价值工程
價值工程
개치공정
VALUE ENGINEERING
2013年
32期
208-208,209
,共2页
低空无人机遥感影像%Harris角点%SIFT特征%自动匹配%错误匹配剔除
低空無人機遙感影像%Harris角點%SIFT特徵%自動匹配%錯誤匹配剔除
저공무인궤요감영상%Harris각점%SIFT특정%자동필배%착오필배척제
low-altitude UAV remote sensing image%Harris corner%SIFT feature%automatic matching%mismatch rejecting
本文针对低空无人机遥感影像的特点,提出一种基于Harris角点和SIFT特征的自动匹配算法,并利用基于单应矩阵和核线几何双重约束的RANSAC相对定向剔除错误匹配。通过一组影像的自动匹配试验,结果具有较高的匹配正确率,验证了算法的有效性,可为区域网平差解算提供良好基础。
本文針對低空無人機遙感影像的特點,提齣一種基于Harris角點和SIFT特徵的自動匹配算法,併利用基于單應矩陣和覈線幾何雙重約束的RANSAC相對定嚮剔除錯誤匹配。通過一組影像的自動匹配試驗,結果具有較高的匹配正確率,驗證瞭算法的有效性,可為區域網平差解算提供良好基礎。
본문침대저공무인궤요감영상적특점,제출일충기우Harris각점화SIFT특정적자동필배산법,병이용기우단응구진화핵선궤하쌍중약속적RANSAC상대정향척제착오필배。통과일조영상적자동필배시험,결과구유교고적필배정학솔,험증료산법적유효성,가위구역망평차해산제공량호기출。
Aiming at the characteristics of low-altitude UAV remote sensing image, this paper proposed an automatic matching algorithm based on Harris corner and SIFT feature, using RANSAC relative orientation based on double constraint of homography and epipolar geometry to reject mismatch. Experiment was performed on a set of images, validated the validity of algorithm with high accuracy result, it can provide a good foundation for Block adjustment.