This will be the repository for the Computer Vision Course for the fall semester 2019 taught at KU.
So far, this is based on the code and structure from the OpenCV 3 Computer Vision with Python Cookbook, published by Packt.
The course is a mix of fundamental theory and applications. It uses
-
Python
-
OpenCV, as a cross-platform library for computer vision.
Examples of the code:
import argparse
import cv2
parser = argparse.ArgumentParser()
parser.add_argument('--path', default='../data/Lena.png', help='Image path.')
params = parser.parse_args()
img = cv2.imread(params.path)
Image formation
Camera optics and calibration
The structure of pixels and images
Low-level processing
- Histograms
- Filtering and enhancing
- Feature detection
Mid-level processing
- Segmentation
- Shape processing
- Optic flow
- Stereo processing
- Tracking
High-level processing
- Recognition and classification
- Face analysis (detection, identification, expressions, age, attractiveness)
- Describing pictures (deep learning)
- Making pictures (GANs)
- The limits of deep learning and where to go next