Sitemap
A list of all the posts and pages found on the site. For you robots out there is an XML version available for digesting as well.
Pages
A flexible two-column Jekyll theme. Perfect for personal sites, blogs, and portfolios hosted on GitHub or your own server. Latest release v4.9.1
Remote sensing: Linking LiDAR and satellite data with machine learning
Handle up-to-date remote sensing data, make spatial predictions with machine learning, and become familiar with advanced remote sensing modelling in R
Splash Page
Bacon ipsum dolor sit amet salami ham hock ham, hamburger corned beef short ribs kielbasa biltong t-bone drumstick tri-tip tail sirloin pork chop.
Posts
examples
unit01
The very basics
Go through a brute force introduction into R, R Markdown and the RStudio IDE to get ready for solving the upcoming assignment problems and submitting your so...
R and RStudio
To start with a clarification: R is the statistical programming language you will use in this course (and which is used by many other scientists). With R yo...
Example: Vector Basics
Vectors are the basis for many data types in R. Creating a vector A vector is created using the c function. Here are some examples: my_vector_1 <- c(1,2...
Example: Data Frame Basics
Data frames are one of the most heavily used data structures in R. Creation of a data frame A data frame is created from scratch by supplying vectors to the...
Example: R Markdown with html output
This page shows how a compiled R markdown file looks like (in fact, all code examples in this course were compiled with R markdown). This is a header This ...
Project management and working environment
A note on data management Within this course, we will use a variety of datatypes, produce R output and save some (temporary) results, plots and markdown file...
Assignment: Hello R
This worksheet introduces you to R, R scripts and R markdown. After completion you should have gained some experience in running R commands within scripts, R...
unit02
More basics
Become familiar with reading and writing data, computing summary statistics and visual data exploration. Get to know the relation between variables and build...
Tabulated Data I/O
Reading or writing tabulated data into or from a data frame is a quite common task in data analysis. You could use the read.table function for this. df <-...
Example: CSV I/O
Reading data from csv files Reading csv files is realized using the read.table function from R’s utils library. The function will return a data frame which ...
Example: Aggregation Statistics
Summarizing a data set The most straight forward function which returns some aggregated statistical information about a data set is summary. a <- c("A",...
Example: Visual Data Exploration
Visual data exploration should be one of the first steps in data analysis. In fact, it should start right after reading a data set. The following examples ar...
Example: Missing Values
Handling missing values is straight forward. Let’s start with a vector with one NA value at position 3. Please note that NA is not inside quotation marks sin...
Example: Date/Time
Coercing data types to date and/or time information is generally performed using as.Date or either as.POSIXct or as.POSIXlt. Let’s start with as.Date: as.Da...
Example: Sorting
Sorting vectors or lists Vectors can be sorted using the sort function. If you want to sort a list, you have to access the actual elements since sort require...
Example: Merging
When thinking about combining two data frames one has to distinguish between merging them by the values given in a specific column or consecutively putting t...
Practice your statistics
Do the following tasks on your own. Write down questions or difficulties you encountered which we can discuss afterwards. This will be a refresher for basic ...
Homework: Prepare R Spatial
If you have not done so already, get yourself an account for the Sentinel Hub of the ESA. You can register here for free. You will need this account ...
unit03
Spatial Data
Get a brute-force introduction to working with different kinds of spatial data in R.
Vector Data
Vector data consists of potentially linked points defined by coordinates that can form complex geometries with assigned attributes. Vector data model In a ...
Raster Data
Raster data consists of cells or pixels in a seamless grid system and usually contain only one value per cell or pixel. Raster data model Raster data is co...
Coordinate Reference Systems
An integral part of all spatial data is the coordinate reference system (CRS), which is used to define meaningful distances between data points irrespective ...
Example: Workflow in R
This example provides a schematic workflow for processing vector and raster data in R. Get raster data Firstly, we import some raster data into our working...
FAQ: sp vs. sf
Which R package should I learn to use? Short answer: currently both! Long answer: here Where can I find additional ressources about sf? The Introduction t...
Exercise: Spatial Data Processing
This exercise will lead you through some of the most basic and important steps for spatial data processing. Things you need for this worksheet R — the in...
unit04
Introduction to Remote Sensing
Learn the basics of satellite remote sensing theory and Sentinel-2 data.
Sensor Types
Active vs. Passive Conservation Technology Series Issue 4: SATELLITE REMOTE SENSING FOR CONSERVATION Passive Sensors The most common remote sensing...
Sensor Platforms
Footprint and Spatial Resolution The three most common satellite platforms for multispectral imagery are MODIS (Moderate Resolution Imaging Spectroradiomete...
The Light Spectrum or why does Remote Sensing work?
For an understanding of why multispectral remote sensing works, it is a good idea to recap some physics (mainly optics) and how light and color work. Light,...
Sentinel-2
What is Sentinel-2 Sentinel-2 is a pair of satellite platforms with multispectral sensor instruments (abbreviated MSI) on board. They cover almost the entir...
Learning by Doing: Sentinel-2 Preparation
Data acquisition and first steps Download the Sentinel-2 scene for Marburg from 23rd March 2020. Make sure you get the L2A dataset. Create a new R proj...
Learning by Doing: Sentinel-2 Application
In this exercise we want to estimate the area of water bodies in our Sentinel scene from Lahntal. As a start, load in your prepared Sentinel stack from Lahnt...
unit05
Reading: Introduction to LiDAR Remote Sensing
Learn the basic of Lidar remote sensing.
Example: LidR I/O and clipping
Since LiDAR data can have quit large file sizes and need some processing power, we only take a look at a small example area. This area was created with the s...
Exercise: Introduction to LidR
Data acquisition Please download “Lidar Data Lahntal” from the Ilias course directory “data” and save it in your project structure. It may be advisable to c...
Exercise: LiDAR products
DEM, DSM, CHM In this exersice we want to calculate three of the most common LiDAR products: the Digital Elevation Model, the Digital Surface Model, and the...
unit06
What is machine learning?
Learn the fundamentals of machine learning and build a random forest model.
Exersice: Data preparation
In this exersice you prepare your mean vegetation height data from LiDAR and the Sentinel bands. Create a rasterstack containing the Sentinel-2 Bands 02,...
The Modelling Workflow
Predictors and Response. Source: Alice Ziegler Converting the rasterstack to a datafram...
Exersice: Data visualisation
Before you build your model, it is a good idea to get a feeling about your data with some visualizations. Plot your whole dataframe. What can you learn f...
Exersice: My first model
Install the caret package Install the ranger package Install the tree package Decision trees Split your data into a training and testing set wit...
Exersice: Hyperparameter tuning
Models in machine learning usually have multiple parameters which decide how the model looks like. A simple example is the slope and intercept of a linar mod...
Exercise: Validation
Visual validation Use the Sentinel-2 data to predict the mean vegetation height for the entire Lahntal. Save the prediction as a tif file and have a lo...
Example: Putting it all together
Predictors and observations Example workflow: library(raster) library(ranger) library(caret) library(RStoolbox) library(viridis) # Prepared Sentinel 2 S...
A word about classifications
In this course, we focused on and made use of regression models. We used Random Forest to predict a continuous variable (the mean vegetation height) and also...
unit07
Projects
Apply your gained skills in your own work project.