【摘要】 从仿生学的角度提出了一种新的直线检测的尺蠖蠕行算法(cankerworm crawlalgorithm)。该算法首先根据梯度方向将边缘像素集合划分成为具有一定梯度范围的子集,然后采用蠕行策略在独立的边缘点的子集内进行直线提取。理论分析和实验结果证明:该方法大大减小了直线特征提取的复杂度,并且具有较好的直线检测效果。更多还原

【Abstract】 A new biologically inspired line detection algorithm called cankerworm-crawl algorithm (CCA) is proposed. CCA at first divides the edge pixels into some subsets, then extracts the line features in individual subsets with cankerworm-crawl strategy. Theoretical analysis and experimental results prove that CCA reduces the complexity of the line feature extraction and achieves good performance. 更多还原