3D Asset Modeling from Point Clouds for Automated Inspection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for inspecting and maintaining assets are time-consuming, labor-intensive, and often require human inspectors to access difficult or hazardous locations, leading to inefficiencies and potential under- or over-inspection.
Innovation Solution
A method involving the generation of a 3D model of an asset using feature matching on 2D images, followed by a space carving operation, facilitated by unmanned devices like drones, to create a detailed and accurate representation of the asset's condition, allowing for automated inspection and identification of defects.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If human inspectors manually inspect assets, then inspection accuracy and defect identification can be achieved, but the process becomes time-consuming and labor-intensive
Solution Approach 1:
The patent replaces manual human inspection with an automated system using unmanned aerial vehicles (drones) equipped with sensors and cameras. The drone captures images and videos of assets, and computer vision algorithms automatically analyze the data to detect defects, substituting the mechanical human inspection process with an automated technological system.
Solution Approach 2:
The patent creates a digital replica or model of the physical asset through photogrammetry and point cloud generation from drone-captured images. This digital copy allows for detailed inspection and analysis without requiring physical access to the actual asset, enabling efficient defect detection while reducing inspection time.
2Measurement precision
If human inspectors access difficult or hazardous locations on assets, then complete inspection coverage can be achieved, but safety risks and operational difficulty increase
Solution Approach 1:
The patent uses unmanned aerial vehicles to access difficult-to-reach and hazardous locations on assets without requiring human inspectors to physically access these areas. The drone can fly close to or around the asset, capturing images from angles that would be dangerous or impossible for human inspectors to obtain.
Solution Approach 2:
The drone acts as an intermediary between the inspector and the hazardous location. It collects data from dangerous areas and transmits it to a safe location for analysis, eliminating the need for human inspectors to directly access hazardous environments while maintaining complete inspection coverage.
3Reliability
If scheduled inspections are performed based on fixed timelines, then regular monitoring can be maintained, but over-inspection or under-inspection may occur
Solution Approach 1:
The automated inspection system provides continuous feedback through regular data collection and analysis. The system can detect changes in asset condition over time and trigger alerts when defects are identified, enabling condition-based maintenance scheduling that responds to actual asset status rather than following fixed timelines, thus preventing both over-inspection and under-inspection.
Data Source
AI summary
The present approach relates to the use of a point cloud of an object to initialize or seed a space carving technique used to generate a 3D model of the object. In one implementation, feature matching is performed on 2D images, with matched features constituting the points of a point cloud model. The point cloud generated in this manner, is one input of a foreground/background segmentation algorithm, which generates a set of segmented 2D images used by a space carving routine to generate a 3D model of the object.