Artificial Intelligence Based Image Corner Detection In Machine Vision
Abstract
The Machine vision depends on image processing methods for obtaining the information necessary
for the tasks dedicated to perceive, sense and measure the world around the system of machine vision.
Corners are the principal local features in image. In general, they are nothing but the points which
include high curvature and appear in intersection of various intensities of images. An Artificial
Neural Network and Support Vector Machine based algorithm to detect corner is presented. The
algorithms will not enclose involvement of computing any complex differential geometric operators.
Both methods have hidden learning capability that results in excellent performance for an extensive
range of images. Experimental results show the proposed corner detection methods outperform the
existing methods and are better than the available traditional methods.