An Image Processing-Enabled Humanoid Robot for Autonomous Guest Reception Using ESP32-CAM

Abstract

This study presents the design and development of a humanoid robot for guest reception based on image processing using the ESP32-CAM module. The robot is capable of detecting human presence visually, responding through an arm-waving gesture, and delivering an automatic voice greeting. The system integrates the ESP32-CAM for real-time face detection, servo motors for arm actuation, and a DFPlayer Mini audio module for verbal responses. Experimental results show that the system achieves a 100% face detection success rate within 30–90 cm, with an average response time of 1.21 seconds. Performance evaluation under different lighting conditions (30–580 lux) indicates optimal operation between 100–300 lux, with stable detection times ranging from 1.5–2.5 seconds. The system fails to detect faces under 10 lux  and above 800 lux. Servo–audio synchronization tests across ten subjects achieved perfect reliability with latency under 50 ms. The robot operates efficiently with power consumption of 1.2 A during active detection and 0.4 A when idle. This research demonstrates that a low-cost embedded architecture can successfully realize a functional humanoid robot for automated public reception tasks in indoor environments.