Image fusion problem

Our LEADTOOLS Medical imaging has full support for the latest DICOM standard. This includes the ability to load, modify and save DICOM datasets.
We have C DLL APIs and also C++ Classes (in addition to our .NET classes).
About your requirements, what exactly do you mean by "merge" the images? Do you want to remove the original pixel data from the data set and add a different image to it? If not, please explain in details.

If you don't have LEADTOOLS, I suggest downloading the free evaluation edition from our website.

Thanks,
Maen Badwan
LEADTOOLS Technical Support

You can use Combine method for image fusion. However, you will need to convert 12-bit signed CT to 24-bit color and 16-bit grayscale PET to 24-bit color (replace the gray LUT entry with values stored in the same bin in color palette and color res to 24) before combining.

Thanks,
Maen Badwan
LEADTOOLS Technical Support

Hello,

The best example that we have of the combine command would be in the C DLL"Image Processing Features" demo that ships with the SDK. You can find this demo through the shortcuts here:
C:\LEADTOOLS 17.5\Shortcuts\C DLL\01 Imaging\04 Image Processing\Image Processing Features

Once you load the two images, you can bring up the combine dialog with this:
Filters me nu -> Calculation -> CombineExt menu item

This can also be seen with the .NET "Image Processing" demo:
D:\LEADTOOLS 17.5\Shortcuts\.NET Class Libraries\.NET Framework\01 Imaging\03 Image Processing\Image Processing
Under the base category, there is a Combine Fast (CombineFastCommand) option.
Under effects, you can also see the Combine (CombineCommand) option

I would also recommend checking out the FeatherAlphaBlendCommand, as this command does a combine of two images using a grayscale mask. This command is basically an extension of the AlphaBlendCommand, which combines two images with a fixed opacity across the entire image. Using the grayscale mask, you can specify how much opacity to use with each pixel; the more black the pixel in the mask is, the more of the underlying image bleeds through to the image on top. The more white the pixel is, the more of the top image is seen as compared to the bottom.

There isn't any methods with our toolkit that can help you line up the images so that you're combining them in the right spot. This calculation would need to be done before any of the aforementioned combine methods can be used. Once you've figured out where to do the combining though, these methods are how you can do it.