Camera Models (3D World to 2D Image Projection)

Ever wondered how cameras turn a 3D world into a 2D image—and how we can reverse that process to get depth? In this video, we break down camera models from first principles, focusing on the pinhole and orthographic projections, and explore how concepts like depth of field (DOF) affect what we capture. Sections: 0:00 Welcome 0:45 How Did Paintings Get Better? 7:20 Camera Obscura 10:15 Pinhole Model 14:42 Using Lenses for Better Images 22:21 Thin Lens Model 26:30 Depth of Field 35:22 Using Blur For Robot Navigation 38:43 3D Projection 50:24 Forced Perspective 52:57 Depth from Single Image 59:59 Camera Intrinsics 1:04:30 Camera Extrinsics 1:09:36 Camera Models 1:10:56 Recap 1:11:37 Understanding Check In this video, we’ll cover: ✅ How the pinhole camera model maps 3D points to the image plane ✅ The orthographic projection and when it’s a useful simplification ✅ How depth of field shapes what’s sharp and what’s blurred in an image ✅ The limits and possibilities of estimating depth from a single image using geometry We’ll connect these models to real-world robotics and computer vision tasks, showing why understanding your camera’s projection and optical limits is crucial for accurate perception. Whether you're new to robotics or an AI enthusiast, this video will give you a clear and fun introduction to the world of robots! 🔔 Subscribe for demystifying and deeper dives into perception, computer vision, AI, and robotics! 👍 Like this video if you enjoy learning about intelligent machines! 📩 Have questions? Drop them in the comments! #robotics #computervision #cameramodels #pinholecamera #orthographic #depthoffield #ai #perception #imageprocessing #robotvision #geometry #depthestimation #visualperception #artificialintelligence #embodiedintelligence