Hi there!
just discovered the meshroom product and its very nice!
I am using the "Panorama HDRI" pipeline for the below.
I am trying to use it to combine seven exposure-bracketed 360 degree equirectangular jpeg images into a combined HDRI and hitting an error with the "FeatureMatching" node with the error stating "[info] No image pair to match". Opening the imageMatches.txt text file and it is empty.
So going to the previous node "ImageMatching" the log states "[info] Number of selected image pairs: 0". I have tested each 'method' without success.
Any insight on what might be going wrong here? I don't actually need/want a camera made as i just want to combine into an HDRI... should i be taking a different approach completely perhaps?
Some more info:
- The 360 images were captured on a Samsung Gear 360, and upon dragging the images in they are translated from the double fisheye image to the equirectangular image correctly.
- the seven images were captured on a static tripod with only the exposure being changed between captures
- within the "FeatureExtraction" node i've tried 'ultra' and each of the 'describer types'
logs: