Curved Prescription Label OCR Using Multi-Scan Layer Reconstruction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Medication non-adherence is a significant challenge, particularly among elderly patients and those with visual and cognitive challenges, contributing to healthcare overspending and poor health outcomes, with existing OCR technologies struggling to accurately read prescription labels on curved surfaces.
Innovation Solution
A voice-interactive avatar-guided system using adaptive personality AI, multi-factor imaging, and context-based optical character recognition (OCR) to digitize prescription label information on curved surfaces, combined with pill image validation and augmented reality for personalized medication management.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional OCR technology is used on curved surfaces, then the system is simple, but the recognition accuracy deteriorates
Solution Approach 1:
The curved surface is divided into multiple planar segments or sections, each of which can be processed independently by standard OCR technology. The system captures images of different sections of the curved surface separately, then reconstructs the complete text by combining these segmented results, thereby maintaining high recognition accuracy on curved surfaces without requiring completely new OCR algorithms
Solution Approach 2:
The system transitions from attempting to recognize text directly on the curved 3D surface to capturing 2D images of the curved surface from multiple angles and positions. By working in the 2D image domain rather than directly on the 3D curved surface, standard OCR algorithms can be applied effectively, and the curved surface information is reconstructed through multi-view image synthesis
2Loss of information
If multiple scans are performed to improve coverage, then the completeness of text recognition is improved, but the time consumption increases
Solution Approach 1:
The system performs preliminary actions by capturing multiple images of the curved surface from different angles and positions before the actual OCR recognition process. These pre-captured images are stored and processed offline, allowing the OCR system to work with a comprehensive set of images without requiring multiple sequential scans during the recognition phase, thereby reducing time consumption while maintaining complete text coverage
Solution Approach 2:
The system creates multiple copies or versions of the curved surface text from different viewing angles and positions. Instead of performing multiple sequential OCR scans, the system captures multiple image copies simultaneously or in rapid succession, then processes these copies in parallel to reconstruct the complete text, significantly reducing the time required compared to sequential scanning
Data Source
AI summary
A method for optical character recognition (OCR) on a surface, comprising: activating an image capture device; scanning the surface to obtain a plurality of scans of sections of the surface; performing OCR on the plurality of scans; separating the OCRed content into layers for each of the plurality of scans; merging these layers into single layers; and combining the single layers into a unified image.