汽車盲點偵測系統(Blind Spot Detection System),主要是針對行車駕駛時,因為汽車環境限制而產生的視覺盲點區域,做預防偵測以增進行車之安全。目前該技術廣泛由各車廠以及許多研究機構發展,並應用於行車安全防護上,其盲點偵測範圍,大多為車側偵測以及車後偵測,而其使用的偵測技術,則多為影像物體辨識偵測及微波雷達回波偵測。本論文主要則針對車輛駕駛,因前方左右支柱所造成的視野盲點及轉彎盲點,以超廣角度之行車紀錄器影像系統來加以偵測,以達到預防警示,提高駕駛行車安全,並降低車禍事件之發生。
Vehicle blind spot detection system is used for road driving, in order to detect the visual blind spot area due to the car environmental constraints to enhance the traffic safety. The technology is widely developed and applied by car manufacturers and research institutions to enhance safety of driving. Most blind spot detection area includes the range of both car sides and rear side, and the mostly used technologies are image object recognition and short-wave radar echo detection. In this thesis, we focus on the detection of the front blind spot areas due to the car pillars and car turning by using ultra wide angle vehicle record image system. We hope the proposed technology can be used for object detection and early warning prevention to improve road safety and to reduce driving accident occurrence.