Forstner特征点分类和精确定位方法探讨

针对无人机影像空三处理对特征点定位精度要求高的问题,基于Forstner点特征提取算子探讨一种Forstner特征点分类和精确定位方法。首先利用一定的方法判定Forstner特征点的类别(即角点、圆点、中间点中的一种),然后根据该特征点的类别进行精确定位。通过对一组无人机数据进行实验,试验表明:相比Forstner算子的初步定位,精确定位的特征点精度显著提高。

Forstner feature point classification and precise location method

Aiming at the problem of the UAV image aerial triangulation processing to the high accuracy of the feature point positioning,based on Forstner point feature extraction operator, a method of Forstner feature points classification and precise location is discussed.Firstly, the classification of Forstner feature points is determined by a certain method(A corner, dots, intermediate point in),Then according to the category of the feature points to accurately locate.Through experiments on a set of UAV data.The result showed that compared to the initial location of the Forstner operator,accurate positioning of the feature point accuracy is significantly improved.