Low-Cost Static LiDAR–IMU Based Real-Time Object Detection for Indoor Environments

Abstract

Reliable detection and tracking of moving objects is essential for indoor robotics and smart environments. This paper presents a real-time 2D object detection and tracking system that uses a fixed planar LiDAR sensor combined with an inertial measurement unit (IMU), all implemented on a Raspberry Pi. The LiDAR is set up in a stationary position, which allows for continuous monitoring of a specific area and helps separate moving objects from static background structures based on motion. Unlike mobile sensing systems, this approach does not require estimating ego-motion. Instead, the IMU identifies and reduces measurement errors caused by mechanical vibrations and slight changes in orientation that can impact fixed LiDAR setups. Moving objects are detected by analysing consecutive range scans over time, grouped spatially using lightweight clustering, and tracked in the 2D plane with an efficient state model. While the system can cater to human-centric applications, it does not impose any object-specific shape or learning-based assumptions. This flexibility allows the framework to apply to a variety of moving objects. Experimental results in indoor settings show stable real-time performance even with limited computational resources. These findings confirm that low-cost LiDAR and IMU sensing is suitable for general object detection and tracking.