我需要将激光雷达数据(3D点云)转换为反射率和距离图(图像),以便将它们用作CNN的通道。
任何已经使用激光雷达数据并可以帮助我的人吗?
提前谢谢您。
答案 0 :(得分:0)
您所指的数据产品通常称为“网格指标”。网格指标是用户定义的网格(例如栅格像元)内计算的LiDAR点云高度和强度数据集的描述性统计数据。我同时使用FUSION
(reference p.72)(一组命令行工具)和lidr
(reference p.29)(R中的LiDAR处理程序包)来生成LiDAR网格度量。
以下是使用lidr
软件包的文档中的示例:
LASfile <- system.file("extdata", "Megaplot.laz", package="lidR")
las = readLAS(LASfile)
colors = height.colors(50)
# Canopy surface model with 4 m^2 cells
metrics = grid_metrics(las, max(Z), 2)
plot(metrics, col = colors)
# Mean height with 400-m^2 cells
metrics = grid_metrics(las, mean(Z), 20)
plot(metrics, col = colors)
# Define your own new metrics
myMetrics = function(z, i)
{
metrics = list(
zwimean = sum(z*i)/sum(i), # Mean elevation weighted by intensities
zimean = mean(z*i), # Mean products of z by intensity
zsqmean = sqrt(mean(z^2)) # Quadratic mean
)
return(metrics)
}
metrics = grid_metrics(las, myMetrics(Z, Intensity))
plot(metrics, col = colors)
plot(metrics, "zwimean", col = colors)
plot(metrics, "zimean", col = colors)
plot(metrics, "zsqmean", col = colors)
这是FUSION
文档中的语法:
GridMetrics [switches] groundfile heightbreak cellsize outputfile datafile1 [datafile2 ...
datafileN]