无人机遥感影像中行道树信息快速提取

Rapid extraction of rural house information from UAV remote-sensing images

行道树具有补充氧气、净化空气、美化城市、减少噪音等作用。虽然高分辨率遥感技术能够提取行道树信息,但其工作量大、成本高。无人机具有飞行成本低、数据分辨率高、外业周期短、机动灵活等优点,根据行道树分布灵活设计航线,可显著降低成本提高效率。文章利用无人机航测遥感系统采集并制作的DEM、DSM与DOM,基于邻域最高过滤法提取了行道树信息,并将提取结果与实地调查数据进行了比较分析。结果表明,邻域最高过滤法测量行道树单株树高的平均误差为4.94,株数提取率也达到95%以上。

The roadside trees have the effect of supplement oxygen, purify the air, beautify the city,reduce the noiseand so on. Although high-resolution remote sensing technology can extract the information of roadside trees, its workload is very big and the cost is high. UAV flight with the advantage of low cost, high resolution data,shortfield cycle flexible, etc., according to the distribution of roadside trees flexible design its routes, it can significantly reduce costs and improve efficiency. This paper uses survey and remote sensing system of unmanned aerial vehicles to collect and produce DEM, DSM and DOM. Information about border trees is extracted on basis of maximum neighborhood filtration method. The extraction outcome and on-site investigation data are compared and analyzed. The outcome indicates the maximum neighborhood filtration method registers an average error of 4.94 in measuring information about the tree height and achieves an extraction rate of above 95% in collecting information about the number of border trees.