Discovering the Ancient Tomb under the Forest Using Machine Learning with Timing-Series Features of Sentinel Images: Taking Baling Mountain in Jingzhou as an Example

Cultural traces under forests are one of the main problems affecting identification archaeological sites in densely forested areas, so it is full challenges to discover ancient tombs buried dense vegetation. The covered can be identified by studying time-series features vegetation covering on multi-time series remote sensing images because deep underground have long-term space structures, which affect intrinsic properties surface soil that growth status different from area without tombs. We first use highly detailed DSM data select cannot visually distinguished optical images. Then, we explored and constructed temporal forest non-ancient images, such as radar timing-series Sentinel 1 multi-spectral index 2. Finally, based these machine learning, designed an automatic algorithm for forest. method has been validated Baling Mountain Jingzhou, China. It very feasible automatically identify using Additionally, large or concentrated more accurate, accuracy improved after adding features. paper concludes with a discussion current limitations future directions method.