Research On Airborne LiDAR Strip Adjustment Based On3D Surface Matching

As a new technology, airborne LiDAR satisfies the applications of remote sensinginformation sources of our needs. An airborne LiDAR is an integration system of laserscanner, GPS, and INS. Errors will make the positioning offset of laser points,systematical is the major errors. In order to attain high-accuracy airborne LiDAR data,the methods of elimination of systematic errors on points has been to a key technique.Strip adjustment, as the common approach, corrects the bias by using model, which ismade up by simple parametric models.Using matching algorithms to achieve airborne LiDAR system’ adjustment isproposed in this paper.①Analyze the error sources of airborne LiDAR system. Because of integrationsystem, the positioning accuracy of laser footprints is affected by errors. Form the errorequation, this paper analyses the errors, such as the ranging, laser scan angle, GPSpositioning, INS, system integration and others. And the theory of those errors is deeplydiscussed.②The common strip adjustment modes are introduced. Two patterns exist: databased and model based. Typical models are listed and clearly explained in paper.③A variety of point cloud matching technology are organized. Three-dimensionalfree surface matching process is elaborated in detail, and focus on the ICP (IterativeClosest Point), LZD (Least Z-Difference) and TIN (Triangular Irregular Network)algorithms.④Three-dimensional free surface matching is achieved based on KD-treetechnology. The reference plane is created using K-D tree. The seven parameters aresolved by least-squares to apply to the surface, eliminating the influence of errors on theoverlapping region. Experimental results show that: the accuracy of airborne LiDARstrip adjustment based on this technology can meet requirements.