Quantitative evaluation of peeling and delamination on infrastructure surfaces by laser signal and image processing of 3D point cloud data
Autor: | Yoshifumi Nagata, Kazutomo Yamamoto, Takahiro Yamaguchi, Tetsuya Ishida, Tsukasa Mizutani, Yuuya Yamaguchi, Suzuki Kiyoshi, Tomoshi Kudo, Hinari Kawamura, Tomoaki Tokuno |
---|---|
Rok vydání: | 2022 |
Předmět: |
Signal processing
Computer science business.industry Delamination Point cloud Image processing Building and Construction Signal symbols.namesake Control and Systems Engineering symbols Computer vision Artificial intelligence Hilbert transform business Envelope (mathematics) ComputingMethodologies_COMPUTERGRAPHICS Civil and Structural Engineering Mobile mapping |
Zdroj: | Automation in Construction. 133:104023 |
ISSN: | 0926-5805 |
Popis: | In this article, an algorithm for the automatic and quantitative evaluation of peeling and delamination on infrastructure surfaces from laser 3D point cloud data is proposed using state-of-the-art signal and image processing methodologies. A Mobile Mapping System (MMS) enables road administrators to collect the geometry of tunnel and road surfaces highspeed. The algorithm consists of two steps: 1D signal processing and 2D image processing. The peaks of anomalies are extracted by the envelope of a signal using a Hilbert transform. Anomalies detected in each direction are integrated into a map and smoothed by a morphology transform. The algorithm was validated by an experimental tunnel and road surface data. The proposed algorithm accurately and efficiently detects delamination and estimates the areas and depth of peeling and potholes by a real time analysis approach. |
Databáze: | OpenAIRE |
Externí odkaz: |