BLESS: BLE based Street Sensing for People Counting and Flow Direction Estimation

Abstract

In urban planning and optimization of commercial facilities, pedestrian flow analysis is essential. Currently, camera-based pedestrian flow analysis methods are widely adopted, but this solutions suffers of high costs and privacy concerns. As a new approach to address these problems, numerous methods utilizing BLE (Bluetooth Low Energy) and Wi-Fi have been proposed. However, many previous studies have focused on the congestion level of spaces, with insufficient attention given to analyzing the direction of movement. In this study, we propose a pedestrian flow estimation method that considers movement direction (upward, downward) using two BLE sensors. In our proposed method, time-series data obtained from the two sensors is used to create differences between the data and perform clustering. The number of BLE addresses belonging to each cluster serve as features for estimating pedestrian flow. To evaluate the effectiveness of this method, experiments were conducted on streets in Osaka Prefecture, Japan. For one sample of the collected data (5 minutes), the average pedestrian flow was 75.5 people moving upward, 81.8 people moving downward, with a total flow of 157.2 people. Evaluating the proposed method in 5-minute increments showed that using the DTW metric for clustering, the upward, downward, and overall pedestrian flow could be estimated with an average absolute error of 18.6, 16.0, and 21.9, respectively.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 6 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.