论文部分内容阅读
应用改进C-V模型,进行桥梁下部结构裂缝图像分割,通过裂缝截取、图像填充和旋转变换精确提取裂缝宽度。对不同光照条件下拍摄的在役混凝土桥梁结构裂缝图像,分别利用改进C-V模型算法、自适应阈值法、形态学算法、C-V模型以及Canny算法进行试验对比。分析结果表明:改进C-V模型算法误分率和运算时间最小,分别为3.02%与89ms;1 000幅桥梁结构裂缝图像试验对比显示裂缝检测准确率大于90.8%,裂缝宽度平均误差小于0.03mm。可见,改进算法可有效提高检测准确率,减少运算时间。
The improved C-V model was used to segment the crack image of the substructure of the bridge, and the crack width was accurately extracted by the crack interception, image filling and rotation transformation. The crack images of the existing concrete bridges under different light conditions were compared by using improved C-V model algorithm, adaptive threshold method, morphology algorithm, C-V model and Canny algorithm. The analysis results show that the error rate and operation time of the improved C-V model algorithm are the smallest, which are 3.02% and 89ms respectively. The comparison of 1000 crack images shows that the accuracy of crack detection is more than 90.8% and the average error of crack width is less than 0.03mm. Visible, improved algorithm can effectively improve the detection accuracy, reduce computing time.