Document Recognition via Connected Component Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing optical reader systems face difficulties in accurately identifying and processing the top document in a stack of forms due to overlapping documents, bending, varying sizes, and inadequate lighting, which complicates the recognition of printed and man-made marks.
Innovation Solution
A computer-based system that captures and digitizes images of stacked forms, uses connected component analysis to identify features, and compares these features to pre-stored templates to determine the top form's identity and orientation, allowing for efficient processing of the top form even when others are partially visible.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If forms are stacked one on top of another to increase reading speed, then productivity improves, but the top form becomes difficult to distinguish and read due to overlapping forms underneath
Solution Approach 1:
The patent applies segmentation by dividing the image processing into distinct stages: first identifying the top form through edge detection and component analysis, then separately processing the marks on that form. This segmentation allows the system to isolate and identify the top form before attempting to read marks, resolving the difficulty of distinguishing the top form from overlapping forms underneath.
Solution Approach 2:
The patent implements preliminary action by performing top form identification and orientation determination before mark recognition. The system first locates edges, identifies components, determines the top form's identity and orientation, and only then proceeds to read marks. This preliminary positioning and identification sequence enables accurate top form detection even when forms are haphazardly stacked.
2Ease of operation
If forms are haphazardly stacked without alignment to improve efficiency, then the need for alignment and registration is eliminated, but the camera scene includes portions of many underneath forms making top form identification difficult
Solution Approach 1:
The patent applies local quality by focusing analysis on specific regions and features rather than the entire image. The system identifies components and features locally within the image, compares them to stored templates, and uses these local matches to determine the top form's identity and orientation. This local feature-based approach enables accurate top form detection despite the presence of other forms in the camera scene.
Solution Approach 2:
The patent employs parameter changes by transforming the image into different representations and adjusting analysis parameters. The system converts the image to grayscale, applies thresholding, performs connected component analysis, and uses feature extraction to transform visual information into identifiable patterns. These parameter transformations enable precise top form edge detection and identification despite haphazard stacking conditions.
3Adaptability or versatility
If forms include different sizes and printed symbols to provide variety, then adaptability improves, but distinguishing the top form becomes more difficult due to varying edges and artifacts
Solution Approach 1:
The patent implements preliminary action by establishing a comprehensive set of stored templates representing various form types, sizes, and printed symbols before processing. These pre-stored templates serve as reference patterns that enable the system to handle form variety. During processing, the system compares detected components against these pre-established templates to identify the top form, allowing adaptation to different form characteristics without increasing difficulty.
Solution Approach 2:
The patent applies universality by creating a multi-functional feature matching system that can handle various form types, sizes, and printed symbols through a single unified approach. The stored templates and component comparison mechanism serve multiple purposes: identifying different form types, determining orientation, locating edges, and recognizing marks. This universal system accommodates form variety while maintaining consistent top form identification capability.
Data Source
AI summary
Templates of known forms are stored in computer system. The templates are digitized pixels on which connected component analyses are performed resulting in a first list of components. Five to ten of those components are selected to create an ordered feature list for each form. The computer system then captures an optical image of a form positioned on the top of a stack of forms. The optical image is digitized and stored in the computer (or processor) system as a captured digital image of pixels. A connected component analysis is performed on the captured digital image that results in a second list of image components. Image components on the second list are compared to those on the first list and then each succeeding feature in one of the ordered feature lists. If the comparison is successful, the form is known and other marks on the form may then be processed. If the comparison is unsuccessful, a new feature list is tried.