Video Description
Discover how to build advanced OpenCV3 projects with Python
About This Video
- Practical end-to-end projects covering an important computer vision problem
- Step-by-step guide to creating computer vision applications
- Program advanced computer vision applications in Python using different features of the OpenCV library
In Detail
OpenCV is a native cross-platform C++ library for Computer Vision, Machine Learning, and image processing. It is increasingly being adopted for development in Python.
This course features some trending applications of vision and deep learning and will help you master these techniques. You will learn how to retrieve structure from motion (sfm) and you will also see how we can build an application to capture 2D images and join them dynamically to achieve street views by capturing camera projection angles and relative image positions. You will also learn how to track your head in 3D in real-time, and perform facial recognition against a goldenset. You will also build an app to capture facial emotions based on a CovNet.
Next, you'll generate panoramas using image stitching and we extend this concept by generating a map based on the trajectory of ISS. You'll also learn to build an application to capture beautiful panoramas and also achieve AR effects. You then delve into one of the most trending domains of computer vision: autonomous cars. You'll learn about various architectures and develop the skills to detect lanes, and segment and track vehicles in traffic.You will be using Carla, which is a open driving simulator by Intel, for your project to train a car learn how to drive itself using an end-to-end model.
By the end of this course you will have learned to perform 3D reconstruction by stitching multiple 2D images and recovering camera projection angles. You will also have learned to capture facial landmark points and recognize emotion in images, including in real time. You will also have learned to generate a panorama of a scene and augment a camera view with virtual objects. You will be familiar with the field of self-driving cars and its history, and will have trained a car to drive itself in a simulator.
Table of Contents
-
Chapter 1 : Structure from Motion
- The Course Overview 00:02:21
- Camera Projection Models 00:07:34
- Multi-View Stereo 00:09:11
- Generating Point Clouds 00:07:53
- 2D-to-3D 00:07:50
- Street View 00:09:49
-
Chapter 2 : Building an Android App with Emotion-Based Selfie Filters
- Real-Time Face Detection Based on Eigenfaces 00:11:57
- 3D Head Pose Estimation 00:13:47
- Detecting Cats and Faces Using Haar Cascades 00:07:47
- Facial Landmark Detection Using Dlib Library 00:10:24
- Face Morphology, Averaging, and Swapping 00:08:50
- Expressions - A Selfie Camera App 00:10:19
-
Chapter 3 : Building a Camera App withPanorama, HDR and AR Features
- Image Stitching 00:10:52
- Aerial Video Montage 00:06:30
- Marker-Based Augmented Reality 00:08:41
- Markerless Augmented Reality 00:06:07
- High-Dynamic Range (HDR) Imaging 00:06:34
- Building a Panorama App 00:08:14
-
Chapter 4 : Imitation Learning
- Introduction to Self-Driving Cars 00:07:03
- Sensors and Measurements 00:08:40
- Self-Driving Car Architectures 00:11:32
- Understanding Perception in Self-Driving Cars 00:10:44
- Learning to Drive Using a CNN 00:07:23
- Building a Self-Driving Car Based on Imitation Learning 00:10:43