Research on River and Lake-Related Construction Project Survey and “Clean-up of Four Types of Illegal Activities” Supervision in Xinjiang Based on Remote Sensing Technology

Large-scale monitoring of river–lake corridors in arid regions requires efficient acquisition and analysis of high-resolution geospatial information. This study proposes a remote sensing–based supervision framework for detecting river-related construction activities and illegal disturbances using high-resolution satellite imagery and object-based image analysis. Multi-dimensional spectral, textural, geometric, and topographic features are integrated to improve the identification of engineering structures and disturbance activities under complex dryland conditions. A closed-loop workflow combining remote sensing detection, administrative information matching, risk assessment, and field verification is established to support large-scale supervision and dynamic tracking. Application to 679 rivers and 46 lakes in Xinjiang identified 10,377 river-related objects and demonstrated reliable detection performance under heterogeneous environmental conditions. The proposed framework enhances the capability of large-area environmental sensing, spatiotemporal information extraction, and intelligent monitoring. The study provides a practical methodology for high-resolution Earth observation applications and data-driven supervision in complex environmental monitoring scenarios.