🤖 AI Summary
This work addresses the challenges of insufficient real-time perception accuracy and low-fidelity offline 3D reconstruction for planetary rovers operating in complex terrains by proposing a lightweight visual system that integrates real-time neural depth estimation with offline monocular depth reconstruction. The system combines CLAHE image enhancement, YOLOv11n object detection, a custom distance estimation network, and the Depth Anything V2 model to generate high-fidelity dense point clouds, which are subsequently fused using Open3D. Experimental results demonstrate a median depth error of only 2.26 cm within the 1–10 meter range and achieve a balanced trade-off between detection precision and recall on lunar grayscale imagery, thereby validating the method’s effectiveness in metric accuracy, geometric detail preservation, and computational efficiency.
📝 Abstract
This study presents a vision system for planetary rovers, combining real-time perception with offline terrain reconstruction. The real-time module integrates CLAHE enhanced stereo imagery, YOLOv11n based object detection, and a neural network to estimate object distances. The offline module uses the Depth Anything V2 metric monocular depth estimation model to generate depth maps from captured images, which are fused into dense point clouds using Open3D. Real world distance estimates from the real time pipeline provide reliable metric context alongside the qualitative reconstructions. Evaluation on Chandrayaan 3 NavCam stereo imagery, benchmarked against a CAHV based utility, shows that the neural network achieves a median depth error of 2.26 cm within a 1 to 10 meter range. The object detection model maintains a balanced precision recall tradeoff on grayscale lunar scenes. This architecture offers a scalable, compute-efficient vision solution for autonomous planetary exploration.