This article aims to improve the study of archaeological sites in the territory of Milena, a town located in Sicily (Italy). Between the 80 s and 90 s different excavation campaigns had documented numerous prehistoric sites. However, most of the ancient settlements have yet to be discovered and studied. The search operations are made difficult by the large size of the territory and its geomophorological characteristics. In fact, many settlements stand on chalky hills that are rapidly crumbling, risking damage to the archaeological remains. To intervene as soon as possible, we have relied on the use of modern approaches, capable of speeding up the work of researchers. Here we propose a new approach based on remote sensing and the GIS technology to detect archaeological sites. In particular, we used multispectral remote sensing data in a GIS environmental both for predictive and post-dictive analyses. We were able to locate areas where there are ceramics and elements that can be recoded to human settlements. The quality of the results depends on the choice of the remotely sensed images, which can have different spectral characteristics, on the acquisition period, but also on the preprocessing algorithms adopted for the classification in the GIS environment.

本文旨在改进对位于西西里岛（意大利）小镇米莱纳境内考古遗址的研究。在 80 年代和 90 年代之间，不同的挖掘活动记录了许多史前遗址。然而，大多数古代定居点尚未被发现和研究。由于领土面积大及其地貌特征，搜索作业变得困难。事实上，许多定居点位于白垩质的山丘上，这些山丘正在迅速崩塌，有损坏考古遗迹的风险。为了尽快进行干预，我们依靠使用现代方法，能够加快研究人员的工作。在这里，我们提出了一种基于遥感和 GIS 技术的考古遗址检测新方法。尤其是，我们在 GIS 环境中使用多光谱遥感数据进行预测和预测分析。我们能够找到可以重新编码为人类住区的陶瓷和元素的区域。结果的质量取决于对具有不同光谱特征的遥感图像的选择、采集周期以及在 GIS 环境中分类所采用的预处理算法。