测绘学报
測繪學報
측회학보
Acta Geodaetica et Cartographica Sinica
2015年
10期
1132-1141
,共10页
均值漂移%自适应带宽%运动结构特征%全景影像匹配
均值漂移%自適應帶寬%運動結構特徵%全景影像匹配
균치표이%자괄응대관%운동결구특정%전경영상필배
mean shift%adaptive bandwidth%structure from motion feature%panoramic image matching
以运动结构特征为约束条件的序列影像匹配,是基于多变量核密度函数,采用非参数均值漂移方法估计最优局部运动相似性结构特征的过程。核密度函数的带宽大小决定了匹配方法的收敛速度和精度。本文提出了一种可变带宽的自适应运动结构特征的车载全景序列影像匹配方法。首先以采样点在空间域和光流域的局部空间结构定义自适应的带宽矩阵。采用局部光流特征向量的距离加权法,描述光流域上运动相似性结构特征的松弛扩散过程。然后给出自适应多变量核密度函数的表达形式,并探讨了均值漂移向量的求解、终止条件以及种子点的选择方法。最后融合多尺度SIFT描述特征与运动结构特征,建立统一的全景影像匹配框架。试验选择车载移动测量系统获取的城市球全景序列影像,结果表明在内点率变化、物方尺度变化等情况下,本文方法可以实现自适应运动结构特征的相似性度量,提高匹配的正确点数和匹配率,算法表现出较强的稳键性。
以運動結構特徵為約束條件的序列影像匹配,是基于多變量覈密度函數,採用非參數均值漂移方法估計最優跼部運動相似性結構特徵的過程。覈密度函數的帶寬大小決定瞭匹配方法的收斂速度和精度。本文提齣瞭一種可變帶寬的自適應運動結構特徵的車載全景序列影像匹配方法。首先以採樣點在空間域和光流域的跼部空間結構定義自適應的帶寬矩陣。採用跼部光流特徵嚮量的距離加權法,描述光流域上運動相似性結構特徵的鬆弛擴散過程。然後給齣自適應多變量覈密度函數的錶達形式,併探討瞭均值漂移嚮量的求解、終止條件以及種子點的選擇方法。最後融閤多呎度SIFT描述特徵與運動結構特徵,建立統一的全景影像匹配框架。試驗選擇車載移動測量繫統穫取的城市毬全景序列影像,結果錶明在內點率變化、物方呎度變化等情況下,本文方法可以實現自適應運動結構特徵的相似性度量,提高匹配的正確點數和匹配率,算法錶現齣較彊的穩鍵性。
이운동결구특정위약속조건적서렬영상필배,시기우다변량핵밀도함수,채용비삼수균치표이방법고계최우국부운동상사성결구특정적과정。핵밀도함수적대관대소결정료필배방법적수렴속도화정도。본문제출료일충가변대관적자괄응운동결구특정적차재전경서렬영상필배방법。수선이채양점재공간역화광류역적국부공간결구정의자괄응적대관구진。채용국부광류특정향량적거리가권법,묘술광류역상운동상사성결구특정적송이확산과정。연후급출자괄응다변량핵밀도함수적표체형식,병탐토료균치표이향량적구해、종지조건이급충자점적선택방법。최후융합다척도SIFT묘술특정여운동결구특정,건립통일적전경영상필배광가。시험선택차재이동측량계통획취적성시구전경서렬영상,결과표명재내점솔변화、물방척도변화등정황하,본문방법가이실현자괄응운동결구특정적상사성도량,제고필배적정학점수화필배솔,산법표현출교강적은건성。
Panoramic image matching method with the constraint condition of local structure from motion similarity feature is an important method ,the process requires multivariable kernel density estimations for the structure frommotion feature used nonparametric mean shift .Proper selection of the kernel bandwidth is a critical step for convergence speed and accuracy of matching method .Variable bandwidth with adaptive structure from motion feature for panoramic image matching method has been proposed in this work .First the bandwidth matrix is defined using the locally adaptive spatial structure of the sampling point in spatial domain and optical flow domain .The rel axation diffusion process of structure from motion simil arity feature is described by distance weighting method of local optical flow feature vector .Then the expression form of adaptive multivari ate kernel density function is given out , and discusses the solution of the mean shift vector ,termination conditions ,and the seed point selection method .The final fusions of multi‐scale SIFT the features and structure features to establish a unified panoramic image matching framework .The sphere panoramic images from vehicle‐borne mobile measurement system are chosen such that a comparison analysis between fixed bandwidth and adaptive bandwidth is carried out in detail .The results show that adaptive bandwidth is good for case with the inlier ratio changes and the object space scale changes .The proposedmethodcanrealizetheadaptivesimilaritymeasureofstructurefrommotionfeature,improvesthe correct matching points and matching rate ,experimental results have shown our method to be robust .