Chapter 5 - Remote sensing data extraction and inversion techniques: A review

Abstract

Data extraction from high resolution multispectral satellites imageries can be carried out with the help of different classification techniques. Furthermore, techniques indispensable for information extraction are elements of visual image interpretation which is essential for specific features identification, band/spectral ratioing increases separability of features with minute spectral variation and reduces brightness variation, principal component analysis (PCA) confines necessary information in few bands and reduces band combination ratios. Spectral/band ratioing and PCA are commonly used for efficient represent information. Remote sensing data is organized in the form of row and columns, also known as raster data. Depending upon the arrangement of pixel, there are three types of file format band interleaved by line (BIL), band interleaved by pixel (BIP), and band sequential (BSQ). Other file formats that are commonly used and self-describing as well as machine independent are geographical tagged image file format (GeoTIFF), network common data form (netCDF), and hierarchical data format (HDF). Digital images recorded by sensors onboard satellite platforms contains errors which may be related to geometry and brightness values of the pixels. Image rectification is process in which removes planimetric distortion for accurate area, distance, and direction measurements. The rectification of satellite images is achieved with the help of geometric transformation and resampling techniques. The pixel values of multispectral satellite images are influenced by sensor specific variations and dominant atmospheric condition which hampers the radiance and reflectance measurements. Radiometric correction is a process of transformation of digital number (DN) in absolute radiance value.