A pointer to the Machine Vision functions list

[Marce]

Hello,

I'd like to know if the Machine Vision module does the image processing that I need and I couldn't find any functions list in the web site.

I'm working on aerial images and I should do geometric corrections (not affine) and mosaicing.

Any comments or pointer will be very welcome.

Thanks,

Marce

[BruceAmmons]

It could be done with the Vision toolkit, but it would not be easy. There are no functions for doing what you want, so you would have to create your own.

I would recommend looking for a package that does this sort of thing for you. There are a lot of tools available for creating mosaics from multiple images.

Are you trying to make this an automated process, or do you just need to be able to manually combine the images?

Bruce

[Marce]

Thanks for your response.

In this case I'm not trying to automate the process. I need a tool to do the geometric correction since I will have a series of aerial photos and the operator will have to match points in the photos with reference points in a basis map. I'm writing the operator interface with LabView but I need the specific tool. Once the operator has identified at least 4, but eventually more, reference points, I need a tool that will correct the image according to the location of the reference points in the map. Most of the packages offer an affine correction that is too simple for photos. The packages that has the right corrections, have so many other things for sophisticated geoprocessing that are usually too expensive for my budget.

Thanks,

Marce

[chilly_charly]

Hi Marce !
I have exactly the same problem, that as to be solved within the next four weeks, and may be we could share...
I'm not a newby in labVIEW, but I am in geodesy, definitely !!!

What I intended to do was to create a triangular mesh on the photo, from a series of points of known coordinates, then to do an affine transformation inside each triangle. Of course this may produce some distorsion, and I would like to find a better approach.
Here is what I understand :
Suppose A, B, C on a straight line, B being between A and C. To move B toward C, you must dilate AB and contratc BC. Means that there will be a discontinuity of the derivative around B. OK, that's the problem. I don't know where to go from there : May be I could use a polynomial (2nd degree) to maintain the continuity of the first derivative ? But what will be the resulting complexity on a 2 D mesh ?
The second problem is to manage the parts of the surface outside the triangular mesh : should I extrapolate the mesh deformation to the outside ? Or should I use information from the adjacent photos, when any ?

Could we discuss this further ?

CC

[Marce]

CC,

Thanks for your response.

The problem is that I'm new to LV but i'm also not an expert in geodessy. And mainly in the next few I have to solve other things that won't let me deal with that problem with the necessary effort. Let's move to regular email so we don't bother the other forum members with photogrametry issues that has nothing to do with LV.

Thanks,

Marce (marcelor@013.net)