Two-Step CFAR-Based 3D Point Cloud Extraction Method for Circular Scanning Ground-Based Synthetic Aperture Radar

Ground-Based Synthetic Aperture Radar (GBSAR) has non-contact, all-weather, high resolution imaging and microdeformation sensing capabilities, which offers advantages in applications such as building structure monitoring mine slope deformation retrieval. The Circular Scanning (CS-GBSAR) is one of its newest developed working mode, the radar rotates around an axis a vertical plane. Such nonlinear observation geometry brings unique advantage three-dimensional (3D) compared with traditional GBSAR modes. However, causes strong sidelobes SAR images, makes it difficult task to extract point cloud data. Conventional Cell Averaging Constant False Alarm Rate (CA-CFAR) algorithm can 3D data layer-by-layer at different heights, time consuming easily influenced by obtain inaccurate results. To address these problems, this paper proposes new two-step CFAR-based extraction method for CS-GBSAR, accurate under influence sidelobes. It first utilizes maximum projection three-view images from image Then, step CA-CFAR applied coarse masks three-views. volume mask original obtained via inverse projection. This remove outside potential target region area intersecting image. second clouds. Finally, further eliminate residual output clouds, modified Density-Based Spatial Clustering Applications Noise (DBSCAN) clustering applied. DBSCAN uses spherical template cluster. covers more points, sidelobe. Hence, results have noise points. Meanwhile, clusters cylindrical accommodate data’s features, reduce false clustering. proposed validated real acquired North China University Technology (NCUT)-developed CS-GBSAR system. laser detection ranging (LiDAR) are used reference ground truth demonstrate method. comparison experiment conventional shows that 95.4% clustered points sidelobes, better performance