自然场景下自动瞄准中的标杆识别方法

Recognition of Automatic Collimation of Norm Pole in Natural Scene

利用图像处理软件实现瞄准的自动化。由于标杆图像背景复杂,采用颜色分割的方法。利用蓝色和红色分量比及红色分量特征将标杆像素分割出来,利用面积和周长等形状特征使图像中仅标杆像素有非零值。对原图像灰度化,在标杆区域内利用像素细分法最终精确平分标杆,实现标杆的自动精确瞄准。

Image processing was used to realize automatic collimation of missile. For the complex background of directional norm pole, the method of color segmentation played an important role. Norm pole pixels were segmented by the character of red weight and the ratio of red weight to blue weight. The characters of area and perimeter were used to make norm pole pixels nonzero value. Automatic collimation of norm pole was realized by precisely halving norm pole on the method of subpixel on former gray image.