panorama

[digital image processing]

hello,

i am using labview 7.0 se ,imaq 3.0 .
could u please help me how to register multiple images to create a panorama preferably infrequncy domain.

[Scott_R]

Hello,

When you say that you want to create a panorama, do you mean that you would like to concatenate the images? If so, the following two links may be of some help:

Make a panorama image from different images
http://forums.ni.com/ni/board/message?board.id=200&message.id=4621&requireLogin=False

Concatenating NI-IMAQ Acquisition Images
http://venus.ni.com/stage/we/niepd_web_display.DISPLAY_EPD4?p_guid=B45EACE3D91356A4E034080020E74861&p_node=DZ52489&p_submitted=N&p_rank=&p_answer=&p_source=Internal

Are you using a National Instruments framegrabber board to acquire the images or are you working with existing image files? What do you mean by registering the images? You can either acquire the images from a framegrabber board by using the IMAQ VIs or, if you have the Vision Development Module, you can open the image files from your hard drive.

Regards,
Scott R.
Applications Engineer
National Instruments

[digital image processing]

thank you for that valuable information,
i am just trying out them,

and the link you sent :
http://venus.ni.com/stage/we/niepd_web_display.DISPLAY_EPD4?p_guid=B45EACE3D91356A4E034080020E74861&p_node=DZ52489&p_submitted=N&p_rank=&p_answer=&p_source=Internal


doesnt seem to be working could u help me with that.
thank you.

[Scott_R]

Hello,

I have attached the example program that was included on that page. Try running it and let me know if you have any additional questions.

Regards,
Scott R.
Applications Engineer
National Instruments

[digital image processing]

i `ve tried with the lib file u `ve sent me .
but this file seems to need one more file
imgcreatebuffer.vi
where do i find this file.

sundar

[Scott_R]

Sundar,

It looks as though that VI is designed to work with previous versions of LabVIEW, Vision and IMAQ. I am currently looking into alternative solutions for this problem. Those VIs used to ship with Vision, but do not anymore. There are most likely newer VIs that perform the same functions, but they may be named differently. I will let you know just as soon as I have discovered anything.

Scott R.
Applications Engineer
National Instruments

[Scott_R]

Sundar,

We may be able to simplify the process by simply using an IMAQ Create to create the final image and then use Set Image Size to determine the final dimensions of the image. We could then use IMAQ Image to Image to copy existing images into the final image, using an offset that changes based on which image we are adding. For example, if we were using three images that were 640x480 and we wanted the final image to be the three images next to each other, the final image would be 1920 x 480. The first offset would be [0,0], the second would be [640, 0], and the third would be [1280, 0]. You could use a loop structure with this such that you only use one instance of the Image to Image function but the offset changes based on which iteration of the loop you are on (multiply the iteration by 640).

Try implementing an example based upon that and let me know if you have any additional questions on this issue.

Scott R.
Applications Engineer
National Instruments

[digital image processing]

respected sir,
i would like to just know the preprocessing stage
wherein i need to find the common region ,
so i that i could proceed further to
concatenate the images.

regards,
sundar,

[Scott_R]

Sundar,

You would need to implement an algorithm that would check the borders of the image and then look for regions that overlap in other images. The VI that I pointed you to before did not perform that processing, as it merely attached one image to the edge of another. It did not look for overlapping image portions.

To do this, I would suggest a similar algorithm to the one that I already mentioned, only you would need to perform some pattern matching before adding each new image. When you add the first image, create a template from the edge of the image and then look for that template within the next image, if it exists, then you know the point at which you can start overwriting the existing image. So instead of incrementing the start point by every 640 (as in the previous algorithm) you would increment it by a varying amount based on where the pattern was found. After adding each image, you would create a new template and search for it in the next image, etc.

Try implementing that algorithm in your code and let me know if you have any additional questions on this issue.

Regards,
Scott R.
Applications Engineer
National Instruments