<![CDATA[Indoor human activity monitoring is essential for Internet of Things (IoT) security and automation. However, Passive Infrared (PIR) sensors and cameras face limitations regarding low-movement detection and privacy issues. This study implements an HLK-LD2410C Human Presence Radar using 24 GHz mmWave and ESP32 to classify real-time activities. Parameters including Presence, Distance, P_Energy, and M_Energy from UART communication are categorized via thresholding into No Presence, Static, Light Movement, and Heavy Movement. Testing reveals detection ranges up to 5.2 m for Static, 5.3 m for Light Movement, and 3.8 m for Heavy Movement. Confusion matrix evaluation yields 97.50% accuracy, 97.73% precision, 97.50% recall, and a 97.49% F1-score. M_Energy values depend on distance and movement intensity, while multi-object scenarios reflect the dominant activity within the detection zone. Consequently, the proposed system provides a non-contact, responsive, and privacy-preserving solution for indoor human activity monitoring.]]>
Copyrights © 2026
