Have you ever thought: “Yes, I would love to learn programing in Python, but it is too difficult and time consuming. It would take me months, maybe years, to master thousands and thousands of lines of code and understand complex mathematical equations. And the books that are written for these courses are either very hard to follow or very superficial. So, I give up!” There is no need to give up, this book brings the solution for all your difficulties.
The 100-page Computer Vision book is perfect for you!
Learn how to write a code while understanding the theory
After reading this book, you will gain a full understanding of complicated code and mathematical formulas in a very short time while understanding the theory behind Computer Vision.
Computer Vision: What it is and why it matters
Computer Vision represents one of the most fascinating subjects. Both engineering and research fields have to deal with it nowadays. In the last few years, there has been significant progress in computer vision achievements. Nowadays computers can recognize people easily. On the latest iPhone you can implement face recognition end fingerprint recognition options. All this is possible thanks do deep learning and computer vision research results. Moreover, computers nowadays can recognize many different objects with very high accuracy. For instance, you can just point your cell phone when taking a picture and it will mark where a face is. You can wave to your smartphone and in that way, the smartphone will recognize that you want to take a photo and it will be saved on your phone.
Step into the future of intelligent robots and self-driving cars
You have probably heard about recent advancement in the self driving car industry. With first experiments nowadays it is almost possible to have a fully autonomous drive. In addition, a famous Boston dynamics start-up company from the US has developed an already well known series of robots. Those robots are human like, like the one called Atlas, but they can also be dog and can be used for many interesting and practical activities. The latest achievements can be seen on YouTube videos of Atlas robot, and it shows that it performs many acrobatic exercises better than an average human. All this will result in a tremendous growth of artificial intelligence end computer vision applications. This will boost new branches of technology and will create jobs of the future. Therefore, whether it is for a hobby or for your professional career starting a new venture in the computer vision area, this book will be an important and beneficial step for you.
Are you intimidated with complex mathematical equations?
However, I do not want to sound overly naive and to preach that Computer Vision, Deep Learning and Image Processing are easy fields to grasp. To fully master them, they do require a lot of persistence, effort and programming, as well as a solid understanding of mathematics. Especially for beginners or amateurs in Computer Vision, it can be very difficult and demotivating. I have witnessed this first hand for several years while teaching Artificial Intelligence, Machine Learning and Computer Vision to graduate students. Books that are written for these courses are not easy to follow. There is a lot of mathematics, equations and complex algorithms. Before you reach even your first hello world code snippet, you have to go over numerous obstacles of integrals and Fourier analysis. It can be intimidating.
The code is not the recipe
On the other hand, there are a lot of materials that teach this subject in a very superficial way. We call this way the “show me the code” way of learning. Although it allows for a practitioner a quick start in the Computer Vision journey, it does not provide an understanding of the algorithms being used. So very quickly someone can become demotivated when the person realizes that they’re only using a bunch of recipes. When it comes to changing certain parameters, optimizing the code, combining more approaches, one usually feels stuck and overwhelmed.
Master complex algorithms without giving yourself a headache
Therefore, my motto while teaching students was always to teach them to make practical things while understanding the theory. By understanding these implications, the goal was to get any intuition about the algorithms without the need of heavy mathematics where it is not needed. For years I’ve been improving my teaching methods and I judged my results by the final students projects that they had to deliver. So, after a few years, I was able to witness a face recognition system, license plate recognition and emotion recognition systems. This experience I now transfer to my blog posts and also accompany it from my understanding of industry needs. I have worked as a computer vision engineer within a camera team, and we were making software for self-driving cars. Hence, I do hope that I’m in a position to instruct you on what is the most beneficial road to take in order to start this journey. You will notice that “The 100-page computer vision book” consists of three things:
- Theory, which provides intuition about the subject
- Python code, being easy for beginners and quick to start with
- C++ code that should serve as a leverage so that you are ready to put your programs into hardware
Learn how to process your images
This book starts out with essentials on how to load and process images and videos. Then, it gives you an idea how you can improve the quality of your images. You will learn how to work with filters and how you can smooth or sharpen your images.
Prepare your data for Machine Learning algorithms
In addition, you will learn how to adjust the contrast of your images using the histogram equalization algorithm. It is one of the best methods for image enhancement, which provides better quality of images without loss of any information.
• You will learn how to detect basic shapes in images, such as lines and circles. For this, you will learn the basics of Hough transform.
• Learn how to manipulate image colors and create a fascinating fade in and fade out effect.
• You will get an intuition and all the needed tools to fully grasp a 2D discrete Fourier transform.
• Get a basic understanding about filtering methods.
Learn how to detect feature points in an image
In many Computer Vision and Machine Learning applications we need some feature points which we will track or which will assist us to compare and detect objects or scenes. A special chapter is dedicated to detection of these characteristic points in an image. We are doing this by the use of the well-known Harris corner detector.
So, a unique difference is that we will present all this knowledge with minimal mathematics needed, while allowing you to fully understand the concepts of these algorithms. After reading this book, you will be able to continue your computer vision mastery journey and be ready to develop some very exciting projects!