L’identification des aménagements funéraires du Fadnoun dans le Sahara central en Algérie par l’utilisation de l’apprentissage profond et les outils de la géomatique

Abstract

The research on funerary arrangements in the Fadnoun region, located in the heart of the Central Sahara, aims to explore and document prehistorical burial sites that are often inaccessible due to the extreme conditions of the desert. By utilizing modern technologies such as satellite imagery and remote sensing, this initiative seeks to shed light on ancient funerary structures and deepen our understanding of burial practices in this region. This study aims to enhance knowledge about the funerary arrangements of the Tassili of Fadnoun by employing convolutional neural networks to detect archaeological mounds shaped like keyholes. The objectives include locating and analyzing ancient funerary structures through high-resolution satellite images, developing a neural network model to recognize and classify these mounds, and contributing to the good documentation of cultural heritage by providing accurate data on burial sites. Preliminary results show that the use of convolutional neural networks has enabled the identification of new archaeological mounds in the Fadnoun region, revealing unprecedented funerary practices. The integration of remote sensing with traditional methods has proven effective in locating hard-to-access sites, thereby enhancing the good documentation of cultural heritage. This research aims to improve our understanding of the civilizations of the central Sahara and better document cultural heritage by using convolutional neural networks to detect archaeological mounds. The results demonstrate increased efficiency in identifying these sites through the analysis of high-resolution satellite images.