<![CDATA[This study proposes and implements an automatic vehicle door locking system based on Raspberry Pi 5 integrated with a real-time blind spot vehicle detection mechanism using a single vision camera without additional sensors such as radar or ultrasonic modules. The system was developed to reduce the risk of dooring accidents caused by passengers opening vehicle doors without observing approaching vehicles from the rear blind spot area. Digital image processing and OpenCV-based object detection techniques were utilized to identify approaching vehicles in real time. When a vehicle is detected within the predefined danger zone, the system automatically activates the door lock actuator through a relay module and simultaneously provides audio-visual warnings via a speaker and TFT display. Experimental evaluations were conducted under daytime and nighttime conditions with vehicle speeds ranging from 10 to 50 km/h, where each scenario was repeated ten times. The proposed system achieved an average response time of 0.198 seconds during daytime operation and 0.23 seconds at night. Distance estimation testing produced an average accuracy of 99.3% with a mean error rate of only 0.7%. In addition, the door lock actuator and warning system achieved a 100% operational success rate across all experimental trials. The results demonstrate that the proposed OpenCV-based safety system is capable of performing reliable real-time vehicle detection and preventive automatic door locking, making it a promising low-cost solution for enhancing passenger safety and reducing blind-spot-related traffic accidents.]]>
