Using the HarrisSIFT techniques I'm able to identify and match key points, but I would like to calculate a homography using RANSAC between the two images, and actually apply that transformation to the second image, in such a way that it would now overlay the original image as perfectly as possible.
I don't really know where to start so any pointers would be fantastic (functions to use, etc).
the example IVT/examples/HarrisSIFTDemo shows the use of the HarrisSIFT features together with a RANSAC approach. All is left to do is use the functions LinearAlgebra::DetermineAffineTransformation(…) or LinearAlgebra::DetermineHomography(…), and afterwards ImageProcessor::ApplyHomography(…).
I hope this helps.
That's fantastic, I will try your suggestions shortly.
Also, do you know of anyone having tried to use IVT on android? I'm not really sure whether that's even possible or what it would entail so if you do know of any resources on this subject that would certainly be very useful.
I don't know if anybody has tried, and unfortunately, I am not familiar with implementing software for Android. The only thing I can say is that if it is possible to compile C++ source code for Android, then the IVT core library will compile and run without any problems. The video capture modules are camera/platform specific and will not work, but it should be simple to implement an IVT video capture module for accessing the camera(s). And the GUI toolkit will only work in case Android supports Qt or GTK.
Log in to post a comment.
Sign up for the SourceForge newsletter:
You seem to have CSS turned off.
Please don't fill out this field.