You've Been Doing Visual Odometry WRONG All Along in Robot Autonomy

Описание: #autonomousvehicles #autonomousnavigation #computervision #autonomousrobots

VISUAL ODOMETRY.

Explore real-time car motion tracking with FLIR BlackFly cameras, a key component in many autonomous vehicles. This setup demonstrates the practical challenges of moving from simulation to real-world robotics, especially in areas like robot navigation and machine vision. Overcoming these hurdles is crucial for robust computer vision and successful SLAM implementations, leading to better visual SLAM.

In this video, I show you how I built a real-world stereo camera rig for autonomous navigation using Fleer Blackfly cameras, OpenCV, and G2O optimization — moving beyond simulations to real, on-car visual odometry. I compare cheap AliExpress cameras, the ZED Mini, and finally a professional setup that actually works for 3D visual navigation using FLIR Blackfly Camera. You’ll see the full process — camera calibration, depth mapping, and real-time motion tracking — plus how to fix the most common issues that stifle your SLAM performance. If you’re into robotics, computer vision, or self-driving car projects, this video will help you bridge the gap between simulation and the real world.


-----------------------------------------------------------------------------------
📝In-video sources
➡️ Robotics 101: https://bit.ly/4qs6F6y
➡️ Computer Vision2: https://bit.ly/46hhu1V

-----------------------------------------------------------------------------------
🌏 Find Me Here:
🔥Linkedin:   / hongyun-elliot-lee  
🔥Discord:   / discord  
🔥ubicoders: https://www.ubicoders.com/courses
🔥Blog: https://www.ubicoders.com/blogs
🔥GitHub: https://github.com/ubicoders/
-----------------------------------------------------------------------------------
⏲️Time Stamps:
0:00 intro
1:11 how to escape from robotics simulator
2:49 why robots need 2 cameras
5:45 why robotics needs you some investment
7:18 how you solve computer vision camera problems
8:10 the USB 3 camea for 3D robotic perception
9:22 general framework for graph optimization
11:59 visual odometry