Automated Image Inspection Using GPU Parallel Processing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current automated inspection methods for detecting image defects in printed products are inefficient, particularly in variable data printing, where each image must be inspected against a unique reference, leading to significant delays or increased costs, and human inspection is time-consuming and only 80% accurate.
Innovation Solution
An automated inspection system that includes a print assembly, image capture arrangement, and processing unit, which captures and compares printed images at lower resolutions to reference images, using GPUs for parallel comparison operations, allowing for real-time defect detection and reducing computational resources.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If automated inspection systems use referential approaches to inspect each image against a reference, then defect detection accuracy is improved, but inspection time and computational resources increase significantly
Solution Approach 1:
The patent segments the inspection process into two distinct stages: (1) a training phase where reference images are created from defect-free samples, and (2) an inspection phase where only anomaly detection is performed. This segmentation allows the system to achieve high defect detection accuracy while reducing inspection time, as the reference images are pre-established and do not need to be recreated for each inspection.
Solution Approach 2:
The system performs preliminary actions by creating reference images during a training phase before actual inspection begins. These reference images are stored and reused for multiple inspections, eliminating the need to perform full reference comparisons for every inspected image. This preliminary preparation significantly reduces the time required for subsequent inspections while maintaining high detection accuracy.
2Reliability
If automated inspection systems inspect variable data prints where each image is unique, then comprehensive defect coverage is achieved, but computational resources and costs increase
Solution Approach 1:
The patent creates reference images that serve as templates for comparison during inspection. These reference images are copies of defect-free printed images that capture the expected appearance of variable data prints. By using these pre-created reference copies, the system can efficiently compare against them without requiring complex real-time processing of each unique variable data image, thus reducing computational resources while maintaining comprehensive defect coverage.
3Adaptability or versatility
If manual inspection is used to detect print defects, then flexibility in handling various defect types is maintained, but inspection speed and accuracy are insufficient for commercial printing
Solution Approach 1:
The patent replaces the manual mechanical inspection process with an automated optical inspection system using cameras and image processing algorithms. This substitution maintains flexibility in detecting various defect types through programmable image analysis while dramatically increasing inspection speed to match commercial printing velocities, achieving both adaptability and productivity.
Data Source
AI summary
Automated inspection method for detecting a defect in a printed image, comprising processing a raster image, sending the raster image to a print process, printing a printed image corresponding to the raster image onto a medium, capturing a target image from at least a part of the printed image at a lower resolution than the printed image, at least in a medium moving direction, converting at least a part of the raster image to a reference image, and comparing the reference image to the target image.