Skip to content

Releases: Sinapsis-AI/sinapsis-ocr

sinapsis-ocr v0.3.0

30 Mar 19:46

Choose a tag to compare

Added two new packages to sinapsis-ocr monorepo

sinapsis-ocr v0.1.0

25 Mar 22:33

Choose a tag to compare

The sinapsis-ocr mono repo introduces a comprehensive set of templates for optical character recognition (OCR). This release includes two packages, sinapsis-easyocr and sinapsis-doctr, prividing robust OCR capabilities. These tools make it easier for developers to integrate OCR functionality into their applications. using the sinapsis framework.

  • Sinapsis DocTR

The sinapsis-doctr package integrates the DocTR library to deliver high-quality OCR using modern deep learning models. This package is designed for developers seeking advanced OCR capabilities with detailed text extraction.
Key Features

High-Quality OCR with DocTR
    Utilizes DocTR library for state-of-the-art OCR performance.
    Supports multiple detection and recognition architectures for flexible OCR solutions.

Detailed Text Extraction
    Provides text extraction with bounding boxes and confidence scores.
    Enables precise control over OCR output for advanced use cases.
  • Sinapsis EasyOCR

The sinapsis-easyocr package simplifies OCR tasks by leveraging the EasyOCR library. This package is ideal for developers seeking a straightforward yet effective OCR solution.
Key Features

Multi-Language OCR
    Supports OCR in multiple languages, making it versatile for global applications.
    Simplifies integration of OCR across different language requirements.

Text Extraction with Precision
    Extracts text with bounding boxes and confidence scores.
    Balances simplicity with accuracy for a wide range of OCR needs.
  • Sinapsis OCR WebApp

To showcase the functionality of these OCR libraries, we have developed a web-based demo application. This webapp provides an easy and straightforward way to test and explore the OCR capabilities of both sinapsis-doctr and sinapsis-easyocr.
Key Features

User-Friendly Interface
    A simple, intuitive web interface for uploading images and testing OCR functionality.
    Requires no coding knowledge to get started.

Multi-Library Support
    Demonstrates OCR capabilities using both DocTR and EasyOCR libraries.
    Allows users to compare results and choose the best option for their needs.

Real-Time OCR
    Provides instant OCR results with bounding box visualization.
    Displays confidence scores for extracted text.

Interactive Visualization
    Highlights detected text regions on images.
    Offers a clear visual representation of OCR output.