Hi!
I am looking for a feature or plug-in for ImageJ that can skew images (differentially shift the image vertical or horizontal) by a given amount. It should also be able to operate with stacks. It sounds so trivial, that there just must be a solution already. Any ideas if or how this can be achieved? Thanks in advance! |
Gabriel Landini's Align RGB planes plugin (discussed yesterday on this list)
can be used for that, as long as you use layer as red, green and blue components of the input image/stack (your input image have to be monochrome then). Christophe Leterrier On Jan 13, 2008 4:33 PM, BoP <[hidden email]> wrote: > Hi! > I am looking for a feature or plug-in for ImageJ that can skew images > (differentially shift the image vertical or horizontal) by a given amount. > It should also be able to operate with stacks. > It sounds so trivial, that there just must be a solution already. Any > ideas > if or how this can be achieved? > Thanks in advance! > |
In reply to this post by BoP-3
The Image>Translate command, added in v1.39l, does this.
-wayne On Jan 13, 2008, at 10:33 AM, BoP <[hidden email]> wrote: > Hi! > I am looking for a feature or plug-in for ImageJ that can skew images > (differentially shift the image vertical or horizontal) by a given > amount. > It should also be able to operate with stacks. > It sounds so trivial, that there just must be a solution already. > Any ideas > if or how this can be achieved? > Thanks in advance! |
Thanks!
>> Gabriel Landini's Align RGB planes plugin (discussed yesterday on this list) >> can be used for that, as long as you use layer as red, green and blue >> components of the input image/stack (your input image have to be monochrome >> then). I will have a look! > The Image>Translate command, added in v1.39l, does this. > > -wayne Unfortunately it does only translate, not skew. The function would have to translate 0y at the first x-pixel and n pixels at the last x-pixel. (and m/x for the mth pixel in between) ciao BoP > On Jan 13, 2008, at 10:33 AM, BoP <[hidden email]> wrote: > > Hi! > I am looking for a feature or plug-in for ImageJ that can skew images > (differentially shift the image vertical or horizontal) by a given > amount. > It should also be able to operate with stacks. > It sounds so trivial, that there just must be a solution already. Any > ideas > if or how this can be achieved? > Thanks in advance! |
In reply to this post by BoP-3
On Sun, 13 Jan 2008 10:33:25 -0500, BoP <[hidden email]> wrote:
>Hi! >I am looking for a feature or plug-in for ImageJ that can skew images >(differentially shift the image vertical or horizontal) by a given amount. >It should also be able to operate with stacks. >It sounds so trivial, that there just must be a solution already. Any ideas >if or how this can be achieved? The Affine plugin of the TransformJ package does this. http://www.imagescience.org/meijering/software/transformj/ Best regards, Erik |
Thanks Erik!
I am just wondering how this matrix has to be filled, because I thought affine transformations are just rotation translation and scaling, but no distortion. Would it be possible to give an example? (besides the "user interface" is not really straight forward, is it?) Cheers! > On Sun, 13 Jan 2008 10:33:25 -0500, BoP <[hidden email]> wrote: > >> >Hi! >> >I am looking for a feature or plug-in for ImageJ that can skew images >> >(differentially shift the image vertical or horizontal) by a given amount. >> >It should also be able to operate with stacks. >> >It sounds so trivial, that there just must be a solution already. Any ideas >> >if or how this can be achieved? >> > > The Affine plugin of the TransformJ package does this. > > http://www.imagescience.org/meijering/software/transformj/ > > Best regards, > > Erik > > > |
In reply to this post by BoP-3
> I am just wondering how this matrix has to be filled, because I
> thought affine transformations are just rotation translation and > scaling, but no distortion. Roughly speaking, an affine transformation is a transformation that preserves parallel lines. This includes translations, rotations, scalings, skewings, and any combinations thereof. The matrix is just a standard 4x4 matrix in homogeneous coordinates. You can read a little more about it on the plugin page: http://www.imagescience.org/meijering/software/transformj/affine.html > Would it be possible to give an example? Here is an example of a matrix that slightly stretches x as a function of y: 1.0 1.001 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 1.0 and an example of a matrix that stretches y as a function of x: 1.0 1.0 0.0 0.0 1.001 1.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 1.0 Just save those four lines in a text file and specify that file name in the plugin dialog. > (besides the "user interface" is not really straight forward, is it?) I guess that depends on what you mean by "straightforward" in this context. The present interface allows you to keep separate matrix files and easily apply them to images by specifying them in the dialog. In the future I may add a possibility to directly enter matrix numbers into the dialog and save/load them as text files. Good luck and best regards. Erik |
I would like to map an arbitrary 2D closed object into a circle, subject to the
constraint that the perimeter of the object remains constant. For example, I might take a crescent and map it into a circle, such that the local neighborhood in a small arc will be identical (or very close) in both the crescent and the circle, except of course at the "points" of the crescent. Has anyone on this list done such a thing? Thanks in advance --aryeh -- Aryeh Weiss School of Engineering Bar Ilan University Ramat Gan 52900 Israel Ph: 972-3-5317638 FAX: 972-3-7384050 |
Hi Aryeh,
Sounds like an interesting problem, but I'm not sure exactly what you mean. Could you clarify? When you say "map [the object] into a circle," do you mean: 1) Find the circumscribed circle around the object? 2) Find a circle that approximates the object (includes as much of the same area as possible) and has the same perimeter/circumference? 3) Find a circle with a "best fit" to the curvature of the object (e.g., in your crescent example, the exterior arc of the crescent would coincide as much as possible with the edge of the circle)? 4) Something else? I don't think any of the above should be too difficult. Thanks, Curtis On Jan 16, 2008 6:52 AM, Aryeh Weiss <[hidden email]> wrote: > I would like to map an arbitrary 2D closed object into a circle, subject to the > constraint that the perimeter of the object remains constant. For example, I > might take a crescent and map it into a circle, such that the local neighborhood > in a small arc will be identical (or very close) in both the crescent and the > circle, except of course at the "points" of the crescent. > > Has anyone on this list done such a thing? > > Thanks in advance > --aryeh > -- > Aryeh Weiss > School of Engineering > Bar Ilan University > Ramat Gan 52900 Israel > > Ph: 972-3-5317638 > FAX: 972-3-7384050 > |
Free forum by Nabble | Edit this page |