Hi,

This week you'll learn about Text Detection and OCR with Amazon Rekognition API.

Image

There can't be enough emphasis on the number of tasks that have incorporated optical character recognition (OCR) as an integral part of its workflow. Sometimes it's to such an extent that some systems are centered around OCR. What is the one thing you simply cannot compromise on in a scenario like that? 

Accuracy

Imagine a scenario where you are instructed to build a number plate recognition system. There isn't really a point if your system works fast but ends up mislabeling certain digits. In the real world, it's always ideal to find the sweet spot between speed and accuracy. However, if you were to prioritize one, it should be accuracy. This will make the application more robust and suitable for the real world.

That brings us to today's lesson on Text Detection and OCR with Amazon Rekognition API. 

The big picture: Peter Sondergaard (Gartner Research) rightly said, "Information is the oil of the 21st century, and analytics is the combustion engine." We have familiarized ourselves with the usage of OCR engines like Tesseract. However, most cloud-based OCR modules created by big companies will show better results due to being trained on more data. 

How it works: The Amazon Rekognition API is hosted on the cloud. To access it for your local projects, you need your Amazon Web Services (AWS) access key. Once you specify your server's region and the access key, your local session will directly get linked to Amazon's cloud services, and you can use the API according to your needs. 

Our thoughts: The Amazon Rekognition OCR API is a huge jump from the Tesseract OCR engine. Since the former has been trained on various scenarios and improved almost constantly. It's more likely to work on whatever image you provide to it. 

Yes, but: There are trade-offs to the high accuracy. For example, using Amazon Web Services will require a constant internet connection, and its resources will be subscription-based. Also, since the system is cloud-based, it might be slower than your localhost engines. 

Stay smart: The learning curve for Amazon Web Services is a bit steep, but the pay-off is undeniably worth it. Mastering AWS can help your progress through the world of computer vision and deep learning. 

Click here to read the full tutorial

PyImageSearch University

This lesson is part of PyImageSearch University, our flagship program to help you master computer vision, deep learning, and OpenCV. PyImageSearch University is updated each week with new lessons.

Don't spend your time trying to piece together code when you could get pristine code right now. Want the full, working code of every blog post plus a Colab Notebook and video walkthrough of the code?

I want the code

Having problems with your local development environment or IDE? Fortunately, our pre-configured Colab Notebooks allow you to run code the moment you join PyImageSearch University. But, of course, you don't want to be a sys-admin, so don't waste time messing with your development environment. 

Want to Master Computer Vision and Deep Learning?

Do you think mastering computer vision and deep learning has to be time-consuming, overwhelming, and complicated? Or has to involve complex mathematics and equations? Or requires a degree in computer science?

That's not the case. All you need to master computer vision and deep learning is for someone to explain things to you in simple, intuitive terms. And that's exactly what we do. Our mission is to change education and how complex Artificial Intelligence topics are taught.

Inside PyImageSearch University, you'll find:

  • Code that works for almost any OpenCV or TensorFlow CV problem
  • 35+ courses on the hottest computer vision, deep learning, and OpenCV topics
  • 35+ Certificates of Completion (one for each course)
  • 39+ hours of on-demand video
  • Pre-configured Jupyter Notebooks running in Google Colab
  • Run all code examples in your web browser — works on Windows, macOS, and Linux (no dev environment configuration required!)
  • Access to centralized code repos for most of the 500+ tutorials on the PyImageSearch blog
  • Easy one-click downloads for code, datasets, pre-trained models, etc.
  • Access on mobile, laptop, desktop, etc.
  • New courses released regularly and new tutorials weekly, ensuring you can keep up with state-of-the-art techniques

Click here to join PyImageSearch University


The PyImageSearch Team

P.S. You may have missed this, but last Wednesday, we published a new post on Choosing the Research Topic and Reading Its Literature.