DTM and orthomosaic of a shallow water area near Lazzareto Nuovo island (Venice Lagoon, Italy) - BRAHMS project

DTM and orthomosaic model of a 6-by-12-m shallow water area (average depth about 1.5 m) along a minor tidal channel branching from Canale Carbonera near Lazzareto Nuovo island, Lagoon of Venice (IT), generated by SfM photogrammetric approach. Photos were acquired on 6 October 2022 by a snorkel diver holding a hydrodynamic board equipped with 4 synchronized action cameras, 70-120 cm from the seafloor. Two pairs of cameras were fixed to the board 70 cm apart, each pair consisting of one camera facing nadir and one inclined approx. 45°. Cameras were set in time-lapse mode (2 fps, wide FoV, ISO 100-400). To cover the area, the diver swimmed back and forth, producing a pattern mostly composed of 12 longitudinal swaths partially overlapping. Reel lines were preliminarily deployed on the seafloor as a guide, together with 5 photogrammetric plates (33x33 cm). For each of 4 vertices and 5 plates, lat/long coordinates were acquired with Trimble SPS585 GNSS receiver (RTK corrections) and z-value as water depth measured with RBR solo3 depth logger. Analyses were perfomed with Agisoft Metashape. Sparse point cloud reconstruction was carried out through incremental alignment procedure, with a total of 3815 photos alligned. DTM (in m, water surface during depth measurements being zero elevation) was created with dense cloud as basis and WGS84/UTM 33N projection. Orthomosaic was created based on DTM and same coordinate system. Scaling and georeferentiation were based on 9 ground control points (4 vertices and 5 plates). The resolution of DTM and orthomosaic are 1.28 mm/pix and 0.32 mm/pix, respectively.

本数据集为意大利威尼斯潟湖拉扎雷托新岛（Lazzareto Nuovo）附近、源自卡博内拉水道（Canale Carbonera）的一条小型潮汐汊道沿岸的6×12米浅水区（平均水深约1.5米）的数字地形模型（DTM, Digital Terrain Model）与正射镶嵌模型（orthomosaic），采用运动恢复结构（SfM, Structure from Motion）摄影测量方法生成。影像采集于2022年10月6日，由一名浮潜潜水员手持搭载4台同步运动相机的水动力浮板完成拍摄，相机距海底70至120厘米。两组相机对以70厘米的间距固定于浮板之上，每对相机分别包含一台天底朝向相机与一台倾角约45°的倾斜相机。相机设置为延时拍摄模式（帧率2fps，宽视场角（FoV, Field of View），ISO感光度100-400）。为覆盖研究区域，潜水员往复巡游，形成以12条部分重叠的纵向条带为主的拍摄路径。预先在海底布设卷轴绳作为导航标识，并同步放置5块33×33厘米的摄影测量标志板。针对4个基准顶点与5块标志板，采用Trimble SPS585全球导航卫星系统（GNSS, Global Navigation Satellite System）接收机（搭载实时动态差分（RTK, Real-time Kinematic）校正服务）获取其经纬度坐标，并通过RBR solo3深度记录仪测量水深以获取Z轴高程值。所有数据分析均采用Agisoft Metashape软件完成。通过增量匹配流程完成稀疏点云重建，共计对齐3815张影像。以密集点云为基础创建数字地形模型（DTM），以水深测量时的水面作为零高程基准，投影坐标系采用WGS84/UTM 33N。正射镶嵌模型基于上述数字地形模型与相同坐标系生成。影像的比例尺校正与地理配准基于9个地面控制点（4个基准顶点与5块标志板）完成。数字地形模型与正射镶嵌模型的分辨率分别为1.28毫米/像素与0.32毫米/像素。