A Geospatial Dataframe of Collapsed Buildings in Antakya City after the 2023 Kahramanmaraş Earthquakes Using Object Detection Based on Yolo and VHR Satellite Images

The earthquake sequence in Kahramanmaraş, Türkiye, on February 6, 2023, with magnitudes of 7.8 and 7.6, resulted in significant destruction, impacting buildings and infrastructure across the region. This study focuses on detecting collapsed buildings post-earthquake in heavily damaged Antakya City, utilizing YOLO models. Trained rigorously on Maxar's VHR satellite imagery, YOLOv7, YOLOv7x, YOLOv8l, and YOLOv8x delivered notable results, especially YOLOv7, achieving a [email protected] of 0.79. The integration of precise geographical coordinates enhances insights into the distribution of 216 detected collapsed buildings within 28.20 km². Challenges include misclassifying non-building structures and advocating for diverse dataset inclusion. While YOLO models efficiently contribute to post-earthquake studies, the generalizability of the models is still context-specific, necessitating the exploration of diverse datasets within different scenarios.