Multi-Camera Laser Line Imaging for Non-Flat Sheet Inspection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current imaging systems for quality control in machines that transport webs or sheets of material face challenges in detecting small defects over large areas due to deformation and varying surface-to-sensor distances, which are not effectively addressed by linear cameras.
Innovation Solution
The use of two-dimensional cameras with a laser line projection and overlapping fields of view to capture and stitch images, allowing for precise height measurement and correction of scale variations, enabling accurate quality control across the entire width of the support.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If linear cameras are used to capture sheets at high speed, then the ability to detect defects over wide sheets is improved, but the system becomes more expensive and vulnerable to image deformation when sheet speed varies
Solution Approach 1:
The system divides the wide sheet into multiple image slices captured by multiple cameras positioned side-by-side. Each camera captures a portion of the sheet width, and the images are stitched together to form a complete view. This segmentation allows using multiple standard cameras instead of expensive linear cameras while maintaining defect detection capability across the entire sheet width.
Solution Approach 2:
The invention uses standard two-dimensional matrix cameras instead of specialized linear cameras. By capturing multiple 2D images and stitching them together, the system creates a composite image that replicates the functionality of linear camera systems but using more readily available, cost-effective components.
2Measurement precision
If linear cameras are used to capture sheets at high speed, then defect detection capability is improved, but image deformation occurs when sheet speed varies from expected
Solution Approach 1:
The system uses a laser line projected onto the sheet surface and captured by the cameras to determine the actual height and position of the sheet. This height information is fed back into the image processing pipeline to correct for perspective distortions and scale variations. By continuously measuring and compensating for sheet position variations, the system maintains reliable defect detection even when sheet speed or position varies from expected conditions.
Solution Approach 2:
The invention dynamically adjusts image processing parameters based on measured sheet height. The height information derived from laser line analysis is used to modify the reproduction scale and perspective correction applied to each image slice. This parameter adaptation allows the system to compensate for varying sheet positions and maintain consistent measurement accuracy across different operating conditions.
3Area of stationary object
If projective optical systems are used in linear cameras to limit sensor size, then local variation of image scale occurs when sheet surface is not perfectly flat
Solution Approach 1:
The system introduces a laser line as an intermediary reference element projected onto the sheet surface. This laser line serves as a mediator between the sheet's actual 3D shape and the 2D camera capture. By analyzing the laser line's appearance in the captured images, the system can determine local sheet height and use this information to correct perspective distortions, thereby maintaining consistent image scale across non-flat surfaces.
Solution Approach 2:
The invention transitions from 2D image capture to 3D-aware processing by using the laser line to extract height information. The laser line provides depth information that allows the system to understand the sheet's 3D surface topology. This additional dimensional information is then used to correct 2D image scale variations, effectively bridging the gap between 2D capture and 3D reality.
4Area of stationary object
If multiple cameras are arranged side-by-side to capture the whole width of the support, then complete coverage is achieved, but image stitching complexity increases
Solution Approach 1:
The system replaces complex mechanical coordination between multiple cameras with a computational approach. Instead of requiring precise mechanical alignment and synchronization of multiple cameras, the invention uses image processing algorithms to automatically detect features, calculate transformations, and stitch images together. This substitution of mechanical precision requirements with computational processing simplifies the overall system integration while achieving complete width coverage.
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
This method provides an inexpensive and effective solution for quality control by ensuring accurate image representation and defect detection, even with non-flat surfaces, by combining images with overlapping fields of view and correcting for height variations, resulting in improved precision and reliability.
Implementation Method 1
The imaging system comprises a laser source that projects a laser line on a support surface
Data Source
AI summary
This invention is about an imaging system that uses conventional cameras and a single laser line to perform quality control at the output of a converting machine or a press. The system uses several cameras distributed over the width of the printed sheets. Thanks to the laser line, it can reconstruct the complete image of the printed matter even when the sheets are not perfectly flat or at varying height, compensating geometric as well as photometric distortions. The use of conventional cameras results in a cost effective system.