。模型的建立为核设施照片三维几何参数的估量提供了参考方案。

For the deficiency of estimating geometric parameter influenced by the shooting scene and the pixel,the relationship between the image scene and the two-dimensional mapping based on the photo scene is proposed. The graphics tools for effective processing of two-dimensional graphics photos,calibration of two-dimensional length by MATLAB,the geometry of the scene to be measured parameters,fitting the contour of the object,are used. Self-calibration of the camera to get the pixel coordinates of the photo scene,correction of geometric parameters estimation error. The SIFT algorithm model is proposed to match the features of the images and to realize the 3 D reconstruction. The spatial coordinates of the feature points are obtained. The establishment of the model provides a reference for the estimation of 3 D geometric parameters of nuclear facilities photos.