This session is a guide for satellite image analysis with R programming. Here we will discuss about different R programming packages mostly used for manipulating Raster and vector data. You can find different sources to learn Those stuffs, Here I will guide you to learn each in quick way.
Overview to the content:
· Guide to R programming language.
· Packages for raster and vector data manipulation.
· Packages used for Machine Learning.
· Some websites to learn Spatial analysis with R
Guide to R programming Language:
R is a high-level programming language for statistical analysis. Out of other programming language like python, it is very easy to learn and data analysis.
To install and use R programming you need two programs or software in your machine. First you have to download R programming language from CRAN and R Studio which is a IDE for R programming which make your package management and coding easy.
Here is the links to learn basic operations with R programming.
Packages for raster and vector data manipulation:
When I started satellite image analysis with programming, R was my favourite programming language; Not only for it is easy to learn, here I can perform most of the analysis directly on raster. The most essential packages are rgdal, rgeos, sp and raster with those packages you can easily handle both of raster and vector data. You can follow this website to learn most of the image analysis with R programming.
R packages for Machine Learning and Statistical Analysis:
Machine Learning is a part of Artificial Intelligence which makes our life easy by deploying machines to do different tasks. Basically, it uses different statistical algorithms to learn from previous tasks and their results to perform a new task. In Spatial analysis ML is used for Land Use classification or predict a phenomenon with time. ML is works in 2 different way (Supervised and Unsupervised) with various statistical Methods (LR, MLR, RF, NN etc.). Here is a brief introduction to ML, you can learn more here.
Various frameworks are available in R programming to perform ML for different works.
Google’s TensorFlow, H2o, Caret, those are mostly used frameworks for ML, but when you are working with Satellite imageries and you are not highly familiar with programming use Caret it is simple then other frameworks. When you are going to perform a LULC classification you need a large dataset which contained the LULC class info and corresponding Band wise pixel values. To extract this you can use a package ExtractTrainData which would be helpful for you.
You also can find different statistical analysis packages for your analysis like:
Stats, e1071, matrics, SPIE, ARIMA, Tidyverse etc.
Some websites to learn Spatial analysis with R:
* If you have any questions regarding this topic comment me bellow.