<![CDATA[The growth of four-wheeled vehicles triggers parking search inefficiencies and security issues in conventional facilities. This research aims to develop an automatic parking slot locking prototype based on the Internet of Things (IoT) and Optical Character Recognition (OCR) using a prototyping approach. The local edge server architecture combines the YOLOv8 algorithm to localize license plates and PaddleOCR for text extraction. The performance of the text extraction system was validated using 192 license pelate image samples. An ESP32 microcontroller is utilized to control the slot-locking servo motor and ultrasonic sensors stabilized by a moving average algorithm. Based on black box testing results, the security automation success rate reached 100%. The Greedy algorithm consistently allocates the nearest empty slots with a total real-time computation latency of 180 milliseconds. OCR accuracy reached 92% in bright light conditions but experienced performance degradation under 20 Lux illumination. This prototype demonstrates consistent reliability in ensuring autonomous space certainty and facility security, with potential for further development through the integration of infrared illumination in low-light environments.]]>
Copyrights © 2026
