驾驶员疲劳检测中的眼睛定位算法

人眼状态与疲劳程度有很好的相关性,眼睛定位在驾驶员疲劳检测中有十分重要的地位。目前眼睛定位算法都是针对睁开的眼睛,不能定位闭合的眼睛。为了能够定位任意状态下的眼睛,提出了Adaboosting算法与几何模板融合的眼睛定位算法。在眼睛粗定位区域中用Ada-boosting算法检测睁开的眼睛,修改几何模板。对于Adaboosting算法中不能检测到眼睛时则用几何模板法进行眼睛定位。实验表明该算法有较好的检测率和运算速度。

Eye location algorithm in driver's fatigue detection

There is a good correlation between eye state and fatigue.Eye location is very important in the driver’s fatigue detection.Eye location algorithms only can detect the opened eyes,but the closed eyes can’t be located.To locate any state of the eyes,a new eye location algorithm that merging Adaboosting and geometric template algorithm is proposed.Opened eyes are detected by Adaboosting algorithm in rough location area,and the geometric template is modifed.The eyes that can not detect by Adaboosting algorithm are detect by geometric template algorithm.The experiments show that the algorithm has good detection rate and fast speed of operation.