Mapping landslides is essential for understanding hazards and risk assessment. This study used a geographic object-based image analysis (GEOBIA) approach with high-resolution lidar data to map forest-covered historical landslides in Jena, Germany. Optimizing the moving-window size for lidar derivatives improved accuracy, detecting more landslides and reducing errors. This method showcases the potential of lidar-based approaches for global landslide inventory and hazard assessment.