【摘要】 为了克服传统的遥感图像匹配算法低鲁棒性的缺陷,本文对传统序贯相似性检测算法进行改进.首先,定义噪声敏感度,自适应判定像素点是否为噪声,并对噪声点进行预处理;其次,计算每个像素点的梯度方向,与初始设定的梯度方向阀值对比,离散待取样像素,减少非匹配区域的计算量;最后,对经过预处理的图像进行图像匹配.实验结果表明,经过改进的算法具有较强的抗噪能力和较快的运算速度.更多还原

【Abstract】 In order to overcome the defects of low robustness in remote sensing image matching,this paper makes some improvement to traditional SSDA algorithm.First,it define noise sensitivity,try to decide whether a pixel is a noise or not adaptively,then make some pretreatment to those pixels;Second,it calculates each pixel’s gradient direction,and make a comparison between the gradient direction and the threshold,thus to reduce the computation in non-matching region;Third,try to match the Pretreated Image.Experimental results show that,the new algorithm has a high robustness and it is faster than the traditional one.更多还原