OCR Apparatus Region-Specific Preprocessing for Low Certainty Text

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current OCR processing systems lack the ability to identify and address regions with low certainty degrees within an image, leading to decreased accuracy in text recognition, as they calculate certainty in units of entire images rather than partial regions.

Innovation Solution

An information processing apparatus that performs first preprocessing on acquired image data and executes additional preprocessing on specified partial regions based on feedback from post-processing, allowing for targeted improvement of certainty degrees in specific areas.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If certainty degree is calculated in units of entire images, then overall processing efficiency is maintained, but accuracy of text recognition in specific low-certainty regions deteriorates

Engineering Contradiction:
Improveprocessing efficiencyVSAvoidtext recognition accuracy
Core Design Contradiction:
ProductivityVSMeasurement precision

Solution Approach 1:

The patent divides the image into multiple regions and calculates certainty degree for each region separately rather than for the entire image. This allows identification of specific low-certainty regions that require additional preprocessing, while maintaining high-certainty regions without unnecessary processing, thus resolving the contradiction between overall processing efficiency and local recognition accuracy.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent applies different preprocessing operations to different regions based on their specific certainty degrees. Low-certainty regions receive additional preprocessing (such as contrast enhancement or noise reduction), while high-certainty regions are processed normally. This localized approach improves text recognition accuracy in problematic areas without sacrificing overall processing efficiency.

Inventive Principle:
Principle #3Local quality

2Measurement precision

If additional preprocessing is applied to all regions, then text recognition accuracy is improved, but processing time and computational resources increase

Engineering Contradiction:
Improvetext recognition accuracyVSAvoidprocessing time
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The patent applies additional preprocessing only to regions where it is necessary (low-certainty regions) rather than to the entire image. By identifying and targeting only the specific regions that require improved processing, the system achieves better text recognition accuracy where needed while avoiding the time and computational overhead of processing the entire image with enhanced operations.

Inventive Principle:
Principle #16Partial or excessive action

Solution Approach 2:

The system automatically identifies low-certainty regions and applies appropriate preprocessing operations without requiring manual intervention. The certainty degree calculation and region selection are performed autonomously by the processing system, enabling it to self-optimize the preprocessing approach for each image based on its specific characteristics.

Inventive Principle:
Principle #25Self-service

Data Source

PatentUS11238305B2Information processing apparatus and non-transitory computer readable medium storing program
Publication Date: 2022.02.01 FUJIFILM BUSINESS INNOVATION CORP
  • US11238305B2 patent drawing
  • US11238305B2 patent drawing
  • US11238305B2 patent drawing

AI summary

An information processing apparatus includes a processor configured to execute first preprocessing on acquired image data, and execute second preprocessing on a specified partial region of the image data as a target in a case where information for specifying at least one partial region in an image corresponding to the image data is received from post processing on which the image data after the first preprocessing is processed.