Comparison of Several Edge Detection Algorithms
ABSTRACT
This project will compare the Roberts, Sobel, Robinson, Canny, and the Hough image detection algorithms based on their ability to detect red squares on a black and white background. An image database consisting of actual grey-scale images taken from a test platform containing red squares on a black and white background will be constructed for the evaluation. The success of each algorithm will be based on its accuracy in detecting the red squares within the images from the database. The results of the algorithms will be then compared statistically to determine which one is the best suited for this application.

J. Burnham, J. Hardy, and K. Meadors
Department of Electrical and Computer Engineering
Mississippi State University
email: {jab2, jeh3, kam1}@ece.msstate.edu
URL: http://isip.msstate.edu/resources/courses/ece_4773/research/isip/projects/1997/group_image/presentation/