多源遥感数字图像复原优化研究

发布时间：2018-02-28 14:04

本文关键词： 多源遥感数字图像复原　出处：《计算机仿真》2017年09期 　论文类型：期刊论文

【摘要】：对多源遥感数字图像进行复原,可有效解决图像部分边缘丢失和纹理信息模糊的问题。对数字图像的复原,需要获得图像初始匹配对,计算出图像特征匹配强度,完成数字图像的复原。传统方法通过组建数字图像模糊近似度,构造模糊特征加权因子,但无法得出特征匹配强度,导致复原效果不理想。提出基于非线性最小二乘的多源遥感数字图像复原方法。结合SIFT算子检测多源遥感数字图像尺度空间极值点,计算出图像模糊特征点对比度值,将图像模糊特征点邻域像素的梯度方向分布特性作为各个图像模糊特征点指定方向参数,结合相关匹配理论获得图像初始匹配对,计算出图像特征匹配强度,以此为依据完成多源遥感数字图像复原。实验结果表明,所提方法能够有效提升多源遥感数字图像复原效果,合成多源遥感数字图像判读的可靠性较优。
[Abstract]:The restoration of multi-source remote sensing digital image can effectively solve the problems of image edge loss and texture information blur. For the restoration of digital image, we need to obtain the initial matching pairs of images, and calculate the intensity of image feature matching. The traditional method constructs the fuzzy feature weighting factor by constructing the fuzzy approximation of digital image, but it can not get the intensity of feature matching. The method of multi-source remote sensing digital image restoration based on nonlinear least squares is proposed. The spatial extremum of multi-source remote sensing digital image is detected by SIFT operator, and the contrast value of image fuzzy feature point is calculated. The gradient direction distribution of the pixels adjacent to the fuzzy feature points of the image is taken as the direction parameter of each fuzzy feature point of the image, and the initial matching pairs are obtained by combining the correlation matching theory, and the intensity of the image feature matching is calculated. The experimental results show that the proposed method can effectively improve the effect of multi-source remote sensing digital image restoration, and the reliability of multi-source remote sensing digital image interpretation is better.
【作者单位】：长安大学地质工程与测绘学院;
【基金】：陕西自然科学基金(2011JM5015)
【分类号】：TP751

【相似文献】