公路隧道衬砌裂缝图像质量增强方法研究

Research on image quality enhancement method of highway tunnel lining crack image

公路隧道在运营过程中承受多种环境作用,导致隧道衬砌裂缝广泛发育。采用搭载图像采集设备的移动式检测装备定检隧道裂缝已成为隧道运行维护的主要手段。但是,由于隧道衬砌裂缝尺度小、灰尘覆盖、涂料反射等影响,导致所获取的衬砌裂缝图像存在信噪比和对比度低的问题,难以采用智能算法提取裂缝,严重降低了隧道检测效率。文章以图像质量及裂缝目标位置为切入点,分别提出了基于均衡化理论的图像质量增强方法以及适用于图像边缘裂缝的边框增强方法,并采用主流的目标检测算法对所提算法开展了算法试验,结果表明:所提算法能够显著提升裂缝识别准确率,其在实际应用中具有较强的普适性。

Highway tunnels are subjected to a variety of harsh environments in the course of operation,resulting in widening tunnel lining cracks.The use of mobile inspection equipment with image acquisition equipment for the regular inspection of tunnel cracks has become a major means of the operation and maintenance.However,because of the small scale of tunnel lining cracks,dust coverage,paint reflection,etc.,the lining crack images from mobile inspection equipment have low signal-to-noise ratio and low contrast which makes it difficult to use intelligent algorithms for the crack extraction,seriously reducing the tunnel detection efficiency.This study,taking the image quality and crack location as the breakthrough point,proposes the image quality enhancement method based on the equilibrium theory and the border enhancement method applicable to image edge cracks respectively,and carries out the algorithm experiments for the algorithm in this study using the mainstream target detection algorithms.The results show that the algorithm in this study can significantly improve the crack recognition accuracy and have strong universality.