Automatic and High-precise Extraction of Water Information Using a Hierarchical Iterative Method

Remote sensing information extraction is the key step of remote sensing application, and water body is one of its main targets. Thus, the automatic and high-precise extraction of water information from remotely sensed images is of great significance and urgently required in research fields of water resources investigation, natural disaster evaluation, even global ecological environment evolution, etc. Due to the various influencing factors of water, such as suspension, chemistry component, imaging condition and so on, water information extraction methods only using fixed models on whole image are always with low accuracy and aren’t suitable for complex conditions. This paper presents a step-by-step iterative transformation mechanism to extract water information, which uses spatial scale transformation mechanism of “whole-local” based on Normalized Difference Water Index (NDWI) first, and then fuses the hierarchical knowledge of water extraction and achieves the gradually approach of the water body’s optimal margin iteratively by combining the segmentation and classification at whole and local scales respectively. Experiment of plateau lake information extraction demonstrates its higher accuracy and efficiency.