基于改进Otsu的桥梁裂缝检测方法

Bridge Crack Detection Method Based on Improved Otsu

摘要:

针对传统的桥梁裂缝检测方法效率低、主观性强,以及费时费力等缺点,提出一种基于改进Otsu的桥梁裂缝检测方法。在目标方差前加入偏重参数和背景倾向系数,利用灰度值梯度累积量获取峰值信息,进而保证在单峰情况下阈值始终在峰值左侧,然后通过灰度值累积量适当调整阈值,最终实现自适应桥梁裂缝检测。实验结果表明,改进算法能够有效检测出桥梁裂缝,且相比于EW,WOV,VE等算法,所提算法获得的误分类值更接近0,缺陷检测率更接近1,有更好的检测效果。

Abstract:

In response to the shortcomings of traditional bridge crack detection methods such as low efficiency, subjectivity, and time-consuming, an improved Otsu bridge crack detection method is proposed. The bias parameter and background tendency coefficient were added before the target variance, and the peak information was obtained by using the gray value gradient cumulant, so as to ensure that the threshold was always on the left side of the peak in the case of a single peak. And then the threshold was adjusted appropriately by the gray value cumulant to finally achieve adaptive bridge crack detection. The experimental results show that the improved algorithm in this paper can effectively detect bridge cracks. Compared to improved algorithms such as EW, WOV, VE, etc., the misclassification value obtained by the improved algorithm in this paper is closer to 0, and the defect detection rate is closer to 1, resulting in better detection performance.