Novel Threshold Self-Regulating Water Extraction Method

Water resources are crucial for human activities and sustainable socioeconomic development. Understanding surface water information can play a key role in resource management, which affects the global cycle ecological environments. Considering Hailar River body as an example, this study proposes new threshold self-learning extraction method (TSLWEM) based on modified normalized difference index (MNDWI) data. The optimal thresholds determined by TSLWEM algorithm four test images were −0.0030, 0, 0.1990, −0.0800. effectively identified target with recognition accuracies of 98.08%, 99.93%, 93.39%, 93.20% images. Moreover, it accurately identify small tributaries, such lakes rivers. is suitable Landsat 8 Operational Land Imager (OLI) data, monitor map complex temperate semiarid regions while improving accuracy identification. study’s findings provide technical support protection well their rational utilization monitoring.