Robust Parameter Estimation from Point Cloud Data with Noises for Augmented Reality

Allowing for a priori optimization of the robot manipulation to improve the performance in the unmanned environment, it is critical for the augmented reality system to estimate the attitude of point clouds in model reconstruction. The estimation of planar parameter is not always faithful for point cloud fitting, because the gross errors and outliers are not considered in by the traditional plane fitting methods. Our attempt is to combine the random sample consensus and the least-squares method to detect and eliminate the outliers of point clouds for the plane fitting. Optical 3D scanning system is applied for collecting the real data of point cloud from the planar sandpaper. After the preprocessing, the point clouds are imported into OpenGL for the 3D model reconstruction. The modeling process is depicted, too. The comparative simulation experimental results between our method and the traditional methods, such as least-squares and eigenvalue method, are provided. Moreover, our method is applied to the experiment of 3D laser scanning point cloud planes for the attitude estimation. The method is validated on both simulated data and real data. Competitive results show that our method has better robustness and accuracy, especially for the point clouds containing various errors and outliers.