A revolutionary multi-dimensional data format for remote sensing

Remote sensing data formats are essential for storing, organizing, and managing imagery collected by satellites and sensors. These formats store remote sensing images and their related information, such as geographic coordinates and band information. It specifies the data storage order, encoding method, header file (which includes the basic information of the image, including the number of rows, columns, bands, and data types), and the organization of the data body.

Traditional remote sensing data formats

Traditional remote sensing datasets commonly use GeoTIFF (tagged image file format) and HDF (hierarchical data format) as their core storage units. TIFF is a widely used universal image storage format, particularly in the field of remote sensing. For instance, GeoTIFF, an extension of TIFF that incorporates geospatial metadata, enables the embedding of geographic coordinate information with image files, thereby facilitating geospatial analysis. NetCDF build on HDF, is a scientific data format specifically designed for handling multi-dimensional and multi-variable array data, making it highly effective for storing remote sensing data. It supports data compression, subset extraction, and various other operations, enhancing data management and analysis capabilities. These formats often rely on complex data structures, such as trees or linked lists, to organize information. However, because most standard remote sensing data formats are built on a three-dimensional cube model, they can be difficult to use for analyzing data over long time periods. The multi-dimensional data (MDD) storage format emerged in this context. Figure 1D summarizes the storage efficiency, dimensionality, and efficiency of traditional formats and MDD.