Image Inspection Master Image Generation With Attribute Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image inspection methods face challenges in accurately matching and detecting defects in images formed on recording sheets with preprint and variable images, particularly due to registration shifts, leading to inefficiencies and increased processing time and costs.
Innovation Solution
An image inspection system generates a master image by combining attribute data from preprint and variable images, allowing for accurate position matching and defect detection by distinguishing between preprint and variable image pixels, thereby reducing registration shift errors and processing complexity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If the master image is generated by combining preprint image and variable image, then the image inspection can be performed on preprint printing, but the position matching becomes difficult due to registration shift
Solution Approach 1:
The patent divides the master image into two distinct layers: a preprint image layer containing static reference information and a variable image layer containing dynamic content. This segmentation allows the system to separately process and match each layer, preventing registration shifts in the variable layer from affecting the precise position matching of the preprint layer.
Solution Approach 2:
The patent introduces an attribute data layer as an intermediary that annotates each pixel in the master image with metadata indicating whether it belongs to the preprint image layer or the variable image layer. This intermediary structure enables the inspection system to selectively match pixels based on their layer origin, resolving the position matching difficulty caused by registration shifts.
2Measurement precision
If the preprint image layer is masked to generate read image with only variable image, then the variable image can be inspected, but the processing time and complexity increase
Solution Approach 1:
The patent performs the layer separation and attribute data generation as preliminary actions during the master image creation process, rather than performing masking operations during the inspection phase. This preliminary processing eliminates the need for time-consuming masking operations during defect detection, reducing overall processing time while maintaining inspection accuracy.
Solution Approach 2:
The patent merges the master image generation process with the attribute data generation process into a single unified operation. By combining these processes, the system avoids separate processing steps for creating the master image and preparing it for inspection, thereby reducing processing time and complexity while maintaining the ability to accurately detect defects in the variable image layer.
3Measurement precision
If multiple master images are generated for preprint image and variable image separately, then the inspection can be comprehensive, but the memory storage requirement increases
Solution Approach 1:
The patent combines the preprint image and variable image into a single master image structure with layered organization and attribute data. This unified approach allows comprehensive inspection of both preprint and variable image components while requiring significantly less memory storage than maintaining separate master images, as the layered structure enables efficient pixel-level access to both layers without duplicating entire image data.
Data Source
Figure 1~3
Figure 4~5
Figure 6~7B
AI summary
An apparatus, system, method, and carrier medium storing an image inspection control program, each of which is capable of inspecting a printed image are provided. While generating an inspection image to be used for inspecting a read image obtained by reading the printed image, attribute data indicating, for each pixel in the inspection image, whether the pixel belongs to drawing data in the inspection image is generated. The attribute data is used to process inspection of the read image based on comparison between the read image and the inspection image.