Web Material Defect Detection Using Dual-Camera Spatial Verification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing web monitoring systems suffer from over-inclusion of reporting triggering events, leading to false positives, as they are unable to precisely identify defects in a web of material.
Innovation Solution
A method utilizing two cameras positioned to have overlapping fields of view, capturing images of the web from different angles, and determining the x-y position of potential defects within a margin of error to accurately identify probable defects.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a single camera is used to monitor the web for defects, then the system can detect triggering events, but it produces false positives by incorrectly identifying non-defect objects (such as water droplets) as defects
Solution Approach 1:
The patent transitions from a single-camera 2D monitoring system to a multi-camera 3D spatial verification system. By capturing images from multiple camera positions and triangulating the spatial coordinates of detected objects, the system determines whether objects lie on the web plane or are false positives (such as water droplets or debris in front of the web). This dimensional addition enables accurate defect identification while eliminating false positives.
2Reliability
If multiple cameras with overlapping fields of view are deployed, then false positives are reduced through spatial verification, but the device complexity increases
Solution Approach 1:
The monitoring system divides the web monitoring task into multiple camera zones, each responsible for capturing images from specific angles. The computing device segments the image processing workload by independently analyzing images from each camera, identifying candidates in each field of view, and then correlating results across cameras. This segmentation reduces the computational burden on any single processing unit while maintaining overall system reliability.
Data Source
AI summary
A method for identifying defects in a web of material translating along a travel path is provided. The method includes receiving a first image from a first camera; receiving a second image from a second camera; identifying a candidate as a possible defect within the first image; determining a first x-y position of the candidate relative to the first image; identifying the candidate within the second image; and determining the candidate is a probable defect based on the first x-y position and a second x-y position of the candidate within the second image. The first x-y position and the second x-y position predict the candidate is located on the web of material within a margin-of-error.