基于无人机与深度学习的少样本混凝土表面裂缝检测方法

Few-sample Concrete Surface Crack Detection Method Based on UAV and Deep Learning

混凝土表面裂缝检测是混凝土建筑安全评估和风险预警的重要手段。传统人工检测方法工作量大,且需考虑复杂环境下影响人身安全等因素。基于无人机的裂缝检测方法得到了应用,但当受不确定因素影响时,无人机无法采集足够的训练样本,限制了其检测性能。为此,基于无人机与深度学习,提出少样本条件下混凝土表面裂缝检测方法,采用主流深度学习网络Faster-RCNN和YOLOX,利用WBF算法将检测结果进行融合,有效弥补了像素级标签信息较少导致的检测性能下降。在少样本裂缝图像库及户外场地进行了试验测试,试验结果表明,在少样本条件下基于无人机与深度学习的裂缝检测方法性能得到有效提升,对裂缝检测的准确率达到58.67%。

Concrete surface crack detection is an important means of concrete building safety assessment and risk early warning.The traditional manual detection method has a large workload,and it is necessary to consider factors such as personal safety in complex environments.The crack detection method based on UAV has been applied,but when affected by uncertain factors,UAV cannot collect enough training samples,which limits their detection performance.Therefore,based on UAV and deep learning,a crack detection method for concrete surface under the condition of few samples is proposed.The mainstream deep networks Faster-RCNN and YOLOX are used,and the detection results are fused by WBF algorithm,which effectively alleviates the detection performance degradation caused by less pixel-level label information.Experimental tests were carried out in a few sample crack image library and outdoor sites.The test results show that the performance of the crack detection method based on UAV and deep learning is effectively improved under the condition of few samples,and the accuracy of crack detection is 58.67%.