基于高分辨率影像的桂林市多时序矢量道路网络数据集（2010、2015、2020、2024年）

本数据产品是基于2010、2015及2020年的桂林市（秀峰、叠彩、象山、七星、雁山、临桂六区）高分遥感影像，通过深度学习道路提取模型处理生成二值图，并矢量化后得到。包含道路空间位置、长度、宽度等信息。基于ENVI与ArcGIS平台生成。处理流程如下：首先对多光谱与全色影像进行融合；随后从融合后的四波段影像中提取RGB三波段，并将其转换为8位格式；继而采用基于深度学习的方法提取道路信息；最后，将得到的二值化道路结果进行矢量化，并计算各条道路的长度与宽度，从而生成最终的道路数据产品。

This road data product is derived from high-resolution remote sensing images of six districts (Xiufeng, Diecai, Xiangshan, Qixing, Yanshan and Lingui Districts) of Guilin City in 2010, 2015 and 2020. The images were processed using deep learning-based road extraction models to generate binary maps, which were then vectorized to obtain the final product. It contains information such as the spatial location, length and width of roads, and was developed based on the ENVI and ArcGIS platforms.

The specific processing workflow is as follows:
1. First, fuse the multispectral and panchromatic remote sensing images;
2. Then, extract the RGB three bands from the fused four-band image and convert it to an 8-bit format;
3. Next, extract road information using deep learning-based methods;
4. Finally, vectorize the obtained binarized road extraction results, calculate the length and width of each individual road, and generate the final road data product.

搜集汇总

背景与挑战

背景概述

该数据集是基于高分辨率遥感影像生成的桂林市主城区多时序矢量道路网络数据，覆盖2010、2015、2020和2024年四个时间点，通过深度学习模型提取道路信息并矢量化为包含空间位置、长度和宽度等属性的矢量数据，适用于城市地理和遥感应用研究。