Procedure for Image Stitching

Share icon

Tue, 07/29/2014 - 09:42 am

Is there any build-in option to stitch the images in Igor.
The purpose is as follows: I have about 5 to 7 images of angle resolved photoemission measured at different emission angles to cover the whole Brillouin zone. Each image has some region at edges common. So, using that common region, is there any build-in procedure that can be used to search and stitch the images in to single image. I thank you for your interest in advance.
I am not aware of any IGOR procedures for stitching images.

Starting back in the 1990's, the availability of specialized applications based on Pano Tools and similar programs reduced the motivation for developing such tools in IGOR. If you want to stitch images in IGOR you need to start by using ImageRegistration to compute the match and transform images.

A.G.
WaveMetrics, Inc.

July 29, 2014 at 10:45 am - Permalink

I don't have a solution for doing automated image stitching. But for ARPES data, where the rough tilt angle is known, I wrote something years ago which I could use to shift and attach the data to a single matrix step by step. The overlap region was blended (i.e., averaged) at the edges. If you are interested in this solution (which was not very high level but did the job of producing a whole map), I could look up what I did back then. Or are you talking about Fermi-Surface maps, where it's rather a sample surface rotation?

July 29, 2014 at 11:04 pm - Permalink

Firstly, thanks for both of your interest.
Whatever you are mentioning is the one I am looking for. Spectra are in Energy vs kparallel, and I know the angle of tilt by roughly +/-2 degree of maximum error. And not for the Fermi surface mapping.
If you wish to share the idea or code that would be great.

July 30, 2014 at 03:35 am - Permalink

Ok, I have dug up the file. First, I think it would be better to first stitch the waves together in angle scaling and do the k-space conversion later (one reason is that the blending region is not slanted). Note that I wrote this one more than five years ago when I didn't understand that much about Igor (I have shortened a bit today, though). The process is done in an interactive way (i.e., control menus) to eyeball the the right intensity and angle values in order to blend the data nicely. Anyway I have attached the procedure here.
add tilted data.ipf

July 30, 2014 at 07:28 pm - Permalink

chozo wrote:
Ok, I have dug up the file. First, I think it would be better to first stitch the waves together in angle scaling and do the k-space conversion later (one reason is that the blending region is not slanted). Note that I wrote this one more than five years ago when I didn't understand that much about Igor (I have shortened a bit today, though). The process is done in an interactive way (i.e., control menus) to eyeball the the right intensity and angle values in order to blend the data nicely. Anyway I have attached the procedure here.

TRK
Hi Chozo, mslv86
I am facing with the similar problem related to the stacking of ARPES images. First I tried to append the images by removing the data at the overlapping regions but there always appears a tilt, perhaps due to variation in the intensity of two waves. Then I tried to make use of your code. I created a new wave, say wave3, and simply tried to combine the two ARPES images, say wave1 and wave2, in wave3; still the the tilt appears. In addition it does not display the right intensity beyond the overlapping regions for the second image (the same color appears). Since I am not that expert in programming and just a beginner in igor pro, may be I am not using the right command line to run the code. Can you provide some more info on command line so that I can run the code and see how it is working.
Thanks.
Tika

April 28, 2017 at 07:47 am - Permalink