影像辨識及物件偵測已廣泛應用在生活當中,像是汽車自動跟車、輔助駕駛、人臉辨識均與兩種技術息息相關。以往車輛的安全防護措施都是採用被動的方式,但此種方式通常皆是在受到撞擊後才起到保護的作用,隨著科技迅速發展,硬體成本大幅下降及深度學習和電腦視覺技術日漸成熟,透過影像辨識及物件偵測來提前預防碰撞的發生變得可行。本研究將利用開放資料庫YOLO(You Only Look Once)、OpenCV(Open Source Computer Vision)進行影像和視訊之檢測,再將此項技術應用於駕駛主動安全防護,當駕駛或是附近車輛有危險行為時發出警告。 關鍵字:YOLO、主動駕駛防護、物件偵測、車道辨識
Image recognition and object detection have been widely used in life, such as Adaptive Cruise Control, Advanced Driver Assistance Systems, and Face Recognition System are closely related to the two technologies. In the past, the vehicle safety protection measures were passive, but this method usually played a protective role after being hit. With the rapid development of technology, the hardware cost decreased significantly and deep learning and computer vision technologies have become increasingly mature, it is more feasible to prevent accident in advance through image recognition and object detection. This research will use the open database YOLO (You Only Look Once) and OpenCV (Open Source Computer Vision) to detect images and video, and then apply this technology to active driving safety protection system, which warns driver when driver or nearby vehicles are dangerous. Keywords: YOLO, Driving active safety protection, Object detection, Lane recognition