LiDAR dataset for object localization

Object tracking systems within closed environments employ light detection and ranging (LiDAR) to address privacy and confidentiality. Data collection occurred in two distinct scenarios. The goal of scenario one is to detect the locations of multiple objects from various locations on a flat surface in a closed environment. The second scenario describes the effectiveness of the technique in detecting multiple objects by using LiDAR data obtained from a single, fixed location. In real-time experiments, human subjects navigate predefined paths while LiDAR is positioned centrally to track their locations. The focus of the second scenario's data collection was a strategically placed LiDAR that can proficiently detect moving objects in its vicinity. These dual scenarios were chosen strategically to ensure a well-rounded and diverse dataset that encompasses a broad spectrum of relevant information. The resulting dataset, enriched by the insights from both scenarios, forms a robust foundation for the comparison of different regression techniques for predicting bounding box coordinates for object localization. These findings pave the way for the extension of various real-time applications, including crowd management in malls, surveillance systems, and diverse Internet of Things scenarios.