CAJ | 학술논문

飞行器视觉导航中图像的复杂性对特征点匹配提出了很高的要求.根据图像HSV色彩空间各分量固有的稳定性提出了一种新的特征点不变向量生成方法,以基于欧氏距离的最近邻准则作为特征点的相似度量将该算法提取的特征应用于图像特征点匹配;为了降低特征点描述向量的维数,提高匹配的实时性,采用了主成分分析(PCA)方法.针对误匹配问题,提出了一种利用惯性导航系统的输出信息进行误匹配特征点检测的方法.最后,通过实验证明,所提出的色彩匹配方法可以提高匹配的准确率,并且通过将PCA方法与上述方法结合不仅可以保持匹配的准确性还能降低计算的复杂度;所提出的误匹配检测方法可以较好的剔除误匹配点并能满足实时性的要求,这为视觉导航提供了一种可靠性更高的特征点匹配方法.
비행기시각도항중도상적복잡성대특정점필배제출료흔고적요구.근거도상HSV색채공간각분량고유적은정성제출료일충신적특정점불변향량생성방법,이기우구씨거리적최근린준칙작위특정점적상사도량장해산법제취적특정응용우도상특정점필배;위료강저특정점묘술향량적유수,제고필배적실시성,채용료주성분분석(PCA)방법.침대오필배문제,제출료일충이용관성도항계통적수출신식진행오필배특정점검측적방법.최후,통과실험증명,소제출적색채필배방법가이제고필배적준학솔,병차통과장PCA방법여상술방법결합불부가이보지필배적준학성환능강저계산적복잡도;소제출적오필배검측방법가이교호적척제오필배점병능만족실시성적요구,저위시각도항제공료일충가고성경고적특정점필배방법.
The complexity of images in vision navigation has brought great challenge to foature poinm matching method.The paper put forward a new local invariant descriptor which utilized the HSV(Hue,Saturation,and Value)color space information of images,and applied the method to feature points matching on the criterion of nearest neighbor based on Euclidean distance.In order to reduce the dimensions of the descriptor vectors the principal component analysis (PCA)was used to propose a new method to detect the mismatched points with the output of inertial navigation system.Finally,the experiments verified the validity of the matching method and pointed out that the method combined with PCA can not only keep the accuracy but also reduce the computation complicacy.The mismatched points eliminating method can detect the mismatched ones effectively and fulfill the requirement of realtime calculation well.The paper provides a more reliable points matching method for vision navigation.