Disaster-Site Crack Assessment Using Skeletonization and LiDAR
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing disaster investigation robots face challenges in accurately identifying crack start and end points, determining crack location and size, and assessing structural damage due to instability in constrained environments, which complicates real-time risk assessment and collapse prediction.
Innovation Solution
A method using skeletonization techniques to analyze crack images, combined with LiDAR-based depth changes, to identify and visualize crack locations and scales, and a learning model to predict crack expansion risk, enabling rapid and accurate damage assessment.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If LiDAR sensors are used to measure distances and detect cracks, then measurement capability is improved, but stability deteriorates due to vibration in constrained environments
Solution Approach 1:
The patent segments the crack detection process into multiple independent components: LiDAR-based depth mapping, vision-based crack image acquisition, skeletonization processing, and branch point analysis. This segmentation allows each component to be optimized independently, with the LiDAR providing depth information while vision sensors capture surface crack details, compensating for LiDAR's vibration sensitivity.
Solution Approach 2:
The patent introduces skeletonization as an intermediary processing step between raw crack images and final crack characterization. The skeletonization algorithm converts complex crack images into simplified skeletal representations, making crack start points, end points, and branch points easily identifiable. This intermediary processing enhances measurement precision without requiring additional stable sensor hardware.
2Loss of information
If feature points are extracted from camera images to identify cracks, then crack location identification is improved, but measurement precision deteriorates because start and end points cannot be accurately identified
Solution Approach 1:
The patent applies skeletonization processing that goes beyond simple feature point extraction by creating a complete skeletal representation of the crack. This excessive processing action ensures that all crack characteristics including start points, end points, and branch points are captured, eliminating information loss while providing precise location data for crack assessment.
Solution Approach 2:
The patent combines 2D vision-based crack images with 3D LiDAR depth information to create a multi-dimensional crack characterization. The skeletonization process operates on the 2D image plane to identify precise start and end points, while LiDAR provides depth dimension data, together enabling accurate crack location and size measurement that overcomes the limitations of single-dimension approaches.
3Measurement precision
If comprehensive crack analysis is performed to assess structural damage, then damage assessment accuracy is improved, but processing time increases
Solution Approach 1:
The patent performs skeletonization processing as a preliminary step before detailed crack analysis. By converting crack images into skeletal representations first, the system pre-identifies all critical features including start points, end points, and branch points. This preliminary action organizes the data structure to facilitate rapid subsequent analysis, reducing overall processing time while maintaining comprehensive damage assessment accuracy.
Solution Approach 2:
The patent extracts key crack characteristics (start points, end points, branch points) from the complete crack image data using skeletonization. This extraction process separates the essential diagnostic information from the redundant pixel data, enabling rapid assessment of crack severity and structural damage risk without processing the entire high-resolution image dataset in detail.
4Loss of information
If multiple sensors are used to collect comprehensive data, then information completeness is improved, but device complexity increases
Solution Approach 1:
The patent designs the sensor module with multi-functional sensors that perform multiple roles. The LiDAR sensor not only provides depth mapping for 3D reconstruction but also contributes to crack detection through depth change analysis. The vision sensor captures both general facility images and specific crack images. This multi-functionality reduces the need for dedicated specialized sensors, lowering device complexity while maintaining information completeness.
Solution Approach 2:
The patent merges LiDAR depth information with vision-based crack image information into a unified crack analysis framework. The skeletonization process integrates both data types to produce comprehensive crack characterization including location, size, depth, and structural risk assessment. This merging approach consolidates multiple sensor inputs into a single coherent analysis pipeline, reducing system complexity compared to separate independent analysis systems.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Enables rapid and optimized assessment of disaster site damage and collapse risk, improving crack prediction accuracy and facilitating real-time risk management in unstable environments.
Implementation Method 1
LiDAR sensors operate by measuring incidents and reflected light to calculate distances to target objects
Data Source
AI summary
The present disclosure relates to a method for assessing the degree of damage to objects at disaster sites using skeletonization techniques, including the steps of moving a position of an investigation robot for information analysis of facility in a disaster site space to a first location by using a sensor module equipped in the investigation robot and including at least one of a LiDAR sensor, an IMU sensor or at least one vision sensor; acquiring sensing information corresponding to the first location based on SLAM; identifying a facility segment of a first space based on the sensing information; acquiring visual crack identification information corresponding to the facility segment, analyzed from vision image information of the sensing information, and unit crack information corresponding to the visual crack identification information; determining a crack expansion risk corresponding to the unit crack information.