Automated Fiber Placement Composite Defect Detection via Point Cloud Grayscale Conversion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for detecting surface defects in automated fiber placement composites are inefficient and inaccurate, relying on visual inspection and being susceptible to light, which limits their applicability and precision in high-tech fields like aerospace and automobiles.
Innovation Solution
A method involving the acquisition of surface point clouds, fitting a plane, calculating distances, generating a grayscale image using an oriented bounding box, and employing a pre-trained semantic segmentation network to recognize defects, allowing for high-accuracy detection and visualization of defects like warping, wrinkling, and twisting.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If eye detection is used to detect surface defects, then the detection method is simple to implement, but the detection efficiency is low and accuracy declines due to visual fatigue
Solution Approach 1:
The patent replaces the mechanical/visual inspection system with a computational system. Specifically, it converts 3D point cloud data into 2D images through projection, then uses image processing algorithms and deep learning models to automatically detect and classify surface defects. This substitution eliminates human visual inspection, thereby resolving the contradiction between operational simplicity and detection efficiency.
2Adaptability or versatility
If color images are used for defect detection, then the detection approach is widely applicable, but it is susceptible to light conditions and cannot accurately evaluate defect levels
Solution Approach 1:
The patent changes the fundamental parameter of data representation from color images to grayscale images derived from point cloud depth information. By converting 3D spatial coordinates into 2D projections with grayscale values representing depth or distance, the method eliminates sensitivity to lighting conditions while maintaining the ability to detect and evaluate surface defects at different levels.
3Device complexity
If manual visual inspection is used, then the equipment requirements are minimal, but long-term operation causes visual fatigue and reduced detection accuracy
Solution Approach 1:
The patent substitutes manual visual inspection with an automated computer vision system that processes 3D point cloud data. The system converts point clouds to 2D images and applies deep learning models to detect and classify defects, thereby maintaining high detection accuracy without the visual fatigue that plagues manual inspection methods.
Data Source
AI summary
A method for detecting and recognizing surface defects of an automated fiber placement composite based on an image converted from point clouds, including: acquiring a surface point cloud of the automated fiber placement composite; fitting a plane to surface point data; calculating a distance from each point of the surface point cloud to a fitted plane; enveloping the surface point cloud by OBB, and generating a grayscale image according to the OBB and the distance; constructing a pre-trained semantic segmentation network for defect of fiber placement, and inputting the grayscale image to segment and recognize defect areas thereon; mapping a segmentation result output by the semantic segmentation network to the point cloud followed by defect evaluation and visualization.