研究目的
The identification, analysis, and survey of archaeological remains and proxy indicators is a complex challenge especially in areas covered by dense vegetation, as wooded and forest land. The constraints are due to (i) vegetation cover type, density, and height, as well as (ii) dimensions and state of conservation of archaeological remains.
研究成果
The study successfully applied LiDAR and derived models to rediscover a medieval fortified settlement and detect an unknown urban area under dense vegetation. The integration of diverse visualization techniques and automatic feature extraction methods provided comprehensive insights into the archaeological landscape of Torre Cisterna.
研究不足
The study acknowledges the limitations of LiDAR in areas with dense vegetation and complex topography, where the detection of subtle archaeological remains can be challenging. The effectiveness of LDMs is influenced by vegetation cover density and the size of archaeological features.
1:Experimental Design and Method Selection
The study applied diverse LiDAR-derived models (LDMs) based on relief visualization techniques to enhance the visibility of archaeological features under dense vegetation. The methodology included noise filtering of Digital Terrain Model (DTM) before obtaining the LDMs.
2:Sample Selection and Data Sources
The study area was the medieval site of Torre Cisterna in Basilicata, Southern Italy, selected for its dense vegetation cover and complex topography.
3:List of Experimental Equipment and Materials
LiDAR survey was carried out using a full waveform scanner (RIEGL LMS-Q560) on board a helicopter. The data were processed using Terrasolid's Terrascan for DTM generation.
4:Experimental Procedures and Operational Workflow
The workflow included LiDAR data acquisition, noise reduction, DTM generation, application of diverse visualization techniques (PCA of Hill Shading, Local Relief Model, Sky View Factor, Anisotropic Sky View Factor, Positive Openness, and Slope), and automatic feature extraction.
5:Data Analysis Methods
Data analysis involved visual inspections of LDMs and automatic feature extraction using an object-oriented approach based on spatial autocorrelation, unsupervised classification, and segmentation.
独家科研数据包,助您复现前沿成果,加速创新突破
获取完整内容