how can we detect horizontal and vertical edges separately in an image ?
Asked by Vasundhara Chari on 17 Aug 2012
Latest activity Edited by Walter Roberson on 21 Aug 2012
i am doing a project on denoising.i am using modified laplacian kernel.i need to show after so many iteratiins the vertical edges are preserved but artifacts are produced in the diagonal edges.so how would i first extract the hor and vert edges separately?
[Merged information from duplicate question]
i am doing project on denoising.i am using an edge adapting laplacian kernel.i need to detect orthogonal(horizontal and vertical) and diagonal edges from the image and show the results of applying the kernel on them.the results for ortho and diag edges are different.
Products
No products are associated with this question.
1 Answer
Answer by Image Analyst on 17 Aug 2012
From the help for edge():
[BW,thresh,g45,g135] = edge(I,'roberts',...) returns 45 degree and 135 degree edge responses to Roberts gradient operators. You can also use these expressions to obtain gradient responses:
For horizontal and vertical edges, you can just do something like:
horizontalEdgeImage = imfilter(grayImage, [-1 0 1]); verticalEdgeImage = imfilter(grayImage, [-1 0 1]');
0 Comments
0 Comments