Towards Semi-automatic Detection of Illegal...
网址 https://doi.org/10.1007/978-3-031-97663-6_36
De Luca G., Arcidiaco L., Corongiu M., De Filippis T., Nati C., Rogai M. and Gianni Picchi Abstract Illegal logging is a global issue with severe ecological, economic, and social consequences. In Europe, it often occurs as small-scale, selective harvesting, which, despite its limited footprint, significantly contributes to forest degradation, biodiversity loss, and ecosystem disruption. Detecting illegal logging is essential for assessing its impacts and supporting sustainable forest management. However, its fragmented nature poses significant detection challenges, requiring advanced monitoring solutions. This study presents an exploratory data analysis and preliminary results toward a semi-automatic monitoring framework developed within the EU Horizon SINTETIC project (Single Item Identification for Forest Production, Protection, and Management). The framework integrates high-resolution satellite data from Sentinel-1 (SAR) and Sentinel-2 (multispectral) to analyze time-series trends in optical spectral indices and dual-polarized SAR backscatter, identifying distinctive patterns associated with logging events. An unsupervised sliding-window breakpoint detection algorithm was implemented to detect logging-induced disturbances in satellite time series. The method was validated using georeferenced ground data from legal mechanized logging operations, provided in StanForD 2010 standard format. Two logging scenarios were examined: clear-cutting and selective logging. The exploratory analysis provided valuable insights into forest disturbance patterns, while breakpoint analysis successfully identified the timing of logging events in both scenarios. This system offers a promising approach for detecting illegal logging.
该资源暂时没有视图
其他信息
| 域 | 价值 |
|---|---|
| Data last updated | 2025年7月7日 |
| Metadata last updated | 2025年7月7日 |
| 创建的 | 2025年7月7日 |
| 格式 | 未知 |
| 授权 | Creative Commons Attribution Share-Alike |
| Datastore active | False |
| Has views | False |
| Id | ac737362-53e2-4890-96ca-09b53ae15547 |
| Package id | 688f28f2-75cf-4dab-8644-643bd1d346fe |
| Position | 0 |
| State | active |