EXPLORING OCR MODEL AND BACKEND SUPPORT IN PYTHON

• Optical Character Recognition (OCR) is a technology that converts images, scanned documents, or PDFs into machine-readable text. In Python, there are many powerful OCR libraries and models available that support different backends and use cases.

• In this blog, I explore the most popular OCR modules available in Python and how they are used in real-world applications.

1.TESSERACT OCR (pytesseract):

Backend: Google Tesseract Engine (C++ based)

Python Wrapper: pytesseract
Tesseract is one of the most widely used open-source OCR engines. It is maintained by Google and supports multiple languages.

HOW TO INSTALL:

pip install pytesseract

SAMPLE CODE:

import pytesseract

from PIL import Image

img = Image.open("sample.png")
text = pytesseract.image_to_string(img, lang="eng")
print(text)

1. EASYOCR

Backend: PyTorch (Deep Learning based)

EasyOCR is a deep learning-based OCR library. It works well with complex images and multiple languages.

HOW TO INSTALL:
pip install easyocr

SAMPLE CODE:

  import easyocr

reader = easyocr.Reader(['en','ta'])
result = reader.readtext('sample.png')

for r in result:
print(r[1])

1. PADDLE OCR:

Backend: PaddlePaddle (Deep Learning Framework)

PaddleOCR is a powerful industrial-level OCR toolkit developed by Baidu.

HOW TO INSTALL:
pip install paddleocr paddlepaddle

SAMPLE CODE:
from paddleocr import PaddleOCR

ocr = PaddleOCR(use_angle_cls=True, lang='en')
result = ocr.ocr('sample.png')

for line in result[0]:
print(line[1][0])

1. KERAS OCR:

Backend: TensorFlow / Keras
Keras-OCR is built using deep learning models and provides both text detection and recognition

HOW TO INSTALL:
pip install keras-ocr

SAMPLE CODE:
import keras_ocr

pipeline = keras_ocr.pipeline.Pipeline()
images = [keras_ocr.tools.read('sample.png')]
prediction = pipeline.recognize(images)

print(prediction)

HOW IT WORKS IN BACKEND:

Image / PDF
↓
Preprocessing (OpenCV)
↓
Text Detection (DL/CNN)
↓
Text Recognition (CRNN / Transformer)
↓
Output Text (JSON / String)

AVAILABLE MODULES :

1.pytesseract

2.easyocr

3.paddleocr

4.opencv-python

5.pillow

6.pdf2image