- NCC Template Matching, Stereo Depth, and KNN
NCC template matching on a Simpsons scene, brute-force stereo depth from two views, and k-nearest neighbors classification.
7 min - Camera Calibration and Homography
DLT camera calibration from 100 2D-3D correspondences and normalized DLT homography from 15 point pairs.
5 min - Harris Corners, FAST Corners, and OpenCV
Harris corner detector from scratch on a checkerboard, FAST feature detection with a Bresenham circle, and comparing both to skimage.
7 min - Tracking Objects with Covariance and Mean-Shift
Covariance-based object matching with generalized eigenvalues, and mean-shift tracking with Epanechnikov-weighted color histograms.
7 min - Motion History Images and Optical Flow
Image differencing with morphological cleanup, MEI and MHI temporal representations, similitude moments on motion images, and Lucas-Kanade optical flow.
8 min - Shape Moments and PCA on a Box of Dots
Computing image moments from scratch, eigenanalysis of 2D data, confidence ellipses, decorrelation, and dimensionality reduction.
6 min - Image Pyramids and Background Subtraction
Gaussian and Laplacian pyramids from scratch, single vs multi-frame background subtraction, and connected components.
8 min - Reading Images and the NTSC Grayscale Formula
Loading images, the NTSC grayscale formula, and a checkerboard. First computer vision homework done in Python instead of MATLAB.
4 min
Back