基于攻击特征的检测具有准确性高的优点；
The way of attack signature detection has advantage of high accuracy.
边缘检测是提取图像特征的首要条件。
Image edge detection is the first step to obtain image feature.
本文提出的复杂特征值检测算法降低了单一特征值检测的误报率。
Complicated signature detection improves false positive rate compared with single signature detection.
因此，有必要对每一个镜头检测决策进行特征约简。
Therefore, it is necessary to perform feature reduction for every decision of the shot boundary.
PERCLOS是有效地检测驾驶员瞌睡的特征。
车辆局部特征提取是车辆检测的关键。
Feature extraction of the local vehicle is the key of vehicle detection.
车辆局部特征提取是车辆检测的关键。
Feature extraction of the local vehicle is the key of vehicle detection.
应用推荐