摘要
本文针对航海雷达图像阴影反演有效波高算法进行了扩展和补充,通过对Rune(Rune Gangeskar)阴影反演海面波高算法进行分析,发现波高随着阴影比例而变化,与固定区域的阴影比例成线性变化关系,基于此本文利用最小二乘法建立起波高与比例间的线性模型反演波高。首先,应用边缘检测提取出阴影阈值求阴影比例;其次,应用最小二乘标定波高与阴影比例间的相关系数;最后,利用实测数据来检验算法,并将检测结果与外部参考作比较。结果表明本文算法与实测参考数据较为接近,其相关系数为0.89,均方根误差为0.42,方差为0.21,证实了本算法的实用性与可行性。
The available algorithms for estimation of wave heights from shadowing in marine radar images are supplemented and expanded.In this paper,the algorithm proposed by Rune Gangeskar is analyzed at first,and we found that the wave heights change linearly with shadowing proportion in specific area.On this basis,a linear relation model is developed using the least square method for wave heights and shadowing proportion,which can be used to estimate the wave heights.Firstly,edge detection is employed to extract the shadowing thresholds and then determine the shadowing proportion.Secondly,the least square method is used to determine the correlation coefficient between wave heights and shadowing proportion.Finally,test data is used to verify the proposed algorithm,and the test results are compared with the external reference data.The comparison shows that the values estimated by using the proposed algorithm resemble the external reference data well.In this case,the proposed algorithm has a correlation coefficient of 0.89,a root mean square error of 0.42,and a variance of 0.21.Therefore,the proposed algorithm is practical and effective.
作者
杨江洪
顾宗山
卢志忠
YANG Jianghong;GU Zongshan;LU Zhizhong(No.38 Research Institute of CETC, Hefei 230088;Harbin Engineering University, Harbin 150001)
出处
《火控雷达技术》
2020年第3期7-12,共6页
Fire Control Radar Technology
基金
公共领域基金(61404130110)。
关键词
雷达图像
阴影比例
最小二乘
有效波高
radar image
shadowing proportion
least square
significant wave height