图像识别技术在水工隧洞收敛变形监测中的应用

Application of Image Recognition Technology in Monitoring Convergence Deformation of Hydraulic Tunnel

基于图像识别技术的水工隧洞收敛变形监测技术,是实现水工隧洞全断面收敛变形简便、易操作的自动化监测的重要手段。在实践应用中,基于图像识别技术实现了靶标自动识别和位移自动测读,测量精度可控制在2 mm之内,通过试验获得了最佳靶标直径和适宜的拍摄距离参数,靶标适宜直径范围为150~250 mm,适宜拍摄距离根据洞径选择3~10 m,并结合某施工期水工隧洞工程进行实际应用,取得了良好的应用效果,可显著减少人工监测工作量,大大提高监测工作效率,研究结果可为隧洞安全监测提供重要参考和技术支撑。

The monitoring technology of hydraulic tunnel convergence deformation based on image recognition technology is an important means to realize the automatic monitoring of the convergence and deformation of the whole section in hydraulic tunnel.In practice,automatic target recognition and displacement measurement have been achieved based on image recognition technology,and the measurement accuracy could be controlled within 2 mm.The optimal target diameter and suitable shooting distance parameters were obtained through experiments.The suitable target diameter ranges from 150 to 250 mm,and the suitable shooting distance is in the range of 3-10 m according to the hole diameter.In combination with the practical application of a hydraulic tunnel project during the construction period,good application results have been achieved,which can significantly reduce the workload of manual monitoring and greatly improve the monitoring efficiency.The research results can provide important reference and technical support for tunnel safety monitoring.