search

Kyle-Xu001 / Multi-Depth-Multi-Camera-Stitching

This project consists of various methods for video stitching from multi-cameras to generate a real-time panorama video.
MIT License
35 stars 5 forks source link

Regarding the generation for hpp and spp #5

Open sandeep1404 opened 4 months ago

sandeep1404 commented 4 months ago

Hi Team,

How to generate this hpp and spp file for homography and stitching parameters for running stiching_custom.py for video stitching from multiple cameras? Can you please share this? It is also the question from most of the people who use this repo: can someone please give clarity on this that would be a great help to me. Hoping for a positive reply.

Thank you in advance.

Kyle-Xu001 commented 4 months ago

Hi there. I can provide the hpp and spp files today. However, those files only work for the specific settings of the videos. It still cannot directly work on the general multi-camera videos.

sandeep1404 commented 4 months ago

Thank you for the reply; please share it so that I can understand the format of the files and change it to match my camera settings. I don't really know how to generate those files for my cameras. Do we need any pre-calibration? Based on your files, I can understand what can be done further. Thank you for the response.

sandeep1404 commented 4 months ago

Hi @Kyle-Xu001, Thanks for adding the config files in the repo; I have checked that. What I am trying to do is video stitching, where each video comes from a different camera. Can you please guide me on how we get homography and stitching parameters files to get generated for my case, where I have four videos coming from 4 different cameras and need to stitch them in real time? It would be a great help for my project. Thank you in advance.

Kyle-Xu001 commented 4 months ago

The homography params are directly from the single-frame stitching image.

sandeep1404 commented 4 months ago

HI @Kyle-Xu001, but how do we do the same for video stitching, since we get continuous frames from the video, is it like extract frame 0 from all videos stich and then extract frame 1 from all videos and stitch it and so on, so how do we generate homography parameters for each frame from all the videos and how to set the stitching parameters, please help me with this I am really confused sorry for asking many questions. Thank you for the help.

sandeep1404 commented 3 months ago

Hi @Kyle-Xu001 , Can you please check the issue, I am not able to understand how do we get the params file for my custom video input dataset. I want to stitch the streams coming from four IP cameras in parallely.