摘要
高分辨率光学卫星遥感影像直线自动匹配算法的具体流程包括:进行影像尺度调整,对影像进行边缘提取,采用Hough变换提取直线;同时基于直线几何特征和灰度特征提取同名直线,利用Hough变换模型,将线转换为参数空间中的点,并建立转换模型;最后采用RANSAC算法剔除错误同名直线,以获取最终的匹配结果。实验结果表明,本文提出的算法能显著提高影像同名直线提取的数量和精度,特别是对异源影像具有抗噪能力强、鲁棒性好等特点。
This paper presented an automatic algorithm for line match in high resolution optical satellite images.Firstly,image scale was adjusted.Secondly,Hough transform was employed for line extracting after image edge extraction.Thirdly,line features from reference image and image to be rectified were matched by geometrical and gray feature.Fourthly,the extracted lines were conversed to the points in parameter space by using Hough transform model,and transform model was built then.At last,RANSAC were imbedded to eliminate mismatched lines.Experimental results verified that the algorithm could be more efficient on the quantity and accuracy of correspondence lines extraction,and richer in robustness and anti-noise,especially for multi-source high resolution optical images.
出处
《测绘科学》
CSCD
北大核心
2014年第5期121-125,共5页
Science of Surveying and Mapping
基金
国家自然基金(41101452)
辽宁省创新创业项目(201210147026)
