Eine Plattform für die Wissenschaft: Bauingenieurwesen, Architektur und Urbanistik
Eine Plattform für die Wissenschaft: Bauingenieurwesen, Architektur und Urbanistik
Pavement defect detection method based on multi-scale gradient selection self-adaption
The invention discloses a pavement defect detection method based on multi-scale gradient selection self-adaption, and belongs to the technical field of defect detection, and the method comprises the following steps: S1, constructing a multi-scale feature selection memory bank; s2, performing coupling hypersphere adaptive training by using a to-be-tested sample and a sample in a memory bank; and S3, performing defect scoring on the to-be-scored sample. According to the pavement defect detection method based on multi-scale gradient selection self-adaption, a lightweight memory bank with more representative characteristics is constructed, so that the defect positioning problem in a complex pavement scene can be solved, and the data distribution deviation between a large-scale natural data set and a target defect data set can be relieved; and more global semantic information can be provided for the model, and the capability of detecting and positioning fine defects of the model in a weak texture scene can be improved.
本发明公开了一种基于多尺度梯度选择自适应的路面缺陷检测方法，属于缺陷检测技术领域，包括以下步骤：S1、构建多尺度特征选择内存库；S2、使用待测样本以及内存库中样本进行耦合超球自适应训练；S3、对待评分样本进行缺陷评分。本发明采用上述的一种基于多尺度梯度选择自适应的路面缺陷检测方法，构建了轻量级且特征更具有代表性的内存库可以解决复杂路面场景下的缺陷定位问题，还可缓解大规模自然数据集与目标缺陷数据集的数据分布偏差，还能够为模型提供更多的全局语义信息和提升模型在弱纹理场景下检测与定位细微缺陷的能力。
Pavement defect detection method based on multi-scale gradient selection self-adaption
The invention discloses a pavement defect detection method based on multi-scale gradient selection self-adaption, and belongs to the technical field of defect detection, and the method comprises the following steps: S1, constructing a multi-scale feature selection memory bank; s2, performing coupling hypersphere adaptive training by using a to-be-tested sample and a sample in a memory bank; and S3, performing defect scoring on the to-be-scored sample. According to the pavement defect detection method based on multi-scale gradient selection self-adaption, a lightweight memory bank with more representative characteristics is constructed, so that the defect positioning problem in a complex pavement scene can be solved, and the data distribution deviation between a large-scale natural data set and a target defect data set can be relieved; and more global semantic information can be provided for the model, and the capability of detecting and positioning fine defects of the model in a weak texture scene can be improved.
本发明公开了一种基于多尺度梯度选择自适应的路面缺陷检测方法，属于缺陷检测技术领域，包括以下步骤：S1、构建多尺度特征选择内存库；S2、使用待测样本以及内存库中样本进行耦合超球自适应训练；S3、对待评分样本进行缺陷评分。本发明采用上述的一种基于多尺度梯度选择自适应的路面缺陷检测方法，构建了轻量级且特征更具有代表性的内存库可以解决复杂路面场景下的缺陷定位问题，还可缓解大规模自然数据集与目标缺陷数据集的数据分布偏差，还能够为模型提供更多的全局语义信息和提升模型在弱纹理场景下检测与定位细微缺陷的能力。
Pavement defect detection method based on multi-scale gradient selection self-adaption
一种基于多尺度梯度选择自适应的路面缺陷检测方法
LI ZUOYONG (Autor:in) / ZHENG XIANGPAN (Autor:in) / ZHANG YIJIE (Autor:in) / LIU WEIXIA (Autor:in) / LAI TAOTAO (Autor:in)
21.11.2023
Patent
Elektronische Ressource
Chinesisch
IPC:
G06T
Bilddatenverarbeitung oder Bilddatenerzeugung allgemein
,
IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
/
E01C
Bau von Straßen, Sportplätzen oder dgl., Decken dafür
,
CONSTRUCTION OF, OR SURFACES FOR, ROADS, SPORTS GROUNDS, OR THE LIKE
/
G06N
COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS
,
Rechnersysteme, basierend auf spezifischen Rechenmodellen
Road pavement and defect detection method for read pavement
| Europäisches Patentamt | 2020
Pavement paving equipment with pavement defect detecting and self-repairing functions
| Europäisches Patentamt | 2023
Analyses of Frost Susceptible Flexible Pavement Adaption for Climate Change
| British Library Conference Proceedings | 2023
Parallax-based pavement defect detection system, method and program product
| Europäisches Patentamt | 2025