Form Digitization

An error occured while load Image
Get StartedContact Sales
Training Data
An error occured while load Image
An error occured while load Image
An error occured while load Image
An error occured while load Image
An error occured while load Image
An error occured while load Image
An error occured while load Image
An error occured while load Image
An error occured while load Image
An error occured while load Image
Accuracy95%
Precision0.99
Recall0.96
Response Time3 s
What does the model do
Use OCR models to read text in images, find relevant fields from a document image, get rid of redundant information.
Learn More
How you can customize
Customize the model for your own use case and use template based or non-template based models depending on your needs and requirements.
Learn More
var data = 'modelId=_MODEL_ID&urls=https://goo.gl/ICoiHc'; var xhr = new XMLHttpRequest(); xhr.addEventListener("readystatechange", function () { if (this.readyState === this.DONE) { console.log(this.responseText); } }); xhr.open("POST", "https://app.nanonets.com/api/v2/ImageCategorization/LabelUrls/"); xhr.setRequestHeader("Content-Type", "application/x-www-form-urlencoded"); xhr.setRequestHeader("authorization", "Basic " + btoa("YOUR_API_KEY:")); xhr.send(data);
nodejspythongojavascriptrubyc#javashell

The Nanonets OCR Process

How to build your own custom OCR model

1

Create

2

Upload

3

Annotate

4

Integrate

  • Begin the process by adding the categories of text you want to detect

  • Upload 50 images of each text category you want to digitize

  • Annotate your images by drawing bounding boxes around the text and typing out the characters

  • Training your model takes 1-2 hours and you will be notified once it’s done

Why Nanonets?

No manual setup

Nanonets OCR generalizes across many invoice formats. As opposed to traditional OCR technologies which require a lot of manual tooling and template creation, reduce post processing and turn around times.

Consistent

Cut down the manual inspection time from hours to seconds using AI-powered automation. Analyze each image in milliseconds, maximizing the time for taking corrective measures which can also help increase throughput.

On-Premise

Deploy trained models on-prem using Docker images ensuring that your data does not have to leave your network or deploy models on our production optimized cloud instances for maximum performance.

Explore our other models

Slide 1 of 3

  • NSFW Classification

    Content Moderation Model for Images/Videos

    .

  • Fashion Apparel / Accessories Detection

    Detect clothing, accessories and other fashion items

    .

  • Pedestrian Detection in Aerial Images

    Detect Pedestrians from Aerial Drone shots

    .

Get in touch - start your free trial today!

Reach out for any questions on our Enterprise plan, pricing, or security. Chat with an enterprise collaboration specialist.