Eine Plattform für die Wissenschaft: Bauingenieurwesen, Architektur und Urbanistik
Eine Plattform für die Wissenschaft: Bauingenieurwesen, Architektur und Urbanistik
Roadbed compaction quality rapid automatic detection method based on three-dimensional geological radar
The invention relates to a rapid and automatic roadbed compaction quality detection method based on a three-dimensional geological radar, and the method comprises the steps: collecting different filling layers of at least three soil samples of each layer in backfilling and spectrum images of the moisture content of different layers through employing a three-dimensional geological radar technology, including front images and back images of the soil samples under different collection distances; amplitude, frequency, phase, propagation velocity, conductivity, dielectric constant and the like are used for auxiliary analysis with the front wave spectrum image, the back wave spectrum image and the wave spectrum image after data augmentation, and the accuracy of a three-dimensional geological radar recognition result is improved. By using the constructed convolutional neural network, the spectrum image and data can be effectively analyzed, the misjudgment of detecting the blurred image by the three-dimensional geological radar is avoided, and the accuracy of image recognition is improved.
本发明为一种基于三维地质雷达的路基压实质量快速自动检测方法，该检测方法利用三维地质雷达技术采集回填中每一层至少三个土样的不同填筑层以及不同层含水率的波谱图像，包括不同采集距离下的土样的正面图像、反面图像，通过振幅、频率、相位、传播速度、电导率以及介电常数等来与正面波谱图像和反面波谱图像、数据增广后的波谱图像辅助分析，提高三维地质雷达识别结果的准确率。利用构建的卷积神经网络，能够对波谱图像和数据进行有效分析，避免了三维地质雷达检测模糊图像的误判，提高了对图像识别的精确度。
Roadbed compaction quality rapid automatic detection method based on three-dimensional geological radar
The invention relates to a rapid and automatic roadbed compaction quality detection method based on a three-dimensional geological radar, and the method comprises the steps: collecting different filling layers of at least three soil samples of each layer in backfilling and spectrum images of the moisture content of different layers through employing a three-dimensional geological radar technology, including front images and back images of the soil samples under different collection distances; amplitude, frequency, phase, propagation velocity, conductivity, dielectric constant and the like are used for auxiliary analysis with the front wave spectrum image, the back wave spectrum image and the wave spectrum image after data augmentation, and the accuracy of a three-dimensional geological radar recognition result is improved. By using the constructed convolutional neural network, the spectrum image and data can be effectively analyzed, the misjudgment of detecting the blurred image by the three-dimensional geological radar is avoided, and the accuracy of image recognition is improved.
本发明为一种基于三维地质雷达的路基压实质量快速自动检测方法，该检测方法利用三维地质雷达技术采集回填中每一层至少三个土样的不同填筑层以及不同层含水率的波谱图像，包括不同采集距离下的土样的正面图像、反面图像，通过振幅、频率、相位、传播速度、电导率以及介电常数等来与正面波谱图像和反面波谱图像、数据增广后的波谱图像辅助分析，提高三维地质雷达识别结果的准确率。利用构建的卷积神经网络，能够对波谱图像和数据进行有效分析，避免了三维地质雷达检测模糊图像的误判，提高了对图像识别的精确度。
Roadbed compaction quality rapid automatic detection method based on three-dimensional geological radar
一种基于三维地质雷达的路基压实质量快速自动检测方法
XIAO CHENGZHI (Autor:in) / XU QIANWEN (Autor:in) / YIN JIANGUANG (Autor:in) / PAN PENG (Autor:in) / DING LUQIANG (Autor:in)
01.11.2024
Patent
Elektronische Ressource
Chinesisch
Roadbed compaction quality detection system based on geological radar
| Europäisches Patentamt | 2023
Roadbed compaction quality detection device and method
| Europäisches Patentamt | 2021
Roadbed compaction quality detection device and detection method thereof
| Europäisches Patentamt | 2024
| Europäisches Patentamt | 2022