COURSE / BCV
Basic Computer Vision
Basic Computer Vision Link to heading
This page will collect work on image processing, feature extraction, matching, and visual geometry.
Status: edge detection lab and marker-based AR final project completed. The AR report now includes parameter sensitivity tests, keypoint and blending views, recorded-frame failure analysis and an approximate independent corner check, alongside the original synthetic and phone-to-webcam comparison of SIFT and ORB.
Work in this course
- Planar AR: matching a marker and replacing it
SIFT and ORB planar augmentation, tested on generated scenes and a phone in front of the webcam. Dimming and partial views expose the limits.
- Edge detection: gradients and Laplacian zero crossings
Comparing edges in buildings, roads, people and a car. The tiger shows where edge detection falls short.