遥感图像的子空间边缘融合算法改进

Improvement of the Edge Fusion Algorithm for Subspace of Remote Sensing Image

遥感图像进行边缘融合处理提高图像的识别和分析能力,受到数据在采集、传输、处理各个环节的影响。传统的遥感图像边缘融合算法采用统计形状模型重构算法,在遥感图像的帧点分布不均下,融合性能不好。本文提出一种基于子空间降噪和梯度边缘差拍控制的遥感图像边缘融合算法。先进行遥感图像采集和子空间降噪处理,采用梯度边缘差拍控制进行边缘亮度均衡处理,提高了遥感图像的色彩饱和度,对遥感图像融合进程进行高频、低频并行处理,提高运行速度。仿真结果表明,采用该算法实现遥感图像的边缘融合,输出信噪比较高,性能优越。

Edge fusion processing for remote sensing image can improve image recognition and analysis. Traditional edge fusion algorithm which uses statistical shape model reconstruction algorithm, has a poor result under the circumstance of uneven distribution of frame point. In this paper, we put forward a fusion algorithm based on subspace noise reduction and gradient edge beat control. It improved color saturation and running speed. Simulation results showed a high signal noise ratio （SNR） and superior performance.