基于同态滤波的高空无人机测绘图像薄云去除方法

Method for removing thin clouds from high-altitude unmanned aerial vehiclemapping images based on homomorphic filtering

高空无人机测绘图像中,薄云对太阳辐射的散射作用会影响遥感相机的成像能力,从而使测绘图像的清晰度水平大幅下降,对此,设计基于同态滤波的高空无人机测绘图像薄云去除方法。建立同态滤波模型,根据测绘图像的分区情况,定义增强处理表达式,实现基于同态滤波的无人机测绘图像增强处理。将经过增强处理后的测绘图像输入到测绘图像生成器中,并通过求解薄云影响损失函数的方式,推导具体的薄云影像去除条件,完成基于同态滤波的高空无人机测绘图像薄云去除方法的设计。实验结果表明,在同态滤波模型的作用下,高空无人机测绘图像中薄云影像的含量下降了45.6%,薄云去除效果较好,不会再对测绘图像的清晰度水平造成严重影响。

In high-altitude unmanned aerial vehicle mapping images,the scattering effect of thin clouds on solar radiation can affect the imaging ability of remote sensing cameras,resulting in a significant decrease in the clarity level of the mapping image.Therefore,a thin cloud removal algorithm based on homomorphic filtering is designed for high-altitude unmanned aerial vehicle mapping images.Establish a homomorphic filtering model,define an enhancement processing expression based on the partition of the surveying and mapping image,and achieve unmanned aerial vehicle surveying and mapping image enhancement processing based on homomorphic filtering.Input the enhanced surveying and mapping images into the surveying and mapping image generator,and derive specific conditions for removing thin cloud images by solving the thin cloud impact loss function.Complete the design of a thin cloud removal algorithm for high-altitude unmanned aerial vehicle surveying and mapping images based on homomorphic filtering.The experimental results show that under the action of homomorphic filtering model,the content of thin cloud images in high-altitude unmanned aerial vehicle mapping images has decreased by 45.6%,and the removal effect of thin clouds is good,which will no longer have a serious impact on the clarity level of the mapping images.