Current Issue Cover


摘 要
目的 像对直线特征匹配是计算机视觉的重要研究内容,现有这类匹配方法均存在不同程度的误匹配问题,为此本文提出一种面向像对直线特征匹配的线特征矫正与提纯方法。方法 首先提取像对的边缘特征获得二值化边缘图,通过边缘梯度图及梯度矢量图(GVF)建立梯度引力图,采用直线检测方法提取像对的直线特征,并通过梯度引力图矫正直线位置。其次,采用点特征匹配结果计算像对极线,并结合直线匹配结果确定最后的局部校验特征区域,通过随机抽样一致小邻域范围内特征相似性校验直线匹配结果,从而剔除误匹配直线。结果 分别从LSD检测直线配准率、直线矫正后的配准率、提纯匹配结果配准率三方面对三组像对进行实验,从实验结果上看,平均匹配准确率提高近30个百分点。结论 采用本文方法可以大幅提高像对直线特征匹配的准确率,同时该方法可以很容易地对其它直线匹配结果进行校正与提纯,具备较高的实用性。
Line feature correction and purification for matching straight lines of image pair

Jia Di,Li Yuxiu,Zhao Mingyuan,zhu ning dan()

Objective Image matching of linear features is an important research content of computer vision. Existing matching methods have different degrees of mismatching. For this reason, a line feature correction and purification method for image matching of linear features is proposed. Method Firstly, the edge features of the pair are extracted to obtain the binarized edge map, the edge gradient map and the gradient vector flow (GVF) are used to establish the gradient gravitational map, the line detection method is used to extract the line features of the pair, and the linear position is corrected by using the gradient gravitational map. Secondly, the point feature matching result is used to calculate the image pair epipolar line, and the line matching result is combined to determine the final check local feature area. The mismatched straight line is eliminated by randomly sampling the feature matching in the small neighborhood to verify the straight line matching result. Result The three pairs of image pairs were tested from the LSD to detect the linear registration rate, the alignment rate after straight line correction, and the matching result of the purification matching results. From the experimental results, the average matching accuracy increased by nearly 30 percentage points. Conclusion The method of this paper can greatly improve the accuracy of matching the line features, and the method can easily correct and purify other line matching results, and has high practicability.