4.49 (86 reviews)
☑ What is Optical Character Recognition (OCR)?
☑ A general OCR pipeline used by most industries.
☑ Different Image Pre-processing techniques used in OCR pipeline.
☑ Different Text Detection techniques used in OCR pipeline such as EAST and CTPN.
☑ Different Text Recognition techniques used in OCR pipeline such as CRNN (CNN+RNN+CTC)
☑ Implementing OCR on real-life examples
Welcome to the course 'Mastering OCR using Deep Learning and OpenCV-Python'. This is the first course of my OCR series.
In this course we will start from the very basics. We will first discuss what is Optical Character Recognition and why you should invest your time in learning this.
Then we will move to the general pipeline used by most of the OCR systems available.
After this we will start learning each pipeline component in detail. We will start by learning some image pre-processing techniques commonly used in OCR systems.
Then we will learn some deep learning based text detection algorithms such as EAST and CTPN. We will also implement the EAST algorithm using OpenCV-Python.
Next we will learn the crux of the CTC which is widely used in developing text recognition systems. We will implement very famous text recognition algorithm that is CRNN.
Finally we will learn the last component of the OCR pipeline that is restructuring. In this we will discuss why is restructuring important for any OCR systems.
We will also discuss an open source end-to-end OCR engine which is pytesseract.
Finally we will run the complete OCR pipeline to extract the data from identification document using pytesseract.
So that's all for this course, see you soon in the class room. Happy learning and have a great time.
Stay safe, stay healthy.
Introduction to Optical Character Recognition
Optical Character Recognition Pipeline
OCR Pipeline -1 : Image pre-processing
Introduction to Image pre-processing
OCR Pipeline -2 : Text Detection
Introduction to Text Detection
EAST : Efficient and Accurate Scene Text detector
Implementation of EAST algorithm using OpenCV-Python
CTPN : Connectionist Text Proposal Network
Text Detection Datasets
Additional Resources for Text Detection Datasets
OCR Pipeline -3 : Text Recognition
Introduction to Text Recognition
CRNN model and Intuition behind CTC
Connectionist Temporal Classification (CTC)
Implementation of CRNN model using OpenCV-Python
Text Recognition Datasets
OCR Pipeline -4 : Restructuring
Implementation of end-to-end OCR pipeline using Pytesseract
A Complete Guide to Pytesseract
ID data Extraction
Congratulations and What's Next?
The lecturer is very competent and conveys this knowledge very professionally. The course content is detailed enough to understand the topic well.
21/05末、このコースはちゃんと最後まで動きました！ 理屈は難しいのでよくわかりませんが、最後まで動かせたのは体験としては非常によかったです。 講師の先生も質問したらすぐ返信してくれますし、いいコースでした。
It was good explanation. Kindly bring the next course as soon as possible. Can you please give any tentative date on that?
The structor goes to great lengths trying to explain models and dnns, however ends up using only pre-built ones. I'd say the course does not live up to expectations, given that there's no actual Deep Learning material to grow from and adapt to real life-scenarios, e.g., how to re-train a model to improve OCR accuracy for a particular use case.
This course was not was I was expecting, no recommended at all. Unfortunately, I paid for this, but the same content is available at google forums.
A good course on OCR, basics have been very lucidly covered with explicit examples. It is worth enrolling in.
Excellent course for everyone with python's basic knowledge and want to explore use of python in real world.