The first open-source library that detects the font of a text in a image.

Last update: Feb 24, 2022

Overview

Typefont

Typefont is an experimental library that detects the font of a text in a image.

Usage

Import the main function and invoke it like in the following script.

import { Typefont } from "./src/index.js";

Typefont("image.png").then((result) => console.log(result));

import { Typefont } from "./src/index.js";

async function getFontFromImage (source) {
    const fonts = await Typefont(source);
    
    return fonts[0]; // Return the most similar font.
}

The first argument of the function can be the path or the base64 of the image. The function returns a Promise that when is resolved returns an array containing each font ordered in descending order (considering the similarity percentage).

Preview

Text on the cover of a book (the language is different because I live in Italy).

Text on the cover of another book.

Version 2

I'm working on a new version which gets the fonts directly from .ttf files and the Google Fonts database. The comparison is made using the Hausdorff Distance and the Shape Context. If you are interested in a collaboration contact me ([email protected]). It's difficult to progress since I work and I have many other projects.

Options

You can pass an object with options to the function as second argument.

Option	Type	Description	Default
`progress`	`Function`	A function that is called every time the comparison with a font is completed.	`undefined`
`minSymbolConfidence`	`Number`	The minimum confidence that a symbol must have to be accepted in the comparison queue (the confidence value is assigned by the OCR engine).	`15`
`analyticComparisonThreshold`	`Number`	The threshold of the analytic comparison.	`0.5`
`analyticComparisonScaleToSameSize`	`Boolean`	Scale the symbols to the same size before the analytic comparison?	`false`
`analyticComparisonSize`	`Number`	Used as dimension when resizing the images to the same size during the analytic comparison.	`128`
`perceptualComparisonSize`	`Number`	Used as dimension when resizing the images to the same size during the perceptual comparison.	`64`
`fontsDirectory`	`String`	The URL of the directory containing the fonts.	`storage/fonts/`
`fontsData`	`String`	The name of the file containing the JSON data of a font.	`data.json`
`fontsIndex`	`String`	The URL of the fonts index JSON file.	`storage/index.json`
`fontRequestTimeout`	`Number`	Font request timeout [ms].	`2000`
`textRecognitionTimeout`	`Number`	Text recognition timeout [s].	`60`
`textRecognitionBinarization`	`Boolean`	Binarize the image before the recognition?	`true`

Example

Example with options.

Typefont("restaurant-logo.jpg", {
    minSymbolConfidence: 50,
    analyticComparisonScaleToSameSize: true,
    analyticComparisonSize: 256
}).then(res => console.log(res));

Todo

Store and load fonts directly from .ttf files.
Implement the Shape Context algorithm to improve comparison results.
Implement the Hausdorff distance algorithm to improve the comparison results.
Import the Google Fonts database.

How it works?

Short summary: the input image is passed to the optical character recognition after some filters based on its brightness. The symbols (letters) are extracted from the input image and compared with the symbols of the fonts in the database using a perceptual comparison and a pixel based comparison in order to obtain a percentage of similarity.

How to add a font

The fonts stored in this database are just a JSON structure with letters as keys and the base64 of the image of the letter of the font as value. If you want to add a new font you must follow this structure.

{
    "meta": {
        "name": "name",
        "author": "author",
        "uri": "uri",
        "license": "license",
        "key": "value",
        ...
    },
    "alpha": {
        "a": "base64",
        "b": "base64",
        "c": "base64",
        ...
    }
}

Then you have to include your font in the index of fonts by adding the font name to the array.

License

MIT License.

Credits

Author: Vasile Pește ([email protected]).

You might also like...

keras复现场景文本检测网络CPTN: 《Detecting Text in Natural Image with Connectionist Text Proposal Network》；欢迎试用，关注，并反馈问题...

keras-ctpn [TOC] 说明预测训练例子 4.1 ICDAR2015 4.1.1 带侧边细化 4.1.2 不带带侧边细化 4.1.3 做数据增广-水平翻转 4.2 ICDAR2017 4.3 其它数据集 toDoList 总结说明本工程是keras实现的CPTN: Detecti

107 Jan 9, 2023

Detecting Text in Natural Image with Connectionist Text Proposal Network (ECCV'16)

Detecting Text in Natural Image with Connectionist Text Proposal Network The codes are used for implementing CTPN for scene text detection, described

1.3k Dec 22, 2022

AdvancedEAST is an algorithm used for Scene image text detect, which is primarily based on EAST, and the significant improvement was also made, which make long text predictions more accurate.https://github.com/huoyijie/raspberrypi-car

AdvancedEAST AdvancedEAST is an algorithm used for Scene image text detect, which is primarily based on EAST:An Efficient and Accurate Scene Text Dete

1.2k Dec 29, 2022

WACV 2022 Paper - Is An Image Worth Five Sentences? A New Look into Semantics for Image-Text Matching

Is An Image Worth Five Sentences? A New Look into Semantics for Image-Text Matching Code based on our WACV 2022 Accepted Paper: https://arxiv.org/pdf/

13 Dec 17, 2022

This pyhton script converts a pdf to Image then using tesseract as OCR engine converts Image to Text

Script_Convertir_PDF_IMG_TXT Este script de pyhton convierte un pdf en Imagen luego utilizando tesseract como motor OCR convierte la Imagen a Texto. p

1 Jan 27, 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 2, 2023

The first open-source library that detects the font of a text in a image.

Related tags

Overview

Typefont

Usage

Preview

Version 2

Options

Example

Todo

How it works?

How to add a font

License

Credits

You might also like...

keras复现场景文本检测网络CPTN: 《Detecting Text in Natural Image with Connectionist Text Proposal Network》；欢迎试用，关注，并反馈问题...

Detecting Text in Natural Image with Connectionist Text Proposal Network (ECCV'16)

AdvancedEAST is an algorithm used for Scene image text detect, which is primarily based on EAST, and the significant improvement was also made, which make long text predictions more accurate.https://github.com/huoyijie/raspberrypi-car

WACV 2022 Paper - Is An Image Worth Five Sentences? A New Look into Semantics for Image-Text Matching

This pyhton script converts a pdf to Image then using tesseract as OCR engine converts Image to Text

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

The Open Source Framework for Machine Vision

A semi-automatic open-source tool for Layout Analysis and Region EXtraction on early printed books.

Tesseract Open Source OCR Engine (main repository)

Releases(v0.1-beta.0)

v0.1-beta.0(May 11, 2017)

v0.1-alpha.0(Apr 30, 2017)

Owner

Vasile Pește

Text recognition (optical character recognition) with deep learning methods.

CTPN + DenseNet + CTC based end-to-end Chinese OCR implemented using tensorflow and keras

Detect text blocks and OCR poorly scanned PDFs in bulk. Python module available via pip.

A simple Security Camera created using Opencv in Python where images gets saved in realtime in your Dropbox account at every 5 seconds

【Auto】原神⭐钓鱼辅助工具 | 自动收竿、校准游标 | ✨您只需要抛出鱼竿，我们会帮你完成一切✨

Augmenting Anchors by the Detector Itself

Shape Detection - It's a shape detection project with OpenCV and Python.

FOTS Pytorch Implementation

OpenCV-Erlang/Elixir bindings

Program created with opencv that allows you to automatically count your repetitions on several fitness exercises.

Handwritten Text Recognition (HTR) system implemented with TensorFlow (TF) and trained on the IAM off-line HTR dataset. This Neural Network (NN) model recognizes the text contained in the images of segmented words.

This is a c++ project deploying a deep scene text reading pipeline with tensorflow. It reads text from natural scene images. It uses frozen tensorflow graphs. The detector detect scene text locations. The recognizer reads word from each detected bounding box.

Histogram specification using openCV in python .

scene-linear test images

Convert PDF/Image to TXT using EasyOcr - the best OCR engine available!

Genalog is an open source, cross-platform python package allowing generation of synthetic document images with custom degradations and text alignment capabilities.

APS 6º Semestre - UNIP (2021)

Brief idea about our project is mentioned in project presentation file.

Official PyTorch implementation for "Mixed supervision for surface-defect detection: from weakly to fully supervised learning"

Contextual speed detection for python