一种基于角点的直线段特征提取新方法

A Novel Line Segments Extracting Method Based on Conner

摘要: 直线特征的提取是图像分析、理解以及计算机视觉等技术的关键技术.然而, 由于直线端点定位难、直线易受损等因素, 使其成了一项挑战性很强的任务.由此提出一种基于SUSAN角点算子的直线段提取方法, 解决了直线两端不易定位的难题.实验结果表明, 该算法效率高、准确性好、抗噪能力强, 检测率达到90%以上.该方法提取的直线对目标的识别与三维重建工作具有重要的意义.

Abstract: The extracting of line segments is a hot in image processing research field since it is primary technology to image analysis, image understanding and computer vision, and so on.However, it is a difficult because the location is not easy of the line endpoint, and the line is easy broken.A line segments extracting algorithm is proposed based on SUSAN, and resolve the location difficult problem of line endpoints.The experiment results show that the algorithm is efficient and accuracy and the success detection rate is larger than 90%.The algorithm is important to object recognition and 3D reconstruction.