A Raster-Based Methodology to Detect Cross-Scale Changes in Water Body Representations Caused by Map Generalization.

Area of Science:

• Geoinformatics
• Remote Sensing
• Cartography

Background:

• Traditional change detection relies on remote sensing images, neglecting cartographic generalization effects in multi-scale maps.
• Detecting changes in multi-scale water systems requires specialized methods beyond standard raster data analysis.

Purpose of the Study:

• To develop a new method for detecting multi-scale water system changes using map tile data.
• To address limitations of traditional remote sensing in capturing cartographic generalization effects.

Main Methods:

• A hierarchical system based on spatial cognition is proposed, with detection levels including body, piece, and slice layers.
• Water area changes are classified using a defined set of indicators and rules for multi-scale raster maps.
• Key technologies for water extraction from tile maps were identified.

Main Results:

• The proposed method was applied to test areas using Tiandi map data.
• Experimental results demonstrated the efficiency and high accuracy of the change detection methodology.
• The hierarchical system effectively detects water area changes across multiple scales.

Conclusions:

• The novel map tile data approach offers a significant improvement for multi-scale water system change detection.
• The hierarchical system and defined rules provide a robust framework for analyzing water dynamics in multi-scale cartographic products.
• This methodology enhances the accuracy and efficiency of detecting spatial changes in water bodies represented in varied map scales.