一种基于HSV色彩空间的加油锥套特征提取方法

A FEATURE EXTRACTION METHOD OF REFUELING DROGUE BASED ON HSV COLOR SPACE

自动空中加油过程中视觉导航图像的复杂性对加油锥套特征的检测与识别提出了很高的要求。为准确获取空中加油对接阶段加油锥套端面的中心坐标,提出一种基于图像HSV(hue,saturation,value)色彩空间信息的加油锥套特征提取方法。引入HSV色彩空间,根据颜色与形状特征选取原则选取加油锥套特征。对图像进行色彩空间转换、色相滤波、二值化以及边缘检测处理,利用最小二乘椭圆拟合的方法得到加油锥套端面的中心坐标。实验结果表明,所提出的方法能够准确提取加油锥套特征,实时解算得到加油锥套端面的中心坐标,为自动空中加油视觉导航提供了一种可靠性更高的特征提取方法。

The complexity of visual navigation images in the process of automated aerial refueling has brought more challenge to feature detection and labeling of the refueling drogue. In order to precisely acquire the central coordinates of refueling drogue during aerial refueling, the article proposes a feature extraction method of refueling drogue based on HSV color space information. By introducing HSV color space, in accordance to color and shape feature selection principles, the feature of refueling drogue is selected. Then the image is processed with color space conversion, color phase filtration, binarization and edge detection. By least square eclipse fitting method the central coordinates of refu- eling drogue are obtained. Experiment result show that the proposed method can accurately extract the feature of refueling drogue and timely compute the central coordinates of refueling drogue, thus providing a more reliable feature extraction method for automated aerial refueling visual navigation.