Technique on Vehicle Damage Assessment After Collisions Using Optical Radar Technology and Iterative Closest Point Algorithm

With the rapid advancement of technology, there is an increasing need for quick and accurate analysis of vehicle accident severity, determination of accident liability, assessment of vehicle damage extent, and efficient calculation of insurance compensation. This study proposes an innovative approac...

Full description

Saved in:
Bibliographic Details
Main Authors: Shih-Lin Lin, Yi-Hsuan Chen
Format: Article
Language:English
Published: IEEE 2024-01-01
Series:IEEE Access
Subjects:
Online Access:https://ieeexplore.ieee.org/document/10750172/
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850219508756643840
author Shih-Lin Lin
Yi-Hsuan Chen
author_facet Shih-Lin Lin
Yi-Hsuan Chen
author_sort Shih-Lin Lin
collection DOAJ
description With the rapid advancement of technology, there is an increasing need for quick and accurate analysis of vehicle accident severity, determination of accident liability, assessment of vehicle damage extent, and efficient calculation of insurance compensation. This study proposes an innovative approach for vehicle damage assessment by utilizing smartphones equipped with Light Detection and Ranging (LiDAR) technology. High-precision scans of vehicles are conducted in two stages: an initial comprehensive exterior scan of the undamaged vehicle, followed by a scan of the same model post-collision. The obtained point cloud data is processed using 3D reconstruction techniques to create virtual vehicle models. We apply the Iterative Closest Point (ICP) algorithm and Singular Value Decomposition (SVD) methods, along with a proposed deep learning neural network optimization model, to perform point cloud alignment between the pre-collision and post-collision vehicle models. The proposed method enhances the alignment accuracy by refining the transformation parameters, effectively handling nonlinear deformations caused by collisions. Key performance indicators, including Root Mean Square Error (RMSE), relative translation, and relative rotation angles, are used to quantify the extent of vehicle damage. Experimental results demonstrate that the proposed method significantly reduces the relative rotation from approximately 4.03° to 0.04° and the RMSE from about 1.27 units to 0.29 units compared to the traditional ICP and SVD methods, indicating a substantial improvement in alignment precision and damage assessment accuracy. The contributions of this study lie in integrating LiDAR technology with advanced point cloud processing algorithms and a deep learning optimization model for vehicle damage assessment, demonstrating high precision and cost-effectiveness. This method not only enhances the accuracy and efficiency of damage assessment but also reduces costs, offering significant practical value for rapid accident evaluation and insurance claims processing. By providing improved accuracy and reliability in damage assessments, this study significantly contributes to the fields of automotive safety, insurance, and repair.
format Article
id doaj-art-2c0f7d07e75341198cf0eb4c7005d04f
institution OA Journals
issn 2169-3536
language English
publishDate 2024-01-01
publisher IEEE
record_format Article
series IEEE Access
spelling doaj-art-2c0f7d07e75341198cf0eb4c7005d04f2025-08-20T02:07:20ZengIEEEIEEE Access2169-35362024-01-011217450717451810.1109/ACCESS.2024.349572110750172Technique on Vehicle Damage Assessment After Collisions Using Optical Radar Technology and Iterative Closest Point AlgorithmShih-Lin Lin0https://orcid.org/0000-0002-0234-8786Yi-Hsuan Chen1Graduate Institute of Vehicle Engineering, National Changhua University of Education, Changhua City, TaiwanGraduate Institute of Vehicle Engineering, National Changhua University of Education, Changhua City, TaiwanWith the rapid advancement of technology, there is an increasing need for quick and accurate analysis of vehicle accident severity, determination of accident liability, assessment of vehicle damage extent, and efficient calculation of insurance compensation. This study proposes an innovative approach for vehicle damage assessment by utilizing smartphones equipped with Light Detection and Ranging (LiDAR) technology. High-precision scans of vehicles are conducted in two stages: an initial comprehensive exterior scan of the undamaged vehicle, followed by a scan of the same model post-collision. The obtained point cloud data is processed using 3D reconstruction techniques to create virtual vehicle models. We apply the Iterative Closest Point (ICP) algorithm and Singular Value Decomposition (SVD) methods, along with a proposed deep learning neural network optimization model, to perform point cloud alignment between the pre-collision and post-collision vehicle models. The proposed method enhances the alignment accuracy by refining the transformation parameters, effectively handling nonlinear deformations caused by collisions. Key performance indicators, including Root Mean Square Error (RMSE), relative translation, and relative rotation angles, are used to quantify the extent of vehicle damage. Experimental results demonstrate that the proposed method significantly reduces the relative rotation from approximately 4.03° to 0.04° and the RMSE from about 1.27 units to 0.29 units compared to the traditional ICP and SVD methods, indicating a substantial improvement in alignment precision and damage assessment accuracy. The contributions of this study lie in integrating LiDAR technology with advanced point cloud processing algorithms and a deep learning optimization model for vehicle damage assessment, demonstrating high precision and cost-effectiveness. This method not only enhances the accuracy and efficiency of damage assessment but also reduces costs, offering significant practical value for rapid accident evaluation and insurance claims processing. By providing improved accuracy and reliability in damage assessments, this study significantly contributes to the fields of automotive safety, insurance, and repair.https://ieeexplore.ieee.org/document/10750172/LiDAR technologyhigh-precision scansvehicle damage assessmentiterative closest point (ICP) algorithmautomated photogrammetry
spellingShingle Shih-Lin Lin
Yi-Hsuan Chen
Technique on Vehicle Damage Assessment After Collisions Using Optical Radar Technology and Iterative Closest Point Algorithm
IEEE Access
LiDAR technology
high-precision scans
vehicle damage assessment
iterative closest point (ICP) algorithm
automated photogrammetry
title Technique on Vehicle Damage Assessment After Collisions Using Optical Radar Technology and Iterative Closest Point Algorithm
title_full Technique on Vehicle Damage Assessment After Collisions Using Optical Radar Technology and Iterative Closest Point Algorithm
title_fullStr Technique on Vehicle Damage Assessment After Collisions Using Optical Radar Technology and Iterative Closest Point Algorithm
title_full_unstemmed Technique on Vehicle Damage Assessment After Collisions Using Optical Radar Technology and Iterative Closest Point Algorithm
title_short Technique on Vehicle Damage Assessment After Collisions Using Optical Radar Technology and Iterative Closest Point Algorithm
title_sort technique on vehicle damage assessment after collisions using optical radar technology and iterative closest point algorithm
topic LiDAR technology
high-precision scans
vehicle damage assessment
iterative closest point (ICP) algorithm
automated photogrammetry
url https://ieeexplore.ieee.org/document/10750172/
work_keys_str_mv AT shihlinlin techniqueonvehicledamageassessmentaftercollisionsusingopticalradartechnologyanditerativeclosestpointalgorithm
AT yihsuanchen techniqueonvehicledamageassessmentaftercollisionsusingopticalradartechnologyanditerativeclosestpointalgorithm