基于边缘方向角的自适应尺度角点检测

Adaptive scale corner detection based on edge direction angle

在边缘方向角的基础上提出了一种新颖的角点检测方法。新模型基于:边缘在角点处必定发生弯折;在一定尺度范围内,角点两边的点均有方向角的最大趋向一致性;合理的放宽尺度,角点两边的点均有方向角最大趋向一致性增强。实验结果表明:新模型能够排除更多的伪角点,利用边缘方向向量判断角点的凸凹性;和其它多种角点检测模型相比,这种方法能够提供更多的角点信息,而且检测准确率高。

A novel method for image corner detection based on edge direction angle（EDA） is proposed.The first step is to extract edges from the original image using region segmentation and edge tracking.The proposed model is based on the following considerations：corners are the turning points of the edges;direction angle of the points on each side of corner tends to the greatest consistency in a certain scale scope;and increasing the scale scope reasonably,the greatest consistency of direction angle of points on each side of corner should be enhanced.More false corners can be removed by the new model.In addition,the concavity and convexity of detected corners can be judged by direction vector.Experiment results show that the new method can provide more accurate information about corners.