Information Processing Apparatus for Accurate Key Value Extraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing character recognition technologies face challenges in accurately extracting key character strings from images, particularly when the recognition is erratic or the degree of certainty is lower than a predetermined criteria, making it difficult to extract corresponding value character strings.
Innovation Solution
An information processing apparatus that uses a processor to acquire and extract key character strings from images, employing machine learning models like CNN and YOLO for object detection, and pattern matching to identify regions corresponding to key character strings, thereby extracting associated value character strings even when key character strings are difficult to extract or have low certainty.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional OCR operation is used to extract character strings, then the extraction process is simple, but the accuracy deteriorates when key character strings are difficult to recognize or have low certainty
Solution Approach 1:
The patent introduces an intermediary detection operation that uses machine learning models (CNN, YOLO) to detect regions corresponding to key character strings. This intermediary detection mechanism bridges the gap between traditional OCR and the needed high-accuracy extraction, allowing the system to handle uncertain or poorly recognized key character strings by using the detection operation as a mediator to locate and extract these critical elements even when standard OCR fails or produces low-confidence results.
2Productivity
If key character strings with low certainty are extracted, then more information can be retrieved, but the reliability of extracted information deteriorates
Solution Approach 1:
The patent implements a feedback mechanism where the detection operation evaluates the certainty of detected key character strings and uses this information to guide the extraction process. The system extracts value character strings based on the reliability assessment of key character strings, and this feedback loop ensures that only reliable extractions are produced, maintaining high reliability while maximizing information retrieval by selectively processing based on confidence levels.
Data Source
AI summary
An information processing apparatus includes a processor configured to: acquire an image corresponding to a key character string from a target image in response to the key character string that serves as a character string specified beforehand as a key and is acquired from results of character recognition performed on the target image including character strings; extract, by using results of acquiring the image corresponding to the key character string, from the results of the character recognition a value character string that serves as a character string indicating a value corresponding to the key character string; and output the key character string and the value character string corresponding to the key character string.