3D LiDAR is widely used for mobile robots environment perception nowadays, which has high precision of distance measurement. In this paper, a calibration method which is suitable for Velodyne VLP-16 is firstly introduced. And then we utilize statistical filtering for isolated points removal. After statistical filtering, a slope location and orientation estimation method is presented based on PROSAC algorithm, which can achieve plane extraction from the 3D point cloud data. Experiments are conducted to testify the performance of the presented method.