Autonomous Car and Advanced Robotic Systems have been very HOT topics with great attention from Scientists in many related fields, such as Robotics, Computer Vision, Sensor Fusion, Machine Learning -AI, etc. A merging-back between Computer Vision, Machine Learning (Deep Learning) and Robotics enables new mechanisms in solving difficult autonomous system problems (these were separated from the Cognitive Vision-Robotics in the past 50 years).
​
Nevertheless, such merging requires both broader knowledge and deeper understanding of researchers to follow. Reading decades of books in each field and trying to connect them might take a number of years of investigation. To help researchers and Ph.D students to quickly grasp sufficient knowledge to catch-up this new or combined fields, I have written books and lectures.
Lectures
​
Lecture 1: Derivation of Kalman Filter Derivation from Hilbert vector space
​
​
Lecture 2: Camera Calibration: Pinhole Model
​
​
Lecture 2: Camera Calibration: Kannala Model
​
​
Lecture 3: Camera Calibration: Unified Model
​
​
Lecture 4: Camera Calibration: Scaramuzza Model
​
​
Lecture 5: Visual Odometry & OdomFusion
​
​
Lecture 6: Re-localization
Book: Fundamental Maths to 3D Vision
Book: Perceptual Inference for Autonomous Navigation
3D vision is important for robotics and autonomous systems, especially for autonomous car topic. For the purpose of supporting researchers and Ph.D students to understand deeply the topics, and thus can contribute to it later, I decide to spend time on writing such text-book for free.
This book is written incrementally due to my limited availability. I will upload chapter-by-chapter to this page.
​
- Book Preface, Scope, Author
​
​
​
- Chapter 1: Understanding Scope of 3D Vision
​
​
​
- Chapter 2: Problem-Setup
​
​
​
- Chapter 3: Fundamental Maths
​
​
- Chapter 4: 3D Geometry & Camera Model
.. to be uploaded
​
​
​
- Chapter 5: Machine Learning for 3D Vision
... to be uploaded
​
​
​
​
I wrote this book in the sense of connecting two fields Computer Vision and Robotics. Since challenging autonomous navigation problems requires understanding of both texture and 3D structure. Therefore, a 3D object analysis is required, which at first needs a 2D-3D mapping, and then an investigation of object properties with a given combined 2D-3D information.
My book can be found in Amazon: