Some bits of javascript to transcribe scanned pages using PageXML

Overview

nashi (nasḫī)

Some bits of javascript to transcribe scanned pages using PageXML. Both ltr and rtl languages are supported. Try it! But wait, there's more: download now and get a complete webapp written in Python/Flask that handles import and export of your scanned pages to and from LAREX for semi-automatic layout analysis, does the line segmentation for you (via kraken) and saves your precious PageXML in a database. All you've got to do is follow the instructions below and help me implement all the missing features... OCR training and recognition is currently not included because of our webhost's limited capacity.

Instructions for nashi.html

  • Put nashi.html in a folder with (or some folder above) your PageXML files (containing line segmentation data) and the page images. Serve the folder in a webserver of your choice or simply use the file:// protocol (only supported in Firefox at the moment).
  • In the browser, open the interface as .../path/to/nashi.html?pagexml=Test.xml&direction=rtl where Test.xml (or subfolder/Test.xml) is one of the PageXML files and rtl (or ltr) indicates the main direction of your text.
  • Install the "Andron Scriptor Web" font to use the additional range of characters.

The interface

  • Lines without existing text are marked red, lines containing OCR data blue and lines already transcribed are coloured green.

Keyboard shortcuts in the text input area

  • Tab/Shift+Tab switches to the next/previous input.
  • Shift+Enter saves the edits for the current line.
  • Shift+Insert shows an additional range of characters to select as an alternative to the character next to the cursor. Input one of them using the corresponding number while holding Insert.
  • Shift+ArrowDown opens a new comment field (Shift+ArrowUp switches back to the transcription line).

Global keyboard shortcuts

  • Ctrl+Space Zooms in to line width
  • Ctrl+Shift+Space toggles zoom mode (always zoom in to line width)
  • Shift+PageUp/PageDown loads the next/previous page if the filenames of your PageXML files contain the number.
  • Ctrl+Shift+ArrowLeft/ArrowRight changes orientation and input direction to ltr/rtl.
  • Ctrl+S downloads the PageXML file.
  • Ctrl+E enters or exits polygon edit mode.

Edit mode

  • Click on line area to activate point handles. Points can be moved around using, new points can be created by drawing the borders between existing points.
  • If points or lines are active, they can be deleted using the "Delete"-key.
  • Hold Shift-key and draw to select multiple points
  • New text lines can be created by clicking inside an existing text region and drawing a rectangle. New lines are always added at the end of the region.

Instructions for the server

  • Install redis. The app uses celery as a task queue for line segmentation jobs (and probably OCR jobs in the future).
  • Install LAREX for semi-automatic layout analysis.
  • Install the server from this repository or from pypi:
pip install nashi
  • Create a config.py file. For more options see the file default_settings.py. If you want the app to send emails to users, change the mail settings there. Here is just a minimal example:
BOOKS_DIR = "/home/username/books/"
LAREX_DIR = "/home/username/larex_books/"
  • Set an environment variable containing your database url. If you don't, nashi will create a sqlite database called "test.db" in your working directory.
export DATABASE_URL="mysql+pymysql://user:[email protected]/mydb?charset=utf8"
  • Create the database tables (and users, if needed) from a python prompt. Login is disabled in the default config file.
from nashi import user_datastore
from nashi.database import db_session, init_db
init_db()
user_datastore.create_user(email="[email protected]", password="secret")
db_session.commit()
  • Run the celery worker:
export NASHI_SETTINGS=/home/user/path/to/config.py
celery -A nashi.celery worker --loglevel=info
  • Run the app, don't forget to export your DATABASE_URl again if you're using a new terminal:
export FLASK_APP=nashi
export NASHI_SETTINGS=/home/user/path/to/config.py
flask run
  • Open localhost:5000, log in, update your books list via "Edit, Refresh".

Planned features

  • Sorting of lines
  • Reading order
  • Creation and correction of regions
  • API for external OCR service
  • Advanced text editing capabilities
  • Help, examples, and documentation
  • Artificial general intelligence that writes the code for me
Owner
Andreas Büttner
Andreas Büttner
Programa que viabiliza a OCR (Optical Character Reading - leitura óptica de caracteres) de um PDF.

Este programa tem o intuito de ser um modificador de arquivos PDF. Os arquivos PDFs podem ser 3: PDFs verdadeiros - em que podem ser selecionados o ti

Daniel Soares Saldanha 2 Oct 11, 2021
Textboxes implementation with Tensorflow (python)

tb_tensorflow A python implementation of TextBoxes Dependencies TensorFlow r1.0 OpenCV2 Code from Chaoyue Wang 03/09/2017 Update: 1.Debugging optimize

Jayne Shin (신재인) 20 May 31, 2019
MeshToGeotiff - A fast Python algorithm to convert a 3D mesh into a GeoTIFF

MeshToGeotiff - A fast Python algorithm to convert a 3D mesh into a GeoTIFF Python class for converting (very fast) 3D Meshes/Surfaces to Raster DEMs

8 Sep 10, 2022
Fast image augmentation library and easy to use wrapper around other libraries. Documentation: https://albumentations.ai/docs/ Paper about library: https://www.mdpi.com/2078-2489/11/2/125

Albumentations Albumentations is a Python library for image augmentation. Image augmentation is used in deep learning and computer vision tasks to inc

11.4k Jan 02, 2023
Repository for playing the computer vision apps: People analytics on Raspberry Pi.

play-with-torch Repository for playing the computer vision apps: People analytics on Raspberry Pi. Tools Tested Hardware RasberryPi 4 Model B here, RA

eMHa 1 Sep 23, 2021
BoxToolBox is a simple python application built around the openCV library

BoxToolBox is a simple python application built around the openCV library. It is not a full featured application to guide you through the w

František Horínek 1 Nov 12, 2021
Handwritten_Text_Recognition

Deep Learning framework for Line-level Handwritten Text Recognition Short presentation of our project Introduction Installation 2.a Install conda envi

24 Jul 15, 2022
轻量级公式 OCR 小工具:一键识别各类公式图片,并转换为 LaTeX 格式

QC-Formula | 青尘公式 OCR 介绍 轻量级开源公式 OCR 小工具:一键识别公式图片,并转换为 LaTeX 格式。 支持从 电脑本地 导入公式图片;(后续版本将支持直接从网页导入图片) 公式图片支持 .png / .jpg / .bmp,大小为 4M 以内均可; 支持印刷体及手写体,前

青尘工作室 26 Jan 07, 2023
Demo for the paper "Overlap-aware low-latency online speaker diarization based on end-to-end local segmentation"

Streaming speaker diarization Overlap-aware low-latency online speaker diarization based on end-to-end local segmentation by Juan Manuel Coria, Hervé

Juanma Coria 185 Jan 01, 2023
Rubik's Cube in pygame with OpenGL

Rubik Rubik's Cube in pygame with OpenGL The script show on the screen a Rubik Cube buit with OpenGL. Then I have also implemented all the possible mo

Gabro 2 Apr 15, 2022
Simple SDF mesh generation in Python

Generate 3D meshes based on SDFs (signed distance functions) with a dirt simple Python API.

Michael Fogleman 1.1k Jan 08, 2023
Deep Learning Chinese Word Segment

引用 本项目模型BiLSTM+CRF参考论文:http://www.aclweb.org/anthology/N16-1030 ,IDCNN+CRF参考论文:https://arxiv.org/abs/1702.02098 构建 安装好bazel代码构建工具,安装好tensorflow(目前本项目需

2.1k Dec 23, 2022
A tool for extracting text from scanned documents (via OCR), with user-defined post-processing.

The project is based on older versions of tesseract and other tools, and is now superseded by another project which allows for more granular control o

Maxim 32 Jul 24, 2022
Page to PAGE Layout Analysis Tool

P2PaLA Page to PAGE Layout Analysis (P2PaLA) is a toolkit for Document Layout Analysis based on Neural Networks. 💥 Try our new DEMO for online baseli

Lorenzo Quirós Díaz 180 Nov 24, 2022
Can We Find Neurons that Cause Unrealistic Images in Deep Generative Networks?

Can We Find Neurons that Cause Unrealistic Images in Deep Generative Networks? Artifact Detection/Correction - Offcial PyTorch Implementation This rep

CHOI HWAN IL 23 Dec 20, 2022
OCR system for Arabic language that converts images of typed text to machine-encoded text.

Arabic OCR OCR system for Arabic language that converts images of typed text to machine-encoded text. The system currently supports only letters (29 l

Hussein Youssef 144 Jan 05, 2023
Implementation of our paper 'PixelLink: Detecting Scene Text via Instance Segmentation' in AAAI2018

Code for the AAAI18 paper PixelLink: Detecting Scene Text via Instance Segmentation, by Dan Deng, Haifeng Liu, Xuelong Li, and Deng Cai. Contributions

758 Dec 22, 2022
Image augmentation library in Python for machine learning.

Augmentor is an image augmentation library in Python for machine learning. It aims to be a standalone library that is platform and framework independe

Marcus D. Bloice 4.8k Jan 04, 2023
CNN+LSTM+CTC based OCR implemented using tensorflow.

CNN_LSTM_CTC_Tensorflow CNN+LSTM+CTC based OCR(Optical Character Recognition) implemented using tensorflow. Note: there is No restriction on the numbe

Watson Yang 356 Dec 08, 2022