基于灰度梯度分割的机头检测算法

在末制导阶段,提出了基于灰度梯度分割的机头检测算法.首先利用平均梯度算子进行边缘检测,分割出目标区域和目标最亮区域,寻找各区域形心位置,通过判断目标区域形心和目标最亮区域形心位置,检测出飞机轴线,确定出候选机头信息;然后利用序列图像帧间较为稳定的相关特征进行D-S合成,对候选机头信息的可信度进行判断,得出准确的机头信息;最后利用数帧不同姿态的目标红外图像数据进行了算法验证.仿真结果表明:在末制导阶段,基于灰度梯度分割的机头检测算法,可有效地识别出机头,并具有较快的运算速度,可实现实时处理.

Detection algorithm of aircraft nose based on gray grads segmentation

In the stages of terminal guidance,the detection algorithm of aircraft nose based on gray grads segmentation was proposed.Firstly,an average gradient operator was used for edge detection to separate the object region and the lightest region in the object in order to search for the heart-shaped position,and through comparing the position of the object region and the lightest region,the aircraft axis was detected so that the information of the candidate aircraft nose was identified.Then the D-S evidence theor...