Ridge Detection Filter
This project implements Ridge Detection similar to the one in Wolfram Mathematica using the eigen values from the Hessian Matrix of the input image using Sobel Derivatives.
​
​
-
The principle behind the ridge filter lays in the idea that pixel values can be locally approximated with a second order polynomial.
-
The second order derivative matrix, also known as the “Hessian Matrix”, describes the ridge like structures.
-
This can be transformed into a sum of two parabola rotated by a certain angle, by the eigen-decomposition of this Hessian Matrix.
-
We look for pixels where the major eigenvalue of the Hessian matrix is larger than the minor. The simplest way of doing this would be to calculate the major eigenvalue at each pixel.
-
The code is now part of the OpenCV-contrib module and is available open source for use under their license norms.
