Text Stitching for Pill Bottle Label Extraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for processing images of pill bottles taken from multiple sides face challenges such as image alignment and orientation issues, leading to inefficient entity extraction, which often requires manual processing.
Innovation Solution
A system and method for extracting information from multiple partial images using text stitching techniques, involving masking small characters, detecting and matching edge text boxes, and recognizing cross-image texts to determine entities, thereby automating the extraction process.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional image stitching methods are used to process multiple partial images, then image alignment and orientation problems occur, but the stitching process itself becomes complex and unreliable
Solution Approach 1:
The patent extracts and processes text regions independently from the image stitching process. By identifying and processing text boxes separately using OCR on individual partial images, the system avoids the complexity and reliability issues of stitching while still achieving the goal of extracting text information from the complete bottle label.
Solution Approach 2:
The patent segments the text extraction process into separate steps: detecting text boxes in each partial image, extracting text from each box, and then combining the results. This segmentation allows each step to be optimized independently, avoiding the need for complex global stitching operations.
2Productivity
If manual entity extraction is used, then extraction accuracy can be maintained, but processing time and labor requirements increase significantly
Solution Approach 1:
The system uses automated OCR (Optical Character Recognition) technology to perform entity extraction without manual intervention. The OCR system automatically detects text boxes, extracts text, and identifies entities, enabling the system to process images independently and efficiently, thus improving productivity while maintaining accuracy through sophisticated automated algorithms.
3Productivity
If all characters in partial images are processed equally, then complete text information is captured, but processing efficiency decreases due to unnecessary computation on insignificant characters
Solution Approach 1:
The patent applies different processing qualities to different parts of the text. By detecting text boxes and processing only the relevant text regions rather than all characters uniformly, the system improves processing efficiency while maintaining information completeness. The text box detection identifies and processes only the significant character regions needed for entity extraction.
Data Source
AI summary
A computer-implemented method including detecting respective one or more text boxes in each of multiple partial images of a text-bearing area. The method also can include determining respective one or more edge text boxes of the respective one or more text boxes in each of overlapping partial images of the multiple partial images, wherein each of the respective one or more edge text boxes comprise a respective incomplete text. The method additionally can include matching one or more pairs of corresponding edge text boxes from the respective one or more edge text boxes of two adjacent images of the overlapping partial images of the multiple partial images. The method also can include determining cross-image texts in the one or more pairs of the corresponding edge text boxes. The method further can include determining one or more entities in the text-bearing area based on entity texts of the cross-image texts and non-edge texts in respective one or more non-edge text boxes of the respective one or more text boxes in the multiple partial images. Other embodiments are described.