红外十字分划线的提取和中心定位是红外观瞄装备实现光轴检测数字化的重要组成部分。针对红外图像对比度差、信噪比低、边缘模糊的特点,提出了基于灰度形态学的十字分划线提取方法,应用一个形态学顶帽算子并对得到的结果使用大津(Otsu)方法进行阈值处理。根据十字分划线的形态特点设计了十字结构元素对提取出的十字分划线进行腐蚀,对红外十字分划线实现了定位。实验结果表明,提出的算法能在图像对比度较低,边缘模糊的情况下有效地提取出十字分划线,并准确定位到十字分划线的中心。

The obtaining and center location of infrared crosshair is an important part of digital detection for infrared instrument optical axis. In order to address the poor contrast of the infrared image, low signal to noise ratio and edge blur of infrared images, a method base on grayscale morphology of obtained cross is proposed. This method uses a morphology top-hat operator and the method of Otsu thresholding on the obtained results. According to the crosshair morphology characteristics, a cross-shaped structure element is designed. The obtained crosshair is eroded by the cross-shaped structure element to center location the crosshair. The experimental results demonstrate that the proposed algorithm can effectively extract crosshair and accurately locate the crosshair center in the case of low image contrast and fuzzy edge.