terraceDL: A geomorpholgy deep learning dataset of agricultural terraces in Iowa, USA
收藏资源简介：
<strong>scripts.zip</strong> <br> <strong>arcgisTools.atbx:</strong> <strong>terrainDerivatives</strong>: make terrain derivatives from digital terrain model (Band 1 = TPI (50 m radius circle), Band 2 = square root of slope, Band 3 = TPI (annulus), Band 4 = hillshade, Band 5 = multidirectional hillshades, Band 6 = slopeshade). <strong>rasterizeFeatures</strong>: convert vector polygons to raster masks (1 = feature, 0 = background). <br> <strong>makeChips.R</strong>: R function to break terrain derivatives and chips into image chips of a defined size. <strong>makeTerrainDerivatives.R</strong>: R function to generated 6-band terrain derivatives from digital terrain data (same as ArcGIS Pro tool). <strong>merge_logs.R</strong>: R script to merge training logs into a single file. <strong>predictToExtents.ipynb</strong>: Python notebook to use trained model to predict to new data. <strong>trainExperiments.ipynb</strong>: Python notebook used to train semantic segmentation models using PyTorch and the Segmentation Models package. <strong>assessmentExperiments.ipynb</strong>: Python code to generate assessment metrics using PyTorch and the torchmetrics library. <strong>graphs_results.R</strong>: R code to make graphs with ggplot2 to summarize results. <strong>makeChipsList.R</strong>: R code to generate lists of chips in a directory. <strong>makeMasks.R</strong>: R function to make raster masks from vector data (same as rasterizeFeatures ArcGIS Pro tool). <br> <strong>terraceDL.zip</strong> <br> <strong>dems</strong>: LiDAR DTM data partitioned into training, testing, and validation datasets based on HUC8 watershed boundaries. Original DTM data were provided by the Iowa BMP mapping project: https://www.gis.iastate.edu/BMPs. <strong>extents</strong>: extents of the training, testing, and validation areas as defined by HUC 8 watershed boundaries. <strong>vectors</strong>: vector features representing agricultural terraces and partitioned into separate training, testing, and validation datasets. Original digitized features were provided by the Iowa BMP Mapping Project: https://www.gis.iastate.edu/BMPs<strong>. </strong>
<strong>scripts.zip</strong><br><strong>arcgisTools.atbx工具箱：</strong><strong>terrainDerivatives</strong>：基于数字地形模型（DTM）生成各类地形导数产品，各波段参数如下：波段1为50米半径圆形邻域的地形位置指数（TPI），波段2为坡度平方根，波段3为环形邻域地形位置指数（TPI），波段4为山体阴影，波段5为多方向山体阴影，波段6为坡度阴影。<strong>rasterizeFeatures</strong>：将矢量多边形转换为栅格掩膜，其中要素区域赋值为1，背景区域赋值为0。<br><strong>makeChips.R</strong>：用于将地形导数数据与现有图像样本块裁剪为指定尺寸图像样本块的R语言函数。<strong>makeTerrainDerivatives.R</strong>：可从数字地形数据生成6波段地形导数产品的R语言函数，功能与ArcGIS Pro的terrainDerivatives工具一致。<strong>merge_logs.R</strong>：用于将多条训练日志合并为单个文件的R语言脚本。<strong>predictToExtents.ipynb</strong>：用于调用已训练模型对新数据进行预测推理的Python交互式笔记本。<strong>trainExperiments.ipynb</strong>：基于PyTorch与分割模型（Segmentation Models）包训练语义分割模型的Python交互式笔记本。<strong>assessmentExperiments.ipynb</strong>：基于PyTorch与torchmetrics库生成模型评估指标的Python代码。<strong>graphs_results.R</strong>：依托ggplot2绘图包生成结果可视化图表以汇总实验结果的R语言代码。<strong>makeChipsList.R</strong>：用于生成指定目录下图像样本块列表的R语言代码。<strong>makeMasks.R</strong>：可从矢量数据生成栅格掩膜的R语言函数，功能与ArcGIS Pro的rasterizeFeatures工具一致。<br><strong>terraceDL.zip</strong><br><strong>dems</strong>：基于8级水文单元流域（HUC8）边界划分为训练集、测试集与验证集的激光雷达（LiDAR）数字地形模型（DTM）数据。原始DTM数据由爱荷华州最佳管理措施（BMP）制图项目提供，来源网址：https://www.gis.iastate.edu/BMPs。<br><strong>extents</strong>：由8级水文单元流域（HUC8）边界定义的训练、测试与验证区域的空间范围。<br><strong>vectors</strong>：代表农业梯田的矢量要素，已按8级水文单元流域（HUC8）边界划分为独立的训练集、测试集与验证集。原始数字化要素由爱荷华州BMP制图项目提供，来源网址：https://www.gis.iastate.edu/BMPs。
