Label Detection Using Special Marks for OCR Accuracy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional OCR systems struggle to recognize product codes in images, especially when they are embedded within graphics or advertisements, due to difficulties in separating text and graphics, leading to inefficient text recognition.
Innovation Solution
The implementation of a system that detects a special mark within an image, using predefined patterns, to locate the product label, allowing for efficient recognition of the label through optical character recognition (OCR) by using the special mark as a reference to isolate the label from other image content.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If conventional OCR systems process images containing product codes embedded in graphics or advertisements, then text recognition is attempted, but the system struggles to separate text from graphics leading to inefficient recognition
Solution Approach 1:
The patent applies preliminary action by first detecting and locating special marks (such as barcodes or QR codes) in the image before attempting OCR recognition. The system identifies the position of these special marks, extracts the surrounding text region, and then performs OCR only on the extracted portion. This preliminary localization step simplifies the overall image processing complexity while improving text recognition efficiency by focusing computational resources on relevant areas only.
2Measurement precision
If the system attempts to recognize all text in an image with embedded product codes, then character recognition is performed, but the accuracy decreases due to the complexity of separating text from graphics
Solution Approach 1:
The patent applies segmentation by dividing the image processing task into distinct stages: first detecting special marks (barcodes/QR codes), then extracting the text region surrounding these marks, and finally performing OCR on the isolated text portion. This segmentation separates the text recognition task from the complex graphics background, significantly improving character recognition accuracy by eliminating the interference of surrounding graphical elements.
3Loss of time
If conventional OCR processes entire images to find product codes, then all image content is analyzed, but this increases processing time and computational resources
Solution Approach 1:
The system performs preliminary detection of special marks (barcodes, QR codes) to quickly locate the position of product codes before attempting full OCR processing. By identifying these distinctive features first, the system can extract only the relevant text regions surrounding the special marks, thereby reducing processing time and improving recognition speed without analyzing the entire image content.
Data Source
AI summary
Determining a label from an image is disclosed, including: obtaining an image; determining a first portion of the image associated with a special mark; determining a second portion of the image associated with a label based at least in part on the first portion of the image associated with the special mark; and applying character recognition to the second portion of the image associated with the label to determine a value associated with the label.