Assessment the Accuracy of Crack Detection Derived from Smartphone LiDAR and TLS Dataset

• Published in: 2024 IEEE INTERNATIONAL CONFERENCE ON AUTOMATIC CONTROL AND INTELLIGENT SYSTEMS, I2CACIS 2024
• Main Authors: Syafie, Fahmi Syazwan; Khalid, Nafisah; Tahar, Khairul Nizam; Naharudin, Nabilah; Jalil, Abdul Hadi Abdul; Kamaruddin, Amirul Hakim
• Format: Proceedings Paper
• Language: English
• Published: IEEE 2024
• Subjects: Automation & Control Systems; Computer Science
• Online Access: https://www-webofscience-com.uitm.idm.oclc.org/wos/woscc/full-recordWOS:001308267400022

The feature of LiDAR technology in smartphones commonly used to improve the photos by adding depth scanning techniques. However, the uses of this technology are tremendous. In the field of engineering and surveying, LiDAR has been widely used in the past decades for various applications and one of the applications is examining crack detection. Thus, this study is intended to assess the accuracy of building surface crack detection derived from smartphone LiDAR and Terrestrial Laser Scanner (TLS). Ridge detection tool in open-source Fiji ImageJ software was used to detect cracks from smartphone LiDAR and TLS dataset. The findings show that both smartphone LiDAR and TLS data have the capability in detecting accurate cracks on building surfaces with the highest accuracy are 98%. However, the smartphone LiDAR sensor paired with a 12 megapixels camera has improved the overall outcome of cracks detection on building surfaces. LiDAR has the lowest RMSE of +/- 1.719 cm in crack length and highest of +/- 2.201 cm while the RMSE of crack length measured using TLS range from +/- 3.212 cm and highest at +/- 13.646 cm. The findings show that with the limitation of scanning area, smartphone LiDAR offers great potential for building inspection especially for building surface crack detection.