论文部分内容阅读
针对机载激光雷达(light detection and ranging,LiDAR)数据与航空彩色影像的数据特点,提出了一种面向对象的多源数据融合分类方法。该方法根据影像光谱特性将航空影像分割为若干个同质区域,通过综合考察每个区域内LiDAR数据的滤波结果、空间离散度、高差值和航空影像光谱信息,判断各区域归属为哪一类。实验表明,该方法能够有效地分离房屋、树木和裸露地3种基本地物。
Aiming at the characteristics of data of airborne lidar (LiDAR) and aerial color image, an object-oriented multi-source data fusion classification method is proposed. The method divides the aerial image into several homogenous regions according to the spectral characteristics of the image, and judges the belonging of each region by comprehensively examining the filtering results, spatial dispersion, height difference and aerial image spectral information of LiDAR data in each region class. Experiments show that this method can effectively separate the three kinds of basic objects such as houses, trees and bare ground.