Adaptive Image Smoothing for Printed Product Inspection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In the printing industry, automatic quality control systems face challenges in accurately inspecting printed products due to irregularities in printing substrate transport, such as sheet fluttering, which causes non-linear distortions in captured images, leading to false positive defects in image inspection systems.
Innovation Solution
The method involves adaptive image smoothing, where the digital reference image is smoothed using locally calculated smoothing factors based on the comparison between captured and reference images, isolating and compensating for non-linear distortions specific to the affected regions, thereby eliminating pseudo-defects during image inspection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If uniform smoothing is applied to the entire reference image, then the reference image becomes smoother, but it fails to compensate for local non-linear distortions in specific regions
Solution Approach 1:
The patent applies different smoothing factors to different regions of the reference image based on local distortion characteristics. The computer identifies distorted regions and calculates region-specific smoothing factors, then applies these locally rather than uniformly across the entire image. This resolves the contradiction by making the smoothing process adaptive to local conditions, improving inspection accuracy in distorted regions while maintaining a relatively simple overall process.
2Ease of manufacture
If the sheet trailing edge is allowed to flutter during transport, then the transport process is simpler, but non-linear distortions occur in the captured images
Solution Approach 1:
The patent performs preliminary identification of distorted regions and calculation of appropriate smoothing factors before the actual image inspection process. By pre-processing the reference image with region-specific smoothing, the system compensates for expected distortions in advance, allowing the sheet to flutter during transport without compromising inspection accuracy. This resolves the contradiction by preparing the reference data beforehand to account for transport irregularities.
3Measurement precision
If manual monitoring of image inspection is used, then false positive defects can be identified, but the inspection process requires human intervention
Solution Approach 1:
The patent implements an automated system that self-identifies distorted regions and self-calculates appropriate smoothing factors without human intervention. The computer automatically compares captured images with the reference image, identifies regions with non-linear distortions, and applies region-specific smoothing factors. This automated self-service approach resolves the contradiction by eliminating the need for manual monitoring while maintaining high defect identification accuracy through intelligent image processing.
Data Source
AI summary
A method for image inspection of printed products in a printing material processing machine includes using an image sensor to capture and digitize printed products during image inspection by an image capture system, using a computer to carry out a comparison of captured digital printed images with a digital reference image, examining a result of the comparison for image regions with blurry regions in the captured digital printed image, calculating suitable smoothing factors for the blurry image regions, and then carry out smoothing of the digital reference image with the calculated, suitable smoothing factors for the blurry image regions, causing the digital reference image to have a blurriness comparable to the captured digital printed image in the image regions. Printed products detected as faulty are removed upon an occurrence of deviations of the captured digital printed image from the digital reference image.