Skip to Content

Advanced Image Pre-processing

Optimize your scanned documents with high-precision image enhancements, ensuring cleaner inputs for superior OCR accuracy.

Free Trialimage of an arrow down

Perfect Image Clarity with Our Advanced Pre-Processing Module

Overview

The Advanced Image Pre-processing Module (IPP) is designed to enhance the quality of images before they undergo Optical Character Recognition (OCR). This module improves OCR accuracy for poor-quality images and enables additional features such as stamp detection. It is installed as a separate application and integrates with Trade AI to process documents upon upload.

Features and Functionality

The system enhances document quality and OCR accuracy through advanced image correction features like deskewing, contrast adjustment, and despeckling. It also improves OCR results by removing vertical lines, detecting text within stamps, and eliminating tick marks, ensuring more precise data extraction.

  • Deskewing & Enhancing: Corrects skewed pages and adjusts contrast for improved clarity.
  • Despeckling & Vertical Line Removal: Cleans up images by removing grains, spots, and vertical lines that interfere with OCR.
  • Stamp Detection & Tick Marks Removal: Detects text from stamps and removes tick marks to ensure accurate text recognition.

Benefits

The system significantly enhances OCR accuracy by improving image quality through various correction features, especially for poor-quality images. It provides additional capabilities such as stamp detection and removal of unwanted marks, crucial for accurate document processing. The selective application of these optimization features based on OCR confidence levels further boosts processing efficiency.

  • Improved OCR Accuracy: Enhances text recognition accuracy by correcting and refining image quality.
  • Additional Features: Facilitates the extraction of text from stamps and removal of marks, improving overall document processing.
  • Selective Application: Optimizes processing efficiency by applying image corrections based on OCR confidence levels.

Summary

The Advanced Image Pre-processing Module thus plays a critical role in enhancing document processing workflows by improving image quality and OCR accuracy.

OverviewFeatures and FunctionalityBenefitsSummary

How does it work ?

Initiate OCR

The OCR process is initiated as follows:

  • IPP Module Installation: The Image Pre-processing (IPP) module is installed as a separate application.
  • Trigger IPP Service: The IPP service is activated when a user uploads and submits a document in Trade AI.
  • Pre-OCR Processing: The document undergoes pre-processing in the IPP module before being sent for Optical Character Recognition (OCR).

Optimize

During the OCR initiation phase:

  • Image Optimization for Low Confidence Pages: Pages with low OCR confidence are flagged and sent for image optimization to enhance clarity and accuracy before reprocessing.
  • Selective Application: The system can apply OCR to specific document sections or modules, such as "Terms and Conditions," selectively. This means:
    • Initial Processing: Only chosen sections are processed initially.
    • Reprocessing Based on Confidence: If OCR confidence scores are low, the document or specific sections may be sent again for further optimization to improve accuracy.

For example, if pages identified as "Terms and Conditions" are deemed unnecessary for detailed OCR processing, they can be excluded from the initial OCR run.

Re-OCR

During the Re-OCR phase:

  • Reprocessing Optimized Pages: Pages that were previously sent for image optimization are re-processed through OCR to extract text more accurately.
  • Selection of Best Results: The system evaluates the OCR results based on confidence scores. The highest quality result, reflecting the best accuracy, is selected for further use.

This ensures that only the most accurate and reliable data is used for subsequent processing stages.