Pretrained Pipeline for Reading and Removing Noise in Mixed Scanned and Digital PDF Documents

Description

This is a pretrained pipeline designed to remove noise from printed documents, improving the clarity and readability of the text for more accurate Optical Character Recognition (OCR). The model effectively filters out background noise, smudges, and other visual distortions commonly found in scanned or photographed documents, ensuring that the text is clean and well-defined for optimal OCR performance.

By preprocessing the document images to reduce noise, this pipeline significantly enhances the accuracy of text extraction, reducing the likelihood of errors caused by imperfections in the original document. It is an essential tool for improving OCR reliability in environments where document quality may vary, making it invaluable for tasks such as document digitization, automated content extraction, and data analysis.

Predicted Entities

Live Demo Open in Colab Download

How to use

pdf_pipeline = PretrainedPipeline('mixed_scanned_digital_pdf_image_cleaner', 'en', 'clinical/ocr')

pdf_path = '/content/pdfs/'
pdf_example_df = spark.read.format("binaryFile").load(pdf_path).cache()

result = pdf_pipeline.transform(pdf_example_df)

val pdf_pipeline = new PretrainedPipeline("mixed_scanned_digital_pdf_image_cleaner", "en", "clinical/ocr")

val pdf_path = "/content/pdfs/"
val pdf_example_df = spark.read.format("binaryFile").load(pdf_path).cache()

val result = pdf_pipeline.transform(pdf_example_df)

Example

Input

Output

Sample specifications written by
 , BLEND CASING RECASING

- OLD GOLD STRAIGHT Tobacco Blend

Control for Sample No. 5030

Cigarettes:

OLD GOLD STRAIGHT

 

John H. M. Bohlken

FINAL FLAVOR MENTHOL FLAVOR

Tars and Nicotine, Taste Panel, Burning Time, Gas Phase Analysis,
Benzo (A) Pyrene Analyses — T/C -CF~ O.C S51: Fee -

Written by -- John H. M. Bohlken
Original to -Mr. C. L. Tucker, dr.
Copies to ---Dr. A. W. Spears

C

~

Model Information