通过将深度学习框架与卷积神经网络相结合,研究桥梁裂检测的效率
Kaifeng Ma1, Xiang Meng1, Mengshu Hao1
1College of Surveying and Geo-Informatics, North China University of Water Resources and Electric Power, Zhengzhou 450046, China.
Sensors (Basel, Switzerland)
|August 26, 2023
概括
这项研究优化了深度学习框架和卷积神经网络,以有效地检测桥梁裂. 研究确定了特定的模型和框架组合,可以显著提高识别结构损坏的准确性和效率.
科学领域:
- 土木工程 土木工程是指土木工程.
- 计算机科学 计算机科学
- 人工智能的人工智能
背景情况:
- 桥梁健康监测对于基础设施安全至关重要.
- 使用深度学习的自动桥梁裂检测带来了重大挑战.
- 现有的方法需要优化,以提高准确性和效率.
研究的目的:
- 研究将深度学习框架 (DLF) 与卷积神经网络 (CNN) 结合起来,以检测桥梁裂的有效性.
- 确定最佳的DLF和CNN模型组合,以加强桥梁健康监测.
- 评估各种CNN架构和深度学习框架的裂检测性能.
主要方法:
- 使用了2068个桥梁裂图像的数据集.
- 图像被随机分为训练,验证和测试组 (8:1:1比率).
- 在PyTorch,TensorFlow2和Keras框架中使用Faster R-CNN,SSD,YOLO-v5(x),U-Net和PSPNet模型进行了实验.
主要成果:
- 更快的R-CNN和SSD (Keras) 在物体检测中实现了高F1值 (0.76,0.67).
- YOLO-v5(x) (TensorFlow2) 达到了一个F1值为0.67.
- 在语义细分方面,U-Net (PyTorch) 实现了98.37%的准确性 (AC).
- 在语义细分方面,PSPNet (TensorFlow2) 实现了97.86%的交流.
结论:
- 确定了DLF和CNN在桥梁裂检测中的最佳合参数.
- 该研究产生了更准确,更有效的DLF和CNN模型,用于实际的桥梁健康检测.
- 这些发现为基础设施维护和安全提供了重要的实际价值.
相关概念视频
01:20Microcracking in Concrete
149
Microcracking in concrete refers to the tiny cracks that can form within the material even before any external load is applied. These microcracks typically occur at the interface between the coarse aggregate and the hydrated cement paste, often as a result of differential volume changes prompted by variations in stress-strain behavior, as well as thermal and moisture movement. Initially, these microcracks remain stable and do not grow substantially until the concrete is stressed to about 30...
149
01:28Types of Non-structural Cracks in Concrete
186
Non-structural cracks are primarily of three types: plastic, early-age thermal, and drying shrinkage cracks. Plastic cracks are further classified into plastic shrinkage cracks and plastic settlement cracks.
Plastic shrinkage cracks typically form within hours after the concrete is poured. The concrete's surface dries faster than the bottom, creating tensile stress that the still-plastic concrete cannot withstand, leading to diagonal or randomly patterned cracks on the concrete surface.
Plastic shrinkage cracks typically form within hours after the concrete is poured. The concrete's surface dries faster than the bottom, creating tensile stress that the still-plastic concrete cannot withstand, leading to diagonal or randomly patterned cracks on the concrete surface.
186