HoliTracer: Holistic Vectorization of Geographic Objects from Large-Size Remote Sensing Imagery

2025-07-22

AI Summary

To address context loss and geometric fragmentation arising from tiling-based processing in vectorization of large-scale remote sensing imagery (RSI), this paper proposes the first end-to-end, full-image vectorization framework. The method integrates a Context-Aware Attention Network (CAN) for long-range spatial modeling via local-to-global contextual attention, and jointly employs a Mask-to-Contour Refiner (MCR) and a Polygon Sequence Tracker (PST) to directly generate complete, topologically consistent vector polygons for buildings, water bodies, and roads from semantic segmentation outputs. Evaluated on multiple large-scale RSI benchmarks, our approach significantly outperforms existing state-of-the-art methods, achieving substantial improvements in vector completeness, geometric accuracy, and structural coherence. It delivers a scalable, high-fidelity automated solution for production of high-resolution open maps.

Technology Category

• Computer Vision: Remote Sensing / Geospatial AI
• Intelligent Robots: Multimodal Perception & Sensor Fusion
• Knowledge Representation and Reasoning: Geometric, Spatial, and Temporal Reasoning

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Systems and Infrastructure for Web, Mobile and WoT: Location- and context-aware Web and WoT applications and services

Abstract

With the increasing resolution of remote sensing imagery (RSI), large-size RSI has emerged as a vital data source for high-precision vector mapping of geographic objects. Existing methods are typically constrained to processing small image patches, which often leads to the loss of contextual information and produces fragmented vector outputs. To address these, this paper introduces HoliTracer, the first framework designed to holistically extract vectorized geographic objects from large-size RSI. In HoliTracer, we enhance segmentation of large-size RSI using the Context Attention Net (CAN), which employs a local-to-global attention mechanism to capture contextual dependencies. Furthermore, we achieve holistic vectorization through a robust pipeline that leverages the Mask Contour Reformer (MCR) to reconstruct polygons and the Polygon Sequence Tracer (PST) to trace vertices. Extensive experiments on large-size RSI datasets, including buildings, water bodies, and roads, demonstrate that HoliTracer outperforms state-of-the-art methods. Our code and data are available in https://github.com/vvangfaye/HoliTracer.

Problem

• Extracting vectorized geographic objects from large-size remote sensing imagery
• Overcoming fragmentation and loss of context in existing patch-based methods
• Enhancing segmentation and vectorization with attention mechanisms and robust pipelines

Innovation

• Holistic vectorization framework for large-size RSI
• Context Attention Net captures contextual dependencies
• Mask Contour Reformer and Polygon Sequence Tracer