Detection of mining relics (Pingen) using LiDAR technology

From the middle of the 17th to the beginning of the 19th century, tin mining was carried out near Bockau and Aue in the Westerzgebirge. The object of the mining was stratiform mineralization. Near Bockau, (underground) tin mining was first mentioned in documents in 1663 and was active, with interruptions, until the beginning of the 19th century. On an area of more than 2 km², hundreds of dumps and pits, numerous abandoned mines, and historically very remarkable underground objects are known. Mining was preferably carried out near the surface. Despite favourable morphological conditions, hardly any deep adits were built and depths of more than 20 m were rarely reached. In this context, Pingen are abandoned ore pits or prospecting sites, where ores and other mineral resources were mined. Geometrically Pingen resemble round depressions created by the collapse of a mine workings (shaft, adit, underground drift), collapsed due to its age, leaving this relic usually funnel-shaped (down-facing cone), often surrounded by an annular dump (0.5 to 3 m) caused by the lowered surface.

The main aim of this application is to identify mining relics (Pingen) using a UAV equipped with LiDAR technology. The LiDAR technology allows to obtain a high-resolution Digital Elevation Model (DEM) and Point cloud of the surveyed area. The DEM is the digital representation of topographic and manmade features located on the surface of the earth.

In this framework, a LiDAR survey was conducted in a flight area of about 1,6 km² within the “Bockau” area during August 2022. The surveyed features include the elevations, colorized Point Cloud (RGB values), transparency levels, reflectance values, and number of returns (significant for the penetration of the vegetation). This information was used to derive the final products: DEM and classified Point cloud. Various spatial analyses were conducted and tested on the DEM and Point Cloud to automatically identify the mining relics. Hydrogeological analysis showed to be the best approach for the automatic identification of Pingen. As a result, the ground depressions were identified and nested surfaces were delineated.

The automatically identified features were validated by examination of randomly selected samples on the surveyed point cloud, comparison to identified features based on the National dataset of the LiDAR Database and field verification. The validation revealed, that around 90% of the Pingen in the study area were successfully identified with the developed workflow. Other features of interests couldn’t be identified due to the similarity of geometric properties with other topographical features, dense vegetation, erosion etc.

The paper has been prepared in the frame of the Horizon 2020 co-funded project GOLDENEYE, which has received funds through the Grant Agreement 869398.