自动3维重构中确定下一最优视点的方法研究
Research on Next Best View in Automatic 3D Reconstruction

提出了基于牛顿万有引力定律矢量场的当前视点下曲面积分最大化的方法,以确定下一最优视点的3维空间位置.在该方法中采用斯托克斯公式计算边界曲线积分的途径间接获得曲面积分,从而方便地确定了视点的位置数据.实验结果表明:该文提出的下一最优视点的确定方法可使机器人自动实现多视点规划,同时解决了单视点规划下存在采集盲区的问题.
Since the classical mass vector chain(MVC)can only solve the determination of the NBV direction,a new approach is proposed,which determines the NBV spatial position through maximizing the surface integral in the current viewpoint based on the vector field of Newton's law of universal gravitation.In the new approach,the Stokes formula is used to calculate the boundary curve integral in place of the surface integral for efficient determination of the NBV position.The experimental results show that the proposed method can help the robot achieve automatic multi-view planning,and can avoid acquisition blind areas which often occur in the single viewpoint planning