Jump to content

Talk:Erosion (morphology)

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia

Erosion result

[edit]

I really think the result of the erosion of A by B in the article should be as following

    1 1 1 1 0 0 0 1 1 1 1   
    1 1 1 1 0 0 0 1 1 1 1  
    1 1 1 1 1 1 1 1 1 1 1
    1 1 1 1 1 1 1 1 1 1 1 
    1 1 1 1 1 1 1 1 1 1 1   
    1 1 1 1 1 1 1 1 1 1 1  
    1 1 1 1 1 1 1 1 1 1 1 
    1 1 1 1 1 1 1 1 1 1 1  
    1 1 1 1 1 1 1 1 1 1 1  
    1 1 1 1 1 1 1 1 1 1 1 
    1 1 1 1 1 1 1 1 1 1 1 

instead of how it is right now:

    1 1 1 1 0 0 0 1 1 1 1   
    1 1 1 1 0 0 0 1 1 1 1  
    1 1 1 1 0 0 0 1 1 1 1
    1 1 1 1 1 1 1 1 1 1 1 
    1 1 1 1 1 1 1 1 1 1 1   
    1 1 1 1 1 1 1 1 1 1 1  
    1 1 1 1 1 1 1 1 1 1 1 
    1 1 1 1 1 1 1 1 1 1 1  
    1 1 1 1 1 1 1 1 1 1 1  
    1 1 1 1 1 1 1 1 1 1 1 
    1 1 1 1 1 1 1 1 1 1 1 

I already did correct the article on 12th September 2013 but my change was reverted. In my humble opinion, the current state is wrong. As the top row in the resulting image is filled with zeros completely, it is removed entirely. The second and the third top rows both have the three zeros in the middle. However, in the fourth top row, the zero in A is not touched by B anymore. Thus the first two top rows of the result should have the middle zeros while the third row shouldn't.

I hope you agree. If not, I would be happy about an answer about why my way of thinking was wrong. Unless somebody answers within a week, I will change the article to the form proposed by me, assuming nobody has a problem with that. -- 46.244.185.167 (talk) 14:42, 9 December 2013 (UTC)[reply]

Yes, you are correct. People are erroneously thinking of the deleted row when computing the entries around the original zero. 50.173.159.230 (talk) 19:43, 29 August 2014 (UTC)[reply]

Erosion.png confusion

[edit]

Having read the page, I found Erosion.png[1] confusing as it puts the structuring element inside the original image, but if one were to perform the erosion on the original image the structuring element centre would be on the edge. Therefore I feel that the depicting image, whilst in all other respects clear should change to show the "in-process" erosion. — Preceding unsigned comment added by 143.167.5.145 (talk) 14:16, 4 July 2018 (UTC)[reply]

References

Interactive grayscale erosion demo

[edit]

I created an interactive grayscale erosion demo for a blog article about two erosion algorithms applied to procedurally-generated fractal terrain. The demo is at the end of this article: https://www.nablu.com/2021/05/simulating-erosion.html

I improved on the grayscale erosion algorithm described in this Wikipedia article, to simulate actual erosion on a landscape (the algorithm described in the Wikipedia article gives rather strange results, which I illustrate in my blog article).

I don't propose that Wikipedia link to my blog (after all, I have no expertise in this the topic and I am not notable), although I won't object if the community deems it worthy. ~Anachronist (talk) 03:35, 4 May 2021 (UTC)[reply]

Generalization to lattice doesn't make much sense

[edit]

I don't understand the last generalization, and no source where one can look at the source material is provided. If it does make sense, could someone provide a source? I couldn't find anything from a search. 46.162.75.255 (talk) 22:11, 30 March 2023 (UTC)[reply]