Talk:Mathematical morphology

Overhaul

I would be interested in seeing this article (and morphological image processing) fleshed out a little. I have studied morphological image processing as applied to binary images, and the extension to greyscale and in particular how it relates to lattice theory is far from obvious. CyborgTosser (Only half the battle) 18:13, 29 May 2005 (UTC)[reply]

I have just made a modification to this article; and intent it to be a first of a number of modifications. I have extensive experience on this field, and find the page to be kind of poor at this point, so I intend to add material substantially over time.Renatokeshet (talk) 06:47, 16 May 2008 (UTC)[reply]

By now, I've finished most of the over-haul I intended for this page. Even though I believe there is still a lot of work to do (input data on applications of mathematical morphology, make pages for granulometries, dilation and others, etc.), I believe that the page is now in a format that better suits Wikipedia. Hope it helps people, and also that people edit it. Renatokeshet (talk) 11:44, 24 June 2008 (UTC)[reply]

The figure that illustrates erosion is not correct: the center of the eroding disk should walk on the boundary of the outer rectangle. Sprocedato (talk) 14:11, 13 August 2009 (UTC)[reply]

The figure that illustrates erosion is correct and exact. Renato (talk) 14:51, 8 August 2010 (UTC)[reply]

The section although correct is hard to understand. I found the following link to be very useful in describing the concept. It seems it is written as a refresher for the academic.

[1]

Attention needed

Redlinks - fix

Chaosdruid (talk) 04:16, 7 August 2010 (UTC)[reply]

I'm not sure I understood. Your are saying the article needs immediate attention because it has red links? Renato (talk) 15:02, 8 August 2010 (UTC)[reply]

Yes - that is what is needed to be fixed immediately. Links are placed to either explain further/clarify points in the article or to link to further reading. Anyone who reads this page will not have that facility available as the links do not work. It is imperative that people can follow links and so they need fixing urgently to ensure that an average reader can understand the material by being able to have further information explained.

There are also other points but they are not immediate. An article riddled with redlinks is one which is obviously not getting a lot of attention.

Chaosdruid (talk) 19:13, 8 August 2010 (UTC)[reply]

I'm curious to find out how one immediately fixes an article riddled with red links. Renato (talk) 21:01, 8 August 2010 (UTC)[reply]

Either by delinking them or finding another article to link to which partially or wholly provides the necessary material. Chaosdruid (talk) 21:37, 8 August 2010 (UTC)[reply]

Delinking just because of red links is certainly something that should not be done in Wikipedia. The solution is to write the articles that are red, which is somethings that cannot be done immediately.

I think it's a bit of an exaggeration to mark this article for immediate attention on the grounds of red links. There are 164 links in the article, from which only 12 are red, and 3 of them are explained in the text immediately following the link. The Lead has no red links, and in fact, up to the middle of the article there are only 4 red links. That's hardly an article "riddled with red links". I wonder how many articles in Wikipedia has less than 7% red links in it.

I hope that what is bothering you will be addressed by the overhaul we discussed below. Meanwhile, I would like to suggest that you remove the need for attention. Thanks. Renato (talk) 22:05, 8 August 2010 (UTC)[reply]

Delinking because of red links is definately something that should be done if the links have been there for a while and it is unlikely that the articles are to be created soon. If an article is to be created later then the link can be created as long as that article is eventually created and two years seemed like a pretty long time. It is one thing to create a redlink to an article you are going to create next week, or one that you hope will eventually be created, but it is another to find redlinks there that have been there for years without any articles having been created.

As for the redlinks I had mentioned - it says fix. That can be linking to other articles than originally intended or creating the articles that were intended but the important thing is that something needed to be done as the timeframe was pretty long. There had been no discussion on here for 11 months, so forgive me for thinking that they had been left here and fogotten about lol, and during my second sweep of the Robotics project articles I found that it had not been looked at since my last visit 3 weeks previously. There were 14 or 15 redlinks and spread throughout so riddled was probably accurate - more importantly they have been there for a long time [2], [3], [4] and the oldest (2 years ago) [5].

It is good that at least now some more work has been done on here - so maybe the "attention needed" served its purpose ? As for removing it, I have no problem with that now that work is progressing. I understand that you put those links there two years ago and probably were fully intending to come back and work on them, as do I understand the problems of having RL to deal with and that it is not always easy to get back to things. Do not kill the messenger though :¬)

Chaosdruid (talk) 00:10, 9 August 2010 (UTC)[reply]

Plan for new overhaul

I'm planning to do a new overhaul to this article. It will take a couple of months, but here is the plan:

Move all the mathematic details of the basic morphological operators (erosion, dilation, opening, and closing), and their definitions on different spaces (binary, grayscale and lattices), to a new page: Basic morphological operators.
Link from this article to the new one using {{Main}}.
Leave only a general description of the basic operators here, with examples.
Elaborate on the composite operators (hit-or-miss transform, gradient, watershed, etc...)
Provide examples of applications.

Even after the overhaul, there will be still a lot of work in illustrating and expanding the examples. But then the article will be somewhat more directed to the general audience, rather to experts in image processing. Comments/remarks/ideas/objections? Renato (talk) 16:15, 8 August 2010 (UTC)[reply]

If you are intending to massively expand a section then fair enough, but to leave just a stub here would only move the problem from here to there. I don't think that that is really necessary as the page is not that big at the moment.

All that is really needed, right now, is an explanation for what the technique is/does in laymans/college language to link the reason for the articles existence to the real world functions to which it applies. I, for one, would really like to see your expertise directed more at helping get all the articles that lack this fundamental link between PhD level and layperson/college explanation fixed. It is merely a case of writing a few paragraphs in the style of "I am talking to my sixteen year old child and explaining how it works to them"

A large proportion of the AI, algorithm, machine learning and similarpages are not explained very well to the reader. I have been examining correlations between a couple of sources and the articles in Wiki as it appears that there may have been some Copyvio a couple of years ago. The articles are in magazines aimed at the already knowledgeable in the field, at MSc or PhD levels, and this would explain why the articles have little or no attempts at explaining the basics for the layperson or even higher school/college educated.

For example the reader asks themselves "But what is this used for?" - where in the article is this information?

Chaosdruid (talk) 19:28, 8 August 2010 (UTC)[reply]

I'm not intending to leave a stub, but to clean the space to "elaborate on the composite operators and to provide example and applications", as I wrote above. Also, as I wrote, my intention is to "direct the article more to the general audience, rather to experts in image processing". Having that in mind, do you have any objections to the plan as described above? Clarification: I don't know about other articles in "the field" - I wrote this one, and only this one, and no "Copyvio" took place here. Renato (talk) 20:54, 8 August 2010 (UTC)[reply]

Fair enough - it was the comments "Move all the mathematical details" and "Leave only a general description" which gave me the impression that not much would be left here after the changes you were proposing. Now you have explained in more detail it seems more reasonable that there will be enough material remaining.

I am not saying that you did any copyvio - merely that as you are involved in the field you may be able to help with the task of adding explanation and examples. There is a great shortage of experts in the fields I mentioned as many of them have had COI issues raised which can cause problems with editors jumping on them and scaring them away. Chaosdruid (talk) 21:45, 8 August 2010 (UTC)[reply]

So, I'll start working on it. But, as I said, it will take some time. Meanwhile, can you point me to one or more of these COI issues? I'm curious about what have happened. Renato (talk) 21:49, 8 August 2010 (UTC)[reply]

I have a trouble with the equations of the dilation in grayscale morphology: isn't it $(f\oplus b)(x)=sup_{z\in B}f(x-z)=sup_{z\in B^{s}}f(x+z)$ rather than $=sup_{z\in B^{s}}f(z-x)$ ? It does not seem to be coherent with the first grayscale equation of the dilation (if we set z = x - y). Plus $sup_{z\in B}f(x-z)$ is coherent with the chap2 of the slides of Jean Serra (p.19). Same for the erosion. What do you think ? RaffiEnficiaud (talk) 23:43, 26 August 2010 (UTC)[reply]

I agree, the dilation equation looks erroneous to me. Please comment on my suggested correction in the talk page here http://en.wiki.x.io/wiki/Talk:Dilation_%28morphology%29 161.84.227.13 (talk) 16:23, 21 August 2014 (UTC)[reply]

Do you think the sentence "Dilation is the opposite of the erosion" is accurate ? One may think about a reversible process, which the pair ero/dil does not meet (both transforms loose some information either on the foreground or the background) RaffiEnficiaud (talk) 23:54, 26 August 2010 (UTC)[reply]

I think there should be some numeric examples, particularly of grayscale morphology. The notation isn't particularly clear- consider dilation for example. f(x) is defined as an image, but in the equation, f(y) is used, followed by f(x-z). y is defined as an element of E, which is further defined as any real in a Euclidean space. So, to me the implication is to take some function f, apply it to every real that is a member of a Euclidean space, and then take the same function and apply it to x-z. Do that for all y, and take the highest sum, and do what with it? I think a rudimentary example with a small structuring element and numeric image would help clarify. Raligan (talk) 14:06, 9 July 2014 (UTC)[reply]

External links modified (January 2018)

Hello fellow Wikipedians,

I have just modified 2 external links on Mathematical morphology. Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:

When you have finished reviewing my changes, you may follow the instructions on the template below to fix any issues with the URLs.

This message was posted before February 2018. After February 2018, "External links modified" talk page sections are no longer generated or monitored by InternetArchiveBot. No special action is required regarding these talk page notices, other than regular verification using the archive tool instructions below. Editors have permission to delete these "External links modified" talk page sections if they want to de-clutter talk pages, but see the RfC before doing mass systematic removals. This message is updated dynamically through the template {{source check}} (last update: 5 June 2024).

If you have discovered URLs which were erroneously considered dead by the bot, you can report them with this tool.
If you found an error with any archives or the URLs themselves, you can fix them with this tool.

Cheers.—InternetArchiveBot (Report bug) 11:23, 21 January 2018 (UTC)[reply]