applying 1x2 operator to my image: cannot find anything or any syntax

Question

Sat m 2013-2-20

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/64314-applying-1x2-operator-to-my-image-cannot-find-anything-or-any-syntax

i am able to apply sobel and prewitt operator to my grayscale image. but when i put in matlab

help edge

then it displayed syntax and format for sobel, prewitt, roberts, laplacian, zero cross and canny. but i could not find any syntax for 1x2 operator. is 1x2 operator a part of any of these above operators? please let me know syntax for 1x2 operator

4 个评论
显示 2更早的评论隐藏 2更早的评论

Jan 2013-2-20

编辑：Jan 2013-2-20

You cannot find the 1x2 operator anywhere, but you ask for it. I see an epistemological problem here. It is like asking for the BlöderNasenMann operator. (Kind regards to Thorsten, who has invented it as counterpart to the dull named Quabla operator some years ago without defining the mathematical details.)

I've (this means: my favorite internet search engine) found the link at stackoverflow also, but in opposite to Walter, I did not recognize that this solves your problem already. @Walter: Kotow.

Image Analyst 2013-2-21

在 MATLAB Online 中打开

So if you want it, and know of the "features" or drawbacks of it, then go ahead and use it:

filteredImage = conv2(double(grayImage), [-1 1], 'same');

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

Walter Roberson 2013-2-20

1
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/64314-applying-1x2-operator-to-my-image-cannot-find-anything-or-any-syntax#answer_75904

Perhaps,

http://stackoverflow.com/questions/5776975/what-is-the-difference-between-1x2-and-1x3-image-gradient-kernel-filter-definiti

0 个评论
显示 -2更早的评论隐藏 -2更早的评论

请先登录，再进行评论。

Answer 2

Image Analyst 2013-2-20

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/64314-applying-1x2-operator-to-my-image-cannot-find-anything-or-any-syntax#answer_75912

编辑：Image Analyst 2013-3-4

There is no 1x2 kernel for Prewitt and Sobel filters. That's not how they're defined. There is a Roberts filter like that though.

edge() applies the filters and then thresholds and thins the result to give skeletonized lines along the edges. I wish there were an option to return the original edge filtered image, but unfortunately there is not. So if you want the true Sobel filter you will have to program it yourself with imfilter() or conv2(), or if you have R2012b you can use imgradient() or imgradientxy().