In recent years, remote sensing technology has developed by leaps and bounds. Multiple remote sensing platforms such as aerospace, aviation, and near space have continued to increase. The spatial, temporal, and spectral resolutions of data have continued to improve. The amount of data has soared. Remote sensing data has become more and more characterized by big data. . The emergence of remote sensing big data provides unprecedented opportunities for related research. At the same time, how to handle these data also poses huge challenges. Traditional workstations and servers are no longer capable of processing large-area, multi-scale massive remote sensing data.
https://mp.weixin.qq.com/s?__biz=Mzg2NDYxNjMyNA==&mid=2247539613&idx=6&sn=838c5feae0c7a3e143cc9d7a3ca590e8&chksm=ce64b176f91338604d3d9615773e8bd7 ca88025ea8afb5bffb67fe83e1b848ebe145933c5c2a&token=1648264366&lang=zh_CN#rdOnline visual computing and analysis cloud platforms for global-scale earth science data (especially satellite remote sensing data) represented by Earth Engine (GEE) and PIE-Engine are becoming more and more widely used. The GEE platform stores and synchronizes more than 80PB of satellite images, climate and weather, geophysics and other data sets commonly used in the field of remote sensing such as MODIS, Landsat and Sentinel. At the same time, relying on millions of super servers around the world, it provides sufficient computing power to process these data. The data is processed. Compared with traditional remote sensing image processing tools such as ENVI, GEE has incomparable advantages in processing massive remote sensing data. On the one hand, it provides abundant computing resources, and on the other hand, its huge cloud storage saves researchers a lot of data downloads. and preprocessing time, it represents the world's most advanced level in the field of calculation, analysis and visualization of remote sensing data, and is a revolution in the field of remote sensing.
In the early stage, we used Javascript and Python as programming languages respectively, and received extensive participation and support from students from multiple industries. At the request of the majority of scientific researchers, we will focus on forestry, one of the most popular fields of remote sensing applications at present, and comprehensively demonstrate the usage skills and powerful functions of the GEE cloud platform with typical application cases to enhance the participants' ability to solve practical problems. The explanation will be based on the JavaScript version of GEE. First, the basic knowledge of GEE will be introduced, and then the key knowledge will be explained in combination with micro-cases. Finally, a comprehensive explanation will be given based on typical forestry application cases. In order to improve the quality of teaching, the most advanced AI natural language models such as ChatGPT and Wen Xinyiyan will be used to assist teaching to help students answer their doubts and provide targeted suggestions and guidance. This will not only allow students to have a deeper grasp of the course content, but also prepare them for future self-help Learn provides an efficient and personalized learning experience.
Details:
Part One: Basic Practice
1. Platform and basic development platform
Introduction to GEE platform and typical application cases;
Introduction to GEE development environment and common data resources;
Introduction to GPT models such as ChatGPT and Wen Xinyiyan, account application and forestry remote sensing applications
Introduction to JavaScript basics;
Important concepts and typical data analysis processes of GEE remote sensing cloud;
Introduction to GEE basic objects, vector and raster object visualization, attribute viewing, API query, basic debugging and other platforms to get started.
2. Interaction between GEE basic knowledge and AI models such as ChatGPT
Basic image operations and operations: mathematical operations, relational/conditional/Boolean operations, morphological filtering, texture feature extraction; image masking, cropping and mosaic, etc.;
Basic calculations and operations on elements: geometric buffer, intersection, union, difference operations, etc.;
Collection object operations: loop iteration (map/iterate), merge, union (Join);
Data integration Reduce: including image and image set integration, image synthesis, image area statistics and domain statistics, group integration and area neighborhood statistics, image set linear regression analysis, etc.;
Machine learning algorithms: including supervised (random forest, CART, SVM, decision tree, etc.) and unsupervised (wekaKMeans, wekaLVQ, etc.) classification algorithms, classification accuracy evaluation, etc.;
Data asset management: including local vector and raster data upload, cloud vector and raster data download, statistical result data export, etc.;
Drawing visualization: including bar charts, histograms, scatter plots, time series and other graphic drawings.
GPT model interaction: Combine the above basic knowledge points with AI tools such as ChatGPT and Wen Xinyiyan to conduct interactive demonstrations, including auxiliary Q&A, code generation and correction techniques.
Part 2: Micro-case lecture on important knowledge points and interactive demonstration of GPT model
1) Automatically remove clouds and shadows from Landsat and Sentinel-2 images in batches
2) Combine Landsat and Sentinel-2 to calculate vegetation index and annual synthesis in batches
3) Statistical analysis of the number of available images and cloud-free observations in the study area
4) Annual NDVI vegetation number synthesis and annual greenest DOY time search in China
5) Moving window smoothing of time series optical image data
6) Stratified random sampling and sample export, sample local evaluation and data upload to the cloud
7) Analysis of rainfall change trends in China in the past 40 years
8) Statistical analysis of annual forest loss in a certain region (based on Hansen Forest Products)
Part Three: Comprehensive Exercises on Typical Cases
Case 1: Forest identification by combining multi-source remote sensing data
This article introduces in detail the complete process of combining Landsat time series optical images and PALSAR-2 radar data, as well as the decision tree algorithm to achieve remote sensing classification of typical land types such as forests. Topics include spatiotemporal filtering of image data, optical image batch cloud masking and vegetation index calculation; stratified random sampling and sample export, local quality control and cloud upload, random sample segmentation, separability analysis, classification algorithm construction and application, Classification post-processing and accuracy assessment, thematic map drawing, etc.
Case 2: Long-term forest status monitoring
Use long-term series of MODIS or Landsat image data to conduct long-term monitoring of forest status and analyze the greening or browning of forest vegetation. Topics include time series image preprocessing, image set connection, image synthesis, non-parametric detection of change trends, significance testing, quantification and classification of change trends, spatial statistics and result visualization, and thematic map drawing.
Case Three: Deforestation and Degradation Monitoring
Combining Landsat series images, spectral separation model and NDFI normalized difference fraction index to monitor forest deforestation and degradation. Topics include image preprocessing, mixed pixel decomposition, NDFI index calculation, function encapsulation, change detection and intensity classification, result visualization, thematic map drawing, etc.
Case 4: Forest fire monitoring
This article introduces in detail the use of Landsat and Sentinel-2 time series optical remote sensing images to monitor forest fire losses and achieve fire intensity classification. Topics include image filtering, pre-processing such as Landsat and Sentinel-2 optical image cloud removal, vegetation index calculation, image synthesis, fire area identification and disaster intensity classification, statistical analysis and visualization of results, etc.
Case 5: Long-term forest disturbance monitoring
Combining 30 years of optical images such as Landsat and the classic LandTrendr algorithm to realize forest disturbance monitoring. Topics include long-term series remote sensing image preprocessing, vegetation index batch calculation, annual image synthesis, array image concepts and usage methods, LandTrendr algorithm principles and parameter settings, forest disturbance result interpretation and spatial statistical analysis, visualization and thematic map drawing, etc.
Case 6: Key physiological parameters of forest (tree height, biomass/carbon storage)
The inversion combines GEDI lidar, Landsat/Sentinel-2 multispectral optical images, Sentinel-1/PALSAR-2 radar images, etc. and machine learning algorithms to invert key physical parameters of the forest, such as tree height, biomass/carbon storage. Topics include introduction to GEDI lidar data, common optical and radar data processing, machine learning algorithm applications, inversion accuracy evaluation and variable importance analysis, result visualization, etc.
