基于多旋翼无人机桥梁裂缝智能化识别分析  

Intelligent Identification of Bridge Cracks Based on Multi-rotor UAV

作　　者：常广利[1] 冷冰[1] 杨凯银 Chang Guangli;Leng Bing;Yang Kaiyin(Beihua University,Jilin 132013,China)

机构地区：[1]北华大学,吉林吉林132013

出　　处：《市政技术》2021年第10期46-48,共3页Journal of Municipal Technology

摘　　要：采用多旋翼无人机辅助拍摄桥梁裂缝病害图像,然后利用计算机图像视觉识别技术对桥梁裂缝病害进行了确认识别,再通过最大类间方差法对图像进行了处理,减小了误差概率。通过试验分析表明,首先对采集图像灰度化进行处理,然后进行直方图均衡化,再利用滤波器去噪,最后利用最大类间方差法进行识别,得到的处理结果误差率显著降低,该方法对图像的识别正确率达到了89.3%,并具有较高的鲁棒性,提高了桥梁裂缝检测效率。Multi-rotor UAV was adopted to take the disease image of bridge cracks.The bridge crack disease were confirmed and recognized by computer image visual recognition technology.The image is processed by maximization interclass variance method(OTSU)to reduce the error probability.The test analysis shows that firstly,the collected image is grayed processing,then histogram equalized,rdenoised with filter,finally identified by OTSU.The processing result error rate is decreased significantly and the recognition accuracy rate is 89.3%with has high robustness.It improves the detection efficiency.

关 键 词：桥梁 裂缝 识别 无人机 最大类间方差法 

分 类 号：U418.6[交通运输工程—道路与铁道工程]