Search

Digital elevation model(DEM) can reflect the topographic characteristics of an area and has a wide range of scientific research applications.Filtering LIDAR point cloud data,extracting the ground points and interpolating are common steps in constructing DEM.The filtering algorithm used in the process of point cloud filtering directly affects the accuracy of the final DEM.As a point cloud filtering algorithm,cloth simulation filtering(CSF) algorithm has the advantages of simple model and high filtering efficiency.It has high filtering accuracy for flat areas.However,when dealing with complex terrain areas,the accuracy of filtering results will be poor due to the internal elasticity and gravity inertia of the cloth model.In view of this,in order to improve the filtering accuracy and terrain adaptability of CSF algorithm in dealing with complex terrain areas,so as to improve the accuracy of constructing DEM,the cloth simulation filtering algorithm with terrain cognition(CSFTC) is proposed.The algorithm proposes a terrain-cognitive model.Based on the local distribution characteristics of point cloud data points,the terrain-cognitive model is constructed and extended to rough digital elevation model(R-DEM),which realizes the separation of macro terrain trend and micro terrain details through point cloud terrain normalization.Finally,the original CSF algorithm combined with R-DEM is used to realize point cloud filtering.Comparison experiment between CSFTC algorithm and the original CSF algorithm is designed.The average total error rate decreases from 9.30% to 5.10%,and the average type-II error rate decreases from 30.02% to 8.46%.Experimental results show that compared with the original CSF algorithm,the accuracy of CSFTC algorithm increases slightly in flat region and increases significantly in complex region,which improves the terrain adaptability of the algorithm.The significant decrease of type-II error is helpful to improve the accuracy of constructed DEM.

Objective The problem of easily falling into "premature" and local optimum values is encountered during endmember extraction via discrete particle swarm optimization (DPSO). Thus, this study introduced the shuffled frog leaping algorithm (SFLA) into DPSO. When SFLA-DPSO (DPSO based on SFLA) was run, the particle swarm was divided into several groups. DPSO was then used for an in-depth search of endmembers, and information was exchanged between groups. SFLA-DPSO combined the universality and parallelism of SFLA with the fast convergence of DPSO. The combined method can prevent the particle swarm from falling in the local optimum. Method SFLA-DPSO, DPSO, and SMACC were applied to extract the endmembers of a hyperspectral RS image in Pulang, Yunnan Province. Then, the probability of global convergence was calculatedby using SFLA-DPSO, DPSO, and N-FINDR to extract the endmembers in the respective search space of the Hyperion and AVIRIS hyperspectral images. Result SFLA-DPSO had higher precision than the other tested methods. The global convergence probability of SFLA-DPSO reached 100% as the number of iterations increased, whereas that of DPSO was only between 50% and 65%. Conclusion SFLA-DPSO showed a much stronger ability to overcome local convergence than DPSO. [ABSTRACT FROM AUTHOR]