REoOS: Rapid Estimation of Open Space in Complex and Dynamic Environments Using Depth Sensor-Based 3D Analysis

In recent years, depth sensors, which capture distance information between a sensor and an object, have garnered significant attention. The three-dimensional data obtained from these sensors is widely used in tasks such as object detection, dimension estimation and moving space analysis. Advances in research on methods like object detection and moving space analysis using depth sensors have significantly improved their accuracy. However, these systems face challenges in achieving real-time performance due to the high computational demands of processing large volumes of depth data, which strain hardware capabilities. To address these challenges, we developed a method called REoOS (Rapid Estimation of Open Space), which estimates open space in complex and dynamic environments using only depth data, with a strong emphasis on real-time performance. Extensive experiments demonstrated the real-time effectiveness of REoOS on small computers, not only under typical indoor and outdoor lighting conditions but also in low-light environments, including nighttime.