Which one is the best book to start learning Opencv?

edit

If you are a beginner in computer vision, start by reading a course or a book. There are plenty of ressources on the net.

It is essential to understand the theory before. Because OpenCV is just a library and once you understand the theory, it is more or less just a matter to call the functions with the correct arguments. Replace OpenCV with Matlab or scikit-image, they are ultimately just some tools to build your computer vision pipeline.

When you start implementing your own computer vision functions, e.g. a function to detect Harris corners, you will understand or grasp a good idea about the implementation details inherent to computer vision. This will allow you to understand and read the OpenCV source code more easily.

If you are not a beginner in computer vision, you should look at the documentation, at the samples and more generally into the source code. Finally, a good book about OpenCV and computer vision is Learning OpenCV 3 by Gary Bradski and Adrian Kaehler.

I have explored around many places and found so many books. Some are good and some are not that good. In my experience, you can go for OpenCV with Python Blueprints book by Michael Beyeler. This is an excellent book for the beginner to semi-advanced users. It includes plenty of solid examples with useful topics. This book practically shows you how to develop real-world OpenCV applications with python.