Explore advanced edge detection techniques in this comprehensive computer vision lecture. Delve into Prewitt and Sobel edge detectors, understanding their derivative masks and image derivatives. Compare Sobel and Prewitt methods before progressing to second derivative techniques. Examine the Marr-Hildreth edge detector, including LOG filters and zero crossings. Study the Canny edge detector in depth, covering gradient orientation, hysteresis thresholding, and the effects of Gaussian smoothing. Conclude with an introduction to edge detection using deep learning, providing a well-rounded understanding of both classical and modern approaches to this fundamental computer vision task.
Edge Detection Techniques - Prewitt, Sobel, Marr-Hildreth, and Canny - Lecture 4
University of Central Florida via YouTube
Overview
Syllabus
Intro
Edge detection - recap
Edge detection - fewer steps
Prewitt and Sobel Edge Detector
Derivative Masks
Image derivative
Prewitt Edge Detector
Sobel vs Prewitt
Second derivate
Marr Hildreth Edge Detector
Finding Zero Crossings
LOG Filter
On the Separability of LOG
Algorithm
Example
Canny Edge Detector
Gradient Orientation
Hysteresis Thresholding [L, H]
Final Canny Edges
Effect of Gaussian Kernel (smoothing)
Edge Detection with Deep Learning
Taught by
UCF CRCV
