无源毫米波海岸线图像的边缘提取方法研究

Research on coastal line edge extraction method from passive radiometric image

分析了无源毫米波辐射图像轮廓特征,讨论了基于二阶微分算子与形态学的图像边缘检测的方法,根据海岸线的无源毫米波辐射特性和成像特点对图像区域标识,然后利用自适应阈值操作实现多目标区域的细分,最后采用形态学边缘检测海岸线特征方法。实验结果表明:对比传统的微分算子处理结果,形态学算法具有较高的检测精度,适用于提取无源毫米波辐射图像边缘。

On basis of analyzing radiometric image outline characteristic and introducing the image segmentation approaches,this paper presents a coastal line extraction approach from 3 mm radiometric image.First it marks the typical objects according to the millimeter-wave radiation characteristic of different objects.Then it realizes image segmentation by the self-adapting threshold.Fi- nally it adopts a sub-zone and multiple structural element morphological processing method.The result shows the method has higher precision and comprehensive applicability by contrasting to LOG processing result.