With the continuous development of multiple remote sensing platforms such as aviation, aerospace, and near-Earth space, remote sensing technology has advanced by leaps and bounds in recent years. As a result, the spatial, temporal, and spectral resolution of remote sensing data has been continuously improved, and the amount of data has also increased significantly, making it more and more characteristic of big data. For related research, the emergence of remote sensing big data provides unprecedented opportunities, but it also poses huge challenges. Traditional workstations and servers have been unable to meet the needs of large-scale, multi-scale massive remote sensing data processing.
In order to solve this problem, many global-scale earth science data (especially satellite remote sensing data) online visualization computing and analysis cloud platforms such as Google Earth Engine (GEE) and Aerospace Hongtu's PIE Engine have emerged at home and abroad. Among them, Earth Engine is the most powerful, capable of accessing and synchronizing satellite images such as MODIS, Landsat and Sentinel, and meteorological reanalysis data sets such as NCEP, which are currently commonly used in the remote sensing field. These data are processed. At present, Earth Engine contains more than 900 public datasets, and about 2 PB of data are added every month, with a total capacity of more than 80 PB. As the most advanced remote sensing cloud platform in China, PIE Engine has developed very rapidly in recent years. It has a wealth of domestic satellite data and other important open source data in China, and has unique advantages in data security and access convenience. Compared with traditional image processing tools (such as ENVI), the remote sensing cloud platform provides abundant computing resources on the one hand; on the other hand, its huge cloud storage capacity saves a lot of data download and preprocessing time for researchers.
Nowadays, remote sensing cloud platforms such as GEE/PIE are attracting the attention of more and more domestic and foreign scientific and technological workers with their powerful functions, and the scope of application is also expanding. This course is dedicated to helping scientific researchers master the practical application capabilities of GEE and PIE. Based on the JavaScript programming language, it uses examples to explain the basic concept knowledge of remote sensing cloud, image big data analysis, and advanced skills in classic application cases. In order to improve the teaching quality, this course will integrate the most advanced AI natural language model such as ChatGPT to assist teaching, assist students to answer doubts, provide targeted suggestions and guidance, not only allow students to grasp the course content more deeply, but also provide efficient self-learning personalized learning experience.
Topic 1: Getting to know GEE and PIE remote sensing cloud platform
(1) Introduction of GEE and PIE platforms and typical application cases
(2) JavaScript basics, including variables, operators, arrays, judgment and loop statements, etc.
(3) Important concepts of remote sensing cloud and typical data analysis process
(4) Getting Started with Basic Objects and Platforms of Remote Sensing Cloud
·Images and image collections
·Geometry, features and feature sets
· Date, character, number
· Arrays, lists, dictionaries
Introduction to the most commonly used APIs for main objects such as image/image set, feature/feature set data query, spatio-temporal filtering, visualization, and attribute viewing
Topic 2: Basics of GEE and PIE Image Big Data Processing
(1) Explanation of key knowledge points
·Mathematical operation of image, relationship/condition/Boolean operation, morphological filtering, texture feature extraction, etc.
· Image masking, cropping and mosaicing
· Loop iteration of collection objects (map/iterate)
·Collection object union (Join)
·Image object-oriented analysis
(2) Lectures and drills on main functions
·Landsat/Sentinel-2 image batch cloud removal
·Landsat/Sentinel-2 sensor normalization, vegetation index calculation, etc.
Smoothing and spatial interpolation of time series optical images
Topic 3: Data Integration Reduce
(1) Explanation of key knowledge points
·Integration of images and image collections, such as annual image synthesis in a specified time window
· Image area statistics and field statistics, classification post-processing
· Feature set attribute column statistics
· Mutual conversion between raster and vector
·Group integration and regional statistics
·Linear regression analysis of image sets, images and feature sets
(2) Lectures and drills on main functions
Statistical analysis of the number of available Landsat images and the number of cloud-free observations in the study area
Synthesis of annual NDVI vegetation numbers in China and search for the greenest DOY time of the year
·Analysis of the temporal and spatial variation trend of rainfall on the national scale and 30-year scale
Topic 4: Cloud Data Visualization
(1) Explanation of key knowledge points
· Feature and feature set attribute mapping (bar chart, histogram, stacked column chart, scatter plot, etc.)
·Image mapping (area statistics, classification map, histogram, scatter plot, line graph, pie chart, etc.)
·Image set mapping (sample point time series diagram, regional statistical time series diagram, etc.)
·Array and linked list drawing (scatter plot, sample line diagram, etc.)
· Graphic style and attribute setting
(2) Lectures and drills on main functions
·Phenological analysis and mapping of different surface vegetation based on MODIS time series images
·Analysis of annual forest spatio-temporal changes and drawing of thematic maps based on Hansen products
Topic 5: Data import and export and asset management
(1) Explanation of key knowledge points
· Upload individual assets with different vector data
·Image data upload personal assets, attribute settings, etc.
Batch image export (Asset and Driver)
· Vector data export (Asset and Driver)
· Export of spatial statistical analysis results
(2) Lectures and drills on main functions
·Domestic satellite data download on PIE platform
Batch export and download of image synthesis
·Export of remote sensing index data corresponding to ground samples
Topic 6: Machine Learning Algorithms
(1) Explanation of key knowledge points
· Sample sampling (random sampling, stratified random sampling)
Supervised classification algorithms (random forest, CART, Bayesian, SVM, decision tree, etc.)
Unsupervised classification algorithms (wekaKMeans, wekaLVQ, etc.)
·Classification accuracy evaluation
(2) Lectures and drills on main functions
·Forest dynamic monitoring based on combined optical and radar time series images
Automatic water extraction and flood monitoring
Topic 7: Special Topic Exercises and Review
(1) A comprehensive case of GEE land use classification, realizing the main functions of the series, including ground sample preparation, multi-source remote sensing image preprocessing, algorithm development, classification post-processing, accuracy assessment, spatial statistical analysis and mapping, etc.
(2) Explanation and learning of classic PIE case codes
· Night light index extraction
· Long-term scale vegetation coverage inversion
·Water dynamic monitoring
·Crop planting area extraction
·Desertification degree extraction
(3) Analysis of dynamic changes in population density
·GEE and PIE platform switching, code optimization, common errors and debugging summary
Original Link: GEE-PIE Remote Sensing Big Data Processing and Typical Case Practical Technology Application