高级检索

基于样条曲线的一种遥感图像去云方法

Cloud elimination method in remote sensing image based on spline curve

  • 摘要: 在可见光遥感图像中,常常因为云的阻挡而很难获得有云区域的信息.在没有预知信息的情况下,为了从一幅图像中获取有云区域的信息,采用基于样条曲线的方法,对有云区域像素的亮度值进行拉伸,得到去云后的遥感图像,既保持了云区与边界的连续性,又容易获取云区域内的信息.再使用高通滤波器与云区内像素亮度值进行卷积,突出了云区内的图像信息.实验结果表明,对于一幅有1/4被乌云覆盖的遥感图像,突出了云区内隐藏的丰富信息,能获取诸如水坝的墙体、水流等重要信息,说明图像增强技术的多样性.

     

    Abstract: Because of the cloud barrier in a remote sensing image,the information couldn't be captured.Based on spline curve,a proposed method kept the continuity between inner and outer edge of cloud and captures more details in the remote sensing image without previous information.Firstly,the brightness in the cloud region was stretched.Secondly,the more details in the cloud region was exposed after the convolution between the cloud region and a high pass filter.Thirdly,the brightness of the pixels in the cloud region was improved with histogram stretching method.The results show the more details,such as the wall of dam,water stream can be exposed in a remote sensing image in which over a quarter of whole size was covered black cloud,which means the image enhancement technology is multiformity.

     

/

返回文章
返回