Estimation of physical activities of people in offices from time-series point-cloud data

TL;DR

This paper introduces an edge computing system that estimates office physical activities, including typing, from time-series point-cloud data collected via LIDAR sensors, demonstrating successful real-world application.

Contribution

It presents a novel edge computing approach for activity estimation using LIDAR-based point-cloud data in office environments.

Findings

Successful model construction for estimating typed characters

Effective use of LIDAR sensors for activity recognition

Demonstrated real-world applicability

Abstract

This paper proposes an edge computing system that enables estimating physical activities of people in offices from time-series point-cloud data, obtained by using a light-detection-and-ranging (LIDAR) sensor network. The paper presents that the proposed system successfully constructs the model for estimating the number of typed characters from time-series point-cloud data, through an experiment using real LIDAR sensors.