Low-Cost Multimodal Fusion of 2D LiDAR and RGB Camera for Accurate Object-Level Perception

Sensor fusion between LiDAR and camera modalities has emerged as an effective approach for improving perception in autonomous robotic systems. However, most existing solutions rely on high-cost sensors and computationally intensive algorithms. This paper presents a low-cost LiDAR–camera fusion framework using a YDLiDAR TG30 2D LiDAR and a Logitech C270 RGB camera to achieve real-time object-level perception. The system is developed using ROS 2 Humble and integrates intrinsic and extrinsic calibration to accurately project LiDAR point clouds onto camera images. A custom fusion node performs coordinate transformation and depth-based visualization in real time. Experimental validation conducted in indoor environments demonstrates reliable projection accuracy within near-range distances, even under low-light conditions. The proposed framework offers an affordable, modular, and computationally efficient solution suitable for educational robotics, indoor navigation, and resource-constrained mobile platforms.