Videos, photos, and AI-derived grain size data associated with “Modular AI and video surveys transform multiscale sediment grain size mapping”

This data package is associated with the manuscript in preparation “Modular AI and video surveys transform multiscale sediment grain size mapping”. This data package includes five data types: 1) raw photos and videos from drone survey and walking smartphone surveys; 2) images derived from raw videos; 3) manual labeling of reference scales; 4) metadata for all images and photo resolution derived from artificial intelligence (AI) models or manual labels, and 5) grain size data obtained from AI models for all photos. Such data is used to 1) demonstrate significant improvements in accuracy, efficiency, and quality control for grain size data collection with the help of AI models and 2) study the spatial heterogeneity of grain size and observation reproducibility based on tens of thousands of data points generated by the AI models. In particular, the data package contains 111 folders and 296290 files. The files include 30 videos in .mov format, 105077 photos in .jpg format, 16451 video-derived photos in .png format, 16201 segmentation mask data in .tif format, 16201 segmentation data in .json format, 49769 .csv files that rerecord metadata and grain size for each individual photo, as well as 5 flight record data in .srt format. The summary for all metadata and grain size statistics information are included in “Scales_V3_NG.csv” and “Statistics_V3_NG.csv”. NOTE: The manuscript associated with this data package is currently in review. The data may be revised based on reviewer feedback. Upon manuscript acceptance, this data package will be updated with the final dataset and additional metadata.本数据集关联筹备中的稿件《模块化人工智能与视频勘测技术赋能多尺度沉积物粒度制图》（Modular AI and video surveys transform multiscale sediment grain size mapping）。本数据集包含五类数据：1) 无人机勘测与步行智能手机勘测获取的原始照片及视频；2) 由原始视频提取得到的衍生图像；3) 参考比例尺的人工标注结果；4) 所有图像的元数据，以及通过人工智能（AI）模型或人工标注得到的照片分辨率数据；5) 所有照片经AI模型处理得到的沉积物粒度数据。该数据集可用于两大研究方向：1) 验证借助AI模型开展粒度数据采集时，在精度、效率与质量管控层面的显著提升效果；2) 基于AI模型生成的数万级数据点，探究沉积物粒度的空间异质性与观测重现性。本数据包共计包含111个文件夹与296290个文件，具体文件类型及数量如下：.mov格式视频30个、.jpg格式照片105077张、.png格式视频衍生图像16451张、.tif格式分割掩膜数据16201份、.json格式分割数据16201份、用于记录单张照片元数据与粒度信息的.csv文件49769个，以及.srt格式飞行记录数据5份。所有元数据与粒度统计信息的汇总文件为"Scales_V3_NG.csv"与"Statistics_V3_NG.csv"。注意：本数据集关联的稿件目前处于审稿阶段，数据可能会根据审稿意见进行修订。待稿件录用后，本数据集将更新为最终版本并补充额外元数据。