<![CDATA[The rising number of vehicles has led to increased traffic accidents, partly due to blind spot areas on trucks that block the driver’s view of surrounding objects. This study aims to develop an Internet of Things (IoT)-based blind spot detection system integrating a JSN-SR04T ultrasonic sensor for early warning and computer vision using YOLOv11n on an ESP32-CAM for object detection. The method includes collecting and augmenting an initial dataset of 417 images, training the YOLOv11n model on Google Colab, and designing hardware consisting of the JSN-SR04T sensor, ESP32-CAM, NodeMCU ESP8266, and LED indicators, with MQTT-based communication for real-time monitoring. Position testing shows the highest accuracy in the front position for cars at 93% (142 cm), while the farthest detection distance is achieved for pedestrians in the rear position up to 329 cm (73% accuracy). Speed testing indicates accuracy at 10–20 km/h reaches 100% (car), 90% (motorcyclist), and 80% (pedestrian), then drops at 50 km/h to 70%, 60%, and 50%. The system is effective for early warnings at low to moderate speeds but requires improvement for small objects at high speeds.]]>
Copyrights © 2026
