Dear Sir or Madam,
I am Rui Zeng, a PhD candidate from Queensland University of Technology. I am now working on a research project which is about panorama generating from a video. Here is my homepage https://sites.google.com/site/ruizenghomepage/.
Suppose that I got 10 consecutive frames (f_1, f_2, …, f_10) that extracted from the video and then these 10 frames are sent into Hugin to generate a panorama (pano). May I ask you guys the question about how to extract homography matrix from Hugin? Assume that f_1 is an anchor frame and we use an incremental method to build the panorama. Would it be possible to extract Homography matrix H_1, H_2, .., H_9 from Hugin, where f_2 = H_1*f_1, f_3 = H_2 * f_2, …, f10 = H_9 * f_9. In other words, how can we get a mapping from the pixel (x_1, y_1) in the frame f_1 to the pixel (x_p, y_p) in the pano. The pixel (x_2, y_2) in the frame f_2 to the pixel (x_p, y_p) in the pano.
Actually I walked through this image stitching code https://github.com/daeyun/Image-Stitching carefully and know the rough process of image stitching. Does Hugin has some similar functions or API to output what I asked above? By the way, could anyone provide me a baseline project like image stitching code https://github.com/daeyun/Image-Stitching as a learning example. I searched on the Google and cannot find a proper tutorial example like http://www.pyimagesearch.com/2016/01/11/opencv-panorama-stitching/
I believe aforementioned problems must be met by many researchers who work on panorama generating.
Is there anyone to do me a favor. Actually, as a student, I do not have much things to offer to givers who help me. I would like to give him an apple store card as an appreciation cause this research project is really important to me and I believe that rewarding contributors is a good way to build a good community and this is a respect to their contribution. Please feel free to contact me.
Thanks for reading.
King regards,
Rui Zeng