Printed Code Inspection Using Candidate Character Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional printed code inspection techniques are computationally demanding, require significant setup time, and struggle with misinterpretation of characters due to lack of continuity and similarity, especially in continuous inkjet printing, and necessitate manual configuration and reconfiguration for each code change.
Innovation Solution
A method for printed code inspection that analyzes images using character identification parameters and candidate character properties to detect and verify printed codes, reducing computational requirements and setup time by focusing on candidate characters, and optionally performing OCR only on verified characters.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If OCR is performed over the entire image to detect printed characters, then character detection completeness is improved, but computational demand and processing time increase significantly
Solution Approach 1:
The patent segments the image processing task by first identifying candidate character regions using dot pattern recognition, then applying OCR only to these segmented regions rather than the entire image. This segmentation approach maintains character detection completeness while significantly reducing computational demand by limiting OCR to small, targeted areas.
Solution Approach 2:
The patent extracts candidate character regions from the broader image context by detecting dot patterns that form characters. This extraction allows the system to isolate and process only the relevant character areas, improving processing speed without sacrificing detection completeness.
2Productivity
If a specific search region is defined to reduce OCR processing area, then processing time is reduced, but up-front configuration time and setup complexity increase
Solution Approach 1:
The system performs self-configuration by automatically detecting dot patterns and determining candidate character regions without requiring manual user input or pre-training. This self-service capability eliminates setup time while maintaining fast processing speeds through automated region identification.
Solution Approach 2:
The patent performs preliminary dot pattern recognition and candidate character identification before applying OCR, automatically establishing the search regions needed for processing. This preliminary action eliminates the need for manual configuration while preparing the image for efficient OCR processing.
3Measurement precision
If OCR pre-training is performed to improve character recognition accuracy, then recognition precision is improved, but time and computational resources required increase significantly
Solution Approach 1:
The patent extracts and processes only candidate character regions identified through dot pattern recognition, rather than performing OCR pre-training on entire images or large datasets. This extraction approach maintains recognition accuracy by focusing computational resources on relevant character areas without requiring extensive pre-training.
Solution Approach 2:
The system applies partial OCR processing only to candidate character regions rather than performing complete OCR analysis on the entire image. This partial action approach maintains sufficient recognition accuracy for code inspection while avoiding the computational overhead and time requirements of full OCR pre-training.
4Measurement precision
If manual masking of variable elements is performed to improve code inspection accuracy, then inspection precision is improved, but setup time and operational complexity increase
Solution Approach 1:
The system automatically handles variable elements by detecting dot patterns and identifying candidate character regions without requiring manual masking or user intervention. This self-service capability maintains inspection accuracy while eliminating the operational complexity and setup time associated with manual masking procedures.
Data Source
AI summary
This specification describes methods and systems for printed code inspection. For instance, the specification describes a computer-implemented method for printed code inspection by a printed code inspection system operating in conjunction with a production line apparatus configured to move objects along a production line comprising: receiving an image of an object to which a printed code comprising one or more printed characters should have been applied, the image having been captured when the object was located at a particular position on the production line; analysing the image to detect, based on a set of one or more character identification parameters, at least one candidate character within the image; determining, for each of the at least one candidate characters and based on a set of one or more candidate character properties, a likelihood that the candidate character is one of the printed characters of the printed code that should have been applied to the object; determining, based on the candidate characters determined as being likely to be one of the printed characters of the printed code that should have been applied to the object, whether the printed code is present and legible on the object; and outputting an indication as to whether the printed code that should have been applied to the object is present and legible on the object.